Archives of internal medicine

Handle with 1

EXTREME CARE

This volume is damaged or brittle
and CAWA/OT be repaired!

• photocopy only if necessary

• return to staff

• do not pui in bookdrop

hV: ! 1 .

ARCHIVES

OF

INTERNAL MEDICINE

EDITORIAL BOARD

JOSEPH L. MILLER, Chicago
RICHARD C. CABOT, Boston LOUIS V. HAMMAN, Baltimore

GEORGE DOCK, St. Louis WARFIELD T. LONGCOPE, New York City

W. S. THAYER, Baltimore

Volume 29
1922

CHICAGO

AMERICAN MEDICAL ASSOCIATION

PUBLISHERS

//

tc;a f^. I soy ^^

A.

100071^

CONTENTS OF VOLUME 29

JANUARY. 1922. NUMBER 1

PAGE

Motor Phenomena Occurmng in Normal Stomachs, in the Presence of
Peptic Ulcer and Its Pain, as Observed Fluoroscopically. Lawrence
Reynolds, M.D., and C. W. McClure, M.D., Boston 1

Clinical Observations on the Capillary Circulation. S. O. Freedlander,
M.D., and C. H. Lenhart, M.D., Cleveland 12

The Protein Requirement in Tuberculosis. William S. McCann, M.D.

(With the Technical Assistance of Estelle M.\gill), New York) . . 33

Observations on the Use of Quinidin in Auricular Fibrillation.
J. A. E. Eyster, M.D., .\nd G. E. Fahr, M.D., Madison, Wis 59

Tuberculosis of the Heart, with the Report of Two Cases. Edward
Weiss, M.D., Philadelphia 64

Observations Following Intravenous Injections Of Hypertonic Salt
Solutions in Cases of Neurosyphilis. James Wynn, Boston 72

A Metabolic Study of Progressive Pseudohypertrophic Muscular
Dystrophy and Other Muscular Atrophies. R. B. Gibson, Ph.D.,
Francis T. Martin, B.S., and Mary Van Rennselaer Buell, Ph.D.,
Iowa City 82

The Nitrogen Requirement for Maintenance in Diabetes Mellitus.
Phil L. Marsh, M.D. ; L. H. Newburgh, M.D., and L. E. Holly, M.D.,
Ann Arbor, Mich 97

Mycotic Embolic Aneurysms of Peripheral Arteries. deWayne C.
Richey, B.Sc, M.D., and W. W. G. MacLachlan, M.D., CM.,
Pittsburgh 131

Book Reviews 141

FEBRUARY. 1922. NUMBER 2

Tr.'\cheal and Bronchial Stenosis as Causes for Emphysema. C. F.
Hoover, M.D., Cleveland 143

A Study of Microlvmphoidocytic Leukemia, with the Report of a
Case. Solomon Fineman, M.D., M.A., Minneapolis 168

Intracutaneous Reactions in Lobar Pneumonia. George H. Bigelow,
M.D., Boston 221

Clinical Studies on the Respiration. VIII. The Relation of
Dyspnea to the Maximum Minute-Volume of Pulmonary Ven-
tilation. Cyrus C. Sturgis, M.D.; Francis W. Peabody, M.D.;
Francis C. Hall, M.D., and Frank Fremont-Smith, Jr., M.D.,
Boston 236

Position and Activities of the Diaphragm as Affected by Changes
of Posture. Roy D. Adams, M.D., and Henry C. Pillsbury, M.D.,
Major, M. C, U. S. Army, Washington, D. C 245

AURICULOVENTRICULAR RhYTHM AND DiGITALIS. HeNRY B. RiCHARDSON,

M.D., New York 253

A Case of Disseminated Miliary Tuberculosis in a Still-Born Fetus.

R. C. Whitman and L. W. Greene, Boulder, Colo 261

A Convenient Electrode for Experimental Electrocardiographic

Work. Carl S. Williamson. M.D., Rochester, Minn 274

Book Review 276

COXTEXTS OF VOLUME 29
MARCH. 1922. NUMBER 3

PAGE

Clinical Studies ox the Respiration. IX. The Effect of E.\ercise
OK THE Metabolism, Heart Rate and Pul.monary Ventil.^tion
OF Normal Subjects and Patients with Heart Dise.\se. Francis
W. Peabody, M.D., and Cyrus C. Sturgis, M.D. With the Assis-
tance of Berth.\ I. Barker and Margaret N. Read, Boston 277

Studies in Diabetes Insipidus. Water Balance, and Water Intoxi-
cation. Study I. James F. Weir, M.D.; E. Eric Larson, M.D..
and Leonard G. Rowntree. M.D., Rochester, Minn 306

Circulatory Compensation for Deficient Oxygen Carrying Capacity
OF the Blood in Severe Anemias. George Fahr and Ethel
Ronzone, Madison, Wis 331

A Hitherto Undescribed Tumor of the B.ase of the Aorta. George
R. Herrmann, M.D., and Montrose T. Burrows, M.D., St. Louis.. 339

Bacillus Acidophilus and Its Therapeutic Application. Leo F.
Rettger, Ph.D., and Harry A. Cheplin, Ph.D., New H.wen, Conn. 357

The Effect of Blood 'from the Carotid Artery of the Dog and Its
Expression by a General Empirical Formula. Halbert L. Dunn.
B.A., M.A., Minneapolis 368

Study of Blood Sugar Curves Following a Standardized Glucose
Meal. W. H. Olmsted. M.D.. and L. P. Gray, M.D., St. Louis.... 384

Book Reviews 401

APRIL. 1922. NUMBER 4

Observations on Paroxysms of Tachycardia. H. M. Marvin, M.D..

and Paul D. White, M.D., Boston 403

Renal Glycosuria. D. S. Lewis, M.D., Montreal, Canada 418

Chemical Studies of the Blood and Urine of Syphilitic P.\tients

Under Arsphenamin Treatment, with a Note on the Mechanism

of Early Arsphenamin Reactions. Charles Weiss, Ph.D., and

Anna Corson, B.Sc, Philadelphia 428

Studies in the Variation of the Length of the Q-R-S-T Interval.

G. K. Fenn, M.D., Chicago Heights, III 441

Postoperative Pulmonary Complications. Elliott C. Cutler, M.D.,

and Alice M. Hunt, R.N., Boston 449

The Pathology of Cirrhosis of the Liver. An Historic-Pathologic

Study. Frederick Epplen, M.D.. Spokane, Wash 482

The Antidiuretic Effect of Pituitary Extract Applied Intranasallv

IN a Case of Diabetes Insipidus. Herrmann L. Blumgart, M.D.,

Boston 508

The Vital Capacity in a Group of College Students. A. W.

Hewlett, M.D., and N. R. Jackson, M.D., San Francisco 515

The Length of Life of Transfused Erythrocytes in Patients with

Primary and Secondary Anemia. Joseph T. Wearn, M.D., Sylvia

Warren and Olivia Ames, Boston 527

Blood Pressure and Pulse Rate Levels. First Paper: The Levels

Under Ba^al and Daytime Conditions. T. Addis, San Francisco 539
Book Reviews 554

COXTEXTS OF VOLUME 29
MAY. 1922. NUMBER S

PAGE

Study of Some Cases of Diabetes Insipidus, with Special Reference
TO the Detection of Changes in the Blood When Water Is
Taken or Withheld. C. D. Christie, M.D., and G. N. Stewart,
M.D., Cleveland 555

Clinical Calorimetry. XXX. Metabolism in Erysipelas. Warren
Coleman, M.D. ; D.avid P. Barr, M.D., and Eugene F. Du Bois,
M.D. With the Technical Assistance of G. F. Soderstrom,
New York ■. 567

Clinical Calorimetry. XXXI. Observations on the Metabolism of
Arthritis. Russell L. Cecil, M.D. ; David P. Barr, M.D., and
Eugene F. Du Bois, M.D. With the Technical Assistance of
G. F. Soderstrom and Estelle Magill, New York 583

Clinic.\l Calorimetry. XXXII. Temper.^ture Regulation after the
Intravenous Injection of Proteose and Typhoid Vaccine. David
P. Barr. M.D. ; Russell L. Cecil, M.D., and Eugene F. Du Bois,
M.D. With the Technical Assistance of G. F. Soderstrom,
New York 608

A Correlated Study of the Indications for Tonsillectomy and of
the P.\thology and Bacteriology or the Excised Tonsils.
Leonora Hambrecht, B.S., and Franklin R. Nuzum, M.D., Santa
Barbara, Calif 635

Blood Pigment Metabolism and Its Rel.\tion to Liver Function.
Chester M. Jones, M.D., Boston 643

A Study of the Hemoglobin Metabolism in Paro.xysmal Hemo-
globinuria, WITH Observations on the Extrahepatic Form.\tion
OF Bile Pigments in Man. Chester M. Jones, M.D., and B.\sil B.
Jones, M.D., Boston 669

Studies of the Cause of Pain in Gastric and Duodenal Ulcers.
II. Peristalsis as the Direct Cause of Pain in Gastric Ulcers
WITH .'^chylia and IN DuoDENAL Ulcers. Leo L. J. Hardt, M.D.,
Rochester, Minn 684

The Se.-\t of the Emetic Action of the Digitalis Bodies. Robert A.
Hatcher and Soma Weiss, New York 690

The .'Klkali Reserve in Pulmonary Tuberculosis. David S. Hachen,
B.S., M.D., Cincinnati 705

JUNE, 1922. NUMBER 6

Pig.ment Met.\bolism and Regeneration of Hemoglobin in the Body.

G. H. Whipple, Rochester, N. Y 711

Ochronosis: With a Study of an Additional Case. B. S. Oppen-

heimer, M.D., AND B. S. Kline, M.D., New York 732

Aids to Basal Metabolic Rate Determinations. H. S. Newcomer,

M.D., Philadelphia 748.

The Nature of the So-Called "Capillary Pulse." Ernst P. Boas,

M.D., New York 763

The Etiology and Development of Glomerulonephritis. E. T. Bell,

M.D., and T. B. Hartzell, M.D., Minneapolis 768

Biochemical Studies in a F.\tal Case of Methyl Alcohol Poisoning.

I. M. Rabinovitch, M.D., Montreal 821

COXTEXTS OF VOLUME 29

JUNE, 1922— Contmued

PAGE

Reversed Rhythm of the Heart. Morris H. Kahn, M.D., New York 828

Some Observations ox Paroxysmal Rapid Heart Action, with Special
Reference to Roentgen-Ray Measurements of the He.\rt in and
Out of Attacks. Samuel A. Levine, M.D.. and Rose Golden, M.D..
Boston 836

A Study in Experimental Diabetes. The Effect of Intravenous
Injection of Pancreatic Perfus.\tes on the D/N Ratio Follow-
ing Pancre.atectomy. Herbert E. Landes, A.B.; Lester E. Gar-
rison, S.B., and James J. Moorhead, M.D., Chicago 853

Book Review 867

Archives

of Internal

Medicine

VOL. 29

JANUARY. 1922

No. 1

MOTOR PHENOMENA OCCURRING IN NORMAL

STOMACHS, IN THE PRESENCE OF PEPTIC

ULCER AND ITS PAIN, AS OBSERVED

FLUOROSCOPICALLY *

LAWRENCE REYNOLDS, M.D., and C. W. McCLURE, M.D.

BOSTON

This communication embodies the principal results of fluoroscopic
observations of the stomachs of normal men and of patients with ulcer
of the stomach or duodenum, after the feeding a meal composed of meat
and barium, which we have recently made. The ulcer patients were
studied in order to obtain detailed information regarding gastric motor
phenomena occurring throughout the period in which the stomach was
emptying itself and during the occurrence of pain due to the presence
of the ulcer; and, also, to attempt to establish an objective method for
determining the effects of therapeutic measures. The normal patients
were studied to obtain further data as to normal motor activity. It
may seem that there is no need for further roeiitgen-ray observations
on the motor activities of the normal human stomach, for considerable
data ^ are available describing such observations after the ingestion of
various kinds of solid foods. But clinicians and physiologists do not
always seem to be cognizant of this fact as judged from statements
found in current textbooks of medicine. For example, a recent leading
system of medicine contains the statement, "then by means of onward
circular constriction the material is pressed toward the pylorus, but the
pylorus opens only at intervals and not with every peristaltic wave.
In fact, many peristaltic waves will frequently be seen before the
pylorus relaxes." ^ While this statement holds true, accepting Cannon's '
observations, for the stomach of the cat the work of Cole * and our-
selves,= among others, shows that in normal man the pyloric sphincter

♦From the Radiographic Department and the Medical Clinic of the Peter
Bent Brigham Hospital.

1. Carman, R. D., and Miller, A. : The Roentgen Diagnosis of Diseases of
the Alimentary Canal, Philadelphia and London, 1920, p. 110.

2. Rchfuss, M. E.: Oxford Medicine, New York 3:29, 1921.

3. Cannon, W. B. : Am. 1. Phvsiol. 1:.359. 1898.

4. Cole, L. G.: J. A. M. A. 61:762 (March 6) 1913; Am. J. Physiol. 43:
618, 1917.

5. McClure, C. W.; Reynolds, L., and Schwartz, C. W.: Arch. Int. Med.
26:410 (Oct.) 1920.

2 ARCHIVES OF INTERNAL MEDICINE

relaxes as each and every peristaltic wave approaches that orifice. This
observation, alone, indicates the advisability of making more studies
on the stomach of normal man.

For the purposes of the present investigation all subjects were fed
one type of meal. The meal consisted of 160 gm. of finely ground, lean
beef and 40 gm. barium sulphate baked in a loaf. Before feeding it, the
loaf was ground up with sufficient water to make a thick mush, or was
given in its dry state along with from 150 to 200 c. c. of water to drink,
while to one subject it was given in the dry state without water to
drink. The stomachs of the subjects were observed fluoroscopically
immediately after the period of ingestion, a few minutes, then at
fifteen minute intervals for one hour, and then at thirty minute intervals
for two hours and then at hourly intervals until the stomach was nearly
empty, when observations were made at more frequent intervals. Five
normal subjects and sixteen patients with peptic ulcer were studied.

Immediately after the meal had been ingested by the subjects the
stomach was seen to be about three-fourths filled, if observations were
made with the patient in the erect posture. The air bubble in the fundus
occupied the remaining fourth of the gastric cavity. As the stomach
emptied, the upper level of the food gradually became lower until only
the outline of the pyloric region was visible. While this was occurring
that part of the outHne of the stomach which was still visible was not
modified appreciably in shape until but a small residue remained, which
latter formed a hemispherical outline along the greater curvature just
proximal to the pyloric sphincter. At this time, if the patient was
standing, the sphincter was seen to occupy a position above the residue,
the latter would roll back over advancing peristaltic waves and conse-
quently would not be ejected through the sphincter into the duodenum.
However, in the reclining position the residue lay up against the
sphincter and each peristaltic wave forced a portion of it over into the
duodenum, except when pylorospasm was present.

In the stomachs of the five normal subjects as soon as the pyloric
region contained food, peristaltic waves were observed to eject barium
containing chyme through the sphincter into the duodenum, as each
wave approached that orifice, except in one subject. In the latter, gastric
peristaltic waves were very shallow during the first twenty-five minutes
after the meal was ingested, and no barium was seen to enter the
duodenum ; peristalsis then became active and the remaining phenomena
correspomled to that seen in the other subjects. Peristaltic waves began
as shallow indentations at about the junction of the upper and middle
thirds of the stomach's outline and progressively deepened to about
the midpyloric region, where the waves approximately half way bisected
the stomach, while in the immediate neighborhood of the pyloric

REVXOLDS-McCLURE— GASTRIC MOTOR PHENOMENA 3

sphincter the gastric outline was either almost or completely bisected.
Waves began at regular intervals, roughly judged to be twenty seconds
each, but accurate time measurements were not made. To the eye each
peristaltic wave went through the same series of phenomena as it coursed
along the stomach from its origin to the pyloric sphincter. The same
peristaltic phenomena were observed in the subject who ate the meat
and barium loaf in a dry state and without drinking water. In the
. stomach of this subject, during the first thirty minutes after ingesting
the meal, the majority of the latter formed a globular mass below the
air bubble in the fundus, while a smaller portion formed a long, narrow
neck extending from the lower end of the globular mass to the pyloric
sphincter, the whole appearing much like an inverted gourd. The
outline of the stomach appeared very irregular due to the dry state
of the contained food. Barium began to leave the stomach as soon
as the meal was ingested, i. e., a few minutes. Within thirty to forty
minutes after the ingestion of the meal this stomach had assumed the
shape similar to that observed in the other four subjects and had, also,
lost the irregularity of its outline.

Four of the normal subjects occupied the erect position throughout
the period of observation. In three of these the stomach contained a
small residue at the end of five hours. Peristalsis was active at this
time in the stomachs of two of the subjects but was unable to eject
barium into the duodenum for the reasons given. In the stomach of the
third subject, peristalsis was no longer present. A medium sized residue
remained in the stomach of the fourth subject at the end of seven
hours. Peristalsis was active in the stomach of this subject for the
first five hours after the meal was ingested, but during the sixth and
seventh hours of the period of observation it was intermittent and feeble
or absent. The stomach of the fifth normal subject, who was reclining
throughout the entire period of observation, emptied itself in five hours.

In all the normal subjects the first portion of the duodenum filled
out fairly well, but its outline was not as smooth nor its circumference
as great as after the liquid meal ordinarily employed for diagnostic
studies. Roentgenograms of the stomach and intestines of the normal
subjects allowed the meat and barium meal to be followed through the
small intestines and into the cecum. In about forty-five minutes the
jejunum was seen to contain much barium. The ileum was reached by
the barium in an hour to an hour and a half and the cecum in four to
five hours. In one of the subjects the head of the barium column had
reached the hepatic flexure at the end of five hours.

Gastric motor phenomena, as observed fluoroscopically, in the
presence of ulcer in the stomach or duodenum showed in fourteen of
the sixteen patients studied, either constantly or intermittently, devia-

4 ARCHIVES OF INTERNAL MEDICINE

tions from those observed in the absence of such a lesion. The abnormal
motor phenomena were the same whether the ulcer was located in the
stomach or in the duodenum, except that in gastric ulcer peristaltic
waves did not course over the area of ulceration ; this has been described
in a previous communication.^ Other than this, five types of abnormal
gastric motor phenomena were observed, as follows: (1) exaggerated
type of normal peristalsis; (2) irregular peristalsis; (3) antiperistalsis ;
(4) pylorospasm; and (5) the presence of an incisura in the greater
curvature.

Exaggerated Type of Normal Peristalsis. — This type of peristalsis
was observed in seven patients with ulceration of the first portion of
the duodenum. It persisted throughout the emptying periods of the
stomachs of four subjects and during a two hour period of observation
of a fifth. In the sixth subject the exaggerated type of peristalsis was
changed to the irregular type, described below, with the onset of pain,
and in the seventh subject this change occurred without the onset of pain.

In the presence of exaggerated peristalsis the number of waves
observed at any one time in a given stomach was constant, but the
number of waves varied from one to three in the different individual
stomachs. By the time a wave had progressed from its point of origin
in the fundus to the beginning of the lower third of the stomach it
either completely or almost completely bisected the stomach. These
deep waves pushed a considerable amount of gastric contents before
them, either filling the pyloric region of the stomach abnormally full
and giving rise to the so-called "prepyloric bulge," or sending large
amounts through the sphincter into the duodenum. Pylorospasm was
discernible intermittently in five of these stomachs.

Irregular Peristalsis. — This type of peristalsis occurred in the
stomachs of seven of the patients studied. It was characterized by
marked variation in the time of appearance, duration and depth of the
peristaltic waves. Beginning at the usual site a wave would course
along the stomach distances varying from a few centimeters to as far as
the plyoric sphincter. An occasional wave, reaching as far as the
sphincter, would eject barium into the duodenum ; this showed the
absence of pylorospasm. More often, either the waves died out before
reaching the pylorus or no peristalsis was visible. Under these circum-
tances the gastric peristalsis was not of such a character as to permit it
to eject barium through the sphincter into the duodenum, and for this
reason it was not always ascertained whether or not pylorospasm was
present. However, certain of the peristaltic waves, which occasionally
reached the pyloric sphincter, did not eject barium into the duodenum ;
at which times pylorospasm was considered to be present.

6. McCliire. C. W.. and Reynolds, L. : J. A. M. A. 74:711 (March 13) 1920.

REYNOLDS-McCLURE— GASTRIC MOTOR PHENOMENA 5

The irregular type of peristalsis persisted throughout the emptying
period of the stomach in one patient and until after the cessation of pain
in another. It began with the onset of pain in five patients ; prior to the
onset of pain peristalsis was of the normal type in three of these patients
and of the exaggerated type in two. After the cessation of pain either
through natural means or after the administration of sodium bicar-
bonate, peristalsis became of the normal or nearly normal type and
the stomachs emptied rapidly.

Reversed Peristalsis. — In two patients the pyloric sphincter was
found to be in a state of spasm during observations made over periods
of about three and a half and six hours after the ingestion of the meal.
During these periods antiperistaltic waves took origin at the pyloric
sphincter and coursed back over the stomach a variable distance.

Pylorospasm. — Pylorospasm is defined as the failure of the pyloric
sphincter to open or to open its normal width, as judged fluoroscopi-
cally, in relation to the advance of a gastric peristaltic wave; i. e.,
the sphincter opened partially, as judged by the small amount of barium
seen to enter the first portion of the duodenum, or it remained closed,
as judged by the failure to see barium enter the duodenum. Prolonged
pylorospasm frequently accompanied the exaggerated type of gastric
peristalsis, while it was observed in but one stomach in which the
peristaltic waves were very shallow. The pylorospasm was usually
intermittent.

Incisiira. — In one patient a small penetrating ulcer occurred at about
the midpoint of the outline of the lesser curvature of the stomach.
Opposite the ulcer an incisura, invaginated from the greater curvature
side, almost bisected the stomach, thereby dividing the latter into a
lower and upper loculus. The patient's predominating symptoms were
nausea and vomiting, unaccompanied by pain. She was unable to ingest
more than a half of the usual meat meal because of the onset of nausea.
On fluoroscopic examination it was found that the portion of the
meal eaten completely filled the upper loculus of the stomach, and that
very little of the food had entered the lower loculus. In spite of the
complete filling of the upper loculus the patient did not experience a
sensation of fulness. Within fifteen minutes after the ingestion of the
meal eaten completely filled the upper loculus of the stomach, and that
incisura was less deep. Active peristalsis was present in the pyloric
region at this time and the sphincter was acting in a normal manner.
Twenty minutes later the upper loculus of the stomach was found to
be empty, and within a total period of three hours the stomach was
found empty.

In four patients a small incisura developed in the greater curvature
of the stomach coincidently with the onset of pain, which will be
discussed later.

6 ARCHIVES OF INTERNAL MEDICINE

Pain. — During the period of observation twelve patients with peptic
ulcer complained of epigastric pain. The character of the pain was
severe in five, moderately severe in two, mild in two and mere discom-
fort in three. The pain developed between one hour and two and a half
hours after the ingestion of the meal in all but one patient in whom
discomfort developed thirty minutes after the meal. The onset of
severe or moderately severe pain in four of the patients was preceded
by milder pain over periods varying from five to thirty minutes. The
onset of all types of pain was accompanied by distinct modifications
in whatever type of motor activities the stomach had previously mani-
fested, except in two patients who will be discussed later. If the
peristalsis was of the exaggerated type with pylorospasm, the onset of
pain was heralded by an increase in the depth of the waves or in the
degree or duration of pylorospasm or both, except in one case in which
the exaggerated peristalsis became the shallow type. If irregular
peristalsis preceded the onset of pain, when the latter developed, the
irregularity became more pronounced or peristalsis ceased altogether.
If peristalsis had been normal before pain developed, it then ceased
or became of the irregular type. In four cases with pain a small but
definite incisura developed in the greater curvature, near the upper
level of the barium shadow, with the onset of pain. In one of these
the incisura accompanied mild pain a half hour before the severer
pain occurred. In two of these the incisura was the only demonstrable
abnormality developing coincidently with the onset of pain.

After cessation of pain peristalsis became normal, or nearly so,
and the stomach rapidly emptied itself, except in two cases in which
exaggerated peristalsis with intermittent pylorospasm remained. After
the onset of severe pain in one patient and mild pain in two others, 3
gms. sodium bicarbonate or 2 gms. sodium bicarbonate with 1 gm.
calcium carbonate suspended in 20 c. c. water were administered.
Within from three to ten minutes after taking, the pain disappeared,
and simultaneously the abnormal motor phenomena in the pyloric
sphincter and stomach, including the incisura in one case, disappeared.
In a fourth subject with the onset of moderate pain an incisura devel-
oped in the greater curvature of the stomach, while peristalsis and the
behavior of the sphincter remained normal. The patient was given
20 c. c. tap water to drink, immediately after which peristaltic waves
became deeper while the pain remained the same. Twenty-five minutes
later 3 gms. sodium bicarbonate suspended in 20 c. c. tap water were
taken. Ten minutes later all pain had ceased but gastric peristalsis
remained unaffected and the incisura persisted. Just prior to the dis-
appearance of pain after the administration of alkalis to these four
patients it was noticed that the gas bubble in the fundus of the stomach
was much increased in size.

REyXOLDS-McCLURE— GASTRIC MOTOR PHEXOMEXA 7

Two patients, in whom ulcers were located in the first portion of
the duodenum, were observed in whose stomachs no modifications of
peristalsis occurred with the onset of pain. The type of peristalsis
present in the stomach of one of these patients was the exaggerated
type and in the other the waves were perhaps somewhat deeper than
normal. Both subjects developed mild epigastric pain lasting fifteen
and thirty minutes, respectively.

Four patients in whom active ulcers were located in the first portion
of the duodenum w-ere studied in whom no pani or discomfort devel-
oped during the period of observation after eating the meal. However,
the ulcers present in these patients were of the acutely painful type and
pain had been present up to the day of observations here reported. Two
of the patients were observed over periods of two and three hours
only, during which times the stomachs showed the exaggerated type of
peristalsis ; pylorospasm was present in one of these and but very little
barium entered the intestines. The stomachs of the other two were
observed throughout the period of emptying. One showed exaggerated
peristalsis without pylorospasm and emptied in three hours and twenty
minutes. The other showed a marked type of irregular peristalsis
without pylorospasm and emptied in five hours.

Pain developed in eight of the patients at times when the amounts
of food present in the stomachs varied from the quantity ingested to
half that amount. The development of pain was delayed until the
stomachs were nearly empty in four patients.

Emptying Time of the Stomach. — While the normal stomach was not
quite empty at the end of five hours, in the presence of peptic ulcer the
stomach was, with few exceptions, completely empty in three and one-
half to four hours. The emptying time of the stomach of one patient
was much delayed in the presence of prolonged pylorospasm. Delayed
emptying time was not observed in any other condition. In one stomach
in which the irregular type of peristalsis was present, there was an
initial delay in emptying. This was evidently the result of the infre-
quency of peristalsis in the presphincteric region, since, whenever a
peristaltic wave reached that region, barium was ejected into the
duodenum. Nevertheless, this stomach emptied itself of barium in four
hours. In another patient, with an ulcer in the first portion of the
duodenum, the stomach emptied itself in three Hours. Peristalsis was
of the one wave type and during the first half hour after the ingestion
of the meal peristaltic waves were regular in time but irregtilar in depth
and almost complete pylorospasm was present. Peristalsis then became
regular in the depth of the waves, spasm of the pylorus ceased and
during the succeeding two and one-half hours the stomach emptied itself.

8 ARCHIVES OF INTERNAL MEDICINE

SUMMARY AND DISCUSSION

Fluoroscopic observations on the normal human stomach, after the
ingestion of finely divided meat mixed with barium, show that it
empties itself in a regularly progressive manner. Peristaltic waves begin
high up in the gastric walls at uniform intervals of about twenty seconds
and gradually deepening progress in an orderly manner to the region of
the pyloric sphincter. As each wave approaches the sphincter the latter
opens, allowing chyme to be ejected into the duodenum over a period
of about ten seconds.' With the subject in the reclining position, one
of the normal stomachs emptied itself in five hours. Three of the
normal stomachs were almost empty in five hours ; under the conditions
of the observations here reported a very small residue remained along
the greater curvature of these stomachs for a longer period. The
stomach of the fifth subject contained a moderate sized residue at the
end of seven hours.

Abnormal phenomena observed in the stomachs of patients with
duodenal or gastric ulcer were modifications of the motor activities
of the stomachs of healthy persons. The abnormalities noted were:
(1) an exaggerated type of normal gastric peristalsis; (2) irregularity
in the time of occurrence, depth and length of the course of peristaltic
waves; (3) partial or complete intermittent spasm of the pyloric
sphincter; (4) localized, permanent, stationary spasm of the gastric
musculature causing the so-called incisura; (5) gastric antiperistalsis ;
(6) delayed emptying time of the stomach; and (7) very rapid empty-
ing of the stomach. The onset of pain was accompanied by modifications
in whatever type of motor activities the stomach had previously
manifested, with two exceptions. The various abnormal motor phenom-
ena were observed in the stomachs of peptic ulcer patients who did not
develop pain during the period of observation.

The abnormal motor phenomena described are familiar to all clinical
roentgenographers. But their relation to the pain of peptic ulcer as
observed fluoroscopically has not been previously systematically studied.
On the other hand, the relation of gastric motor phenomena to the pain
of peptic ulcer has been studied by the well-known balloon method by
Carlson,' Hardt,'* Hamburger," Homans,^" and others. Carlson and
Hardt state that the pain of ulcer is the result of contractions of the
musculature of the stomach, pylorus (pyloric sphincter) or first portion
of the duodenum. Their evidence that pain is accompanied by pyloro-

7. Carlson, A. J.: Am. J. Physiol. 45:80, 1917.

8. Hardt. L. L. J.: J. A. M. A. 70:837 (March 23) 1918.

9. Hamburger, W. W.; Tumpowsky, I., and Ginsburg, H. : J. .\. M. .\. 67:
990 (Sept. 30) 1916.

10. Homans, J.: Am. J. M. Sc. 157:74, 1919.

REYXOLDS-McCLURE— GASTRIC MOTOR PHENOMENA 9

spasm or contraction of the duodenum is entirely indirect and conse-
quently its existence in their experiments is problematical. Further-
more, Homans obsen^ed pain at a time when the balloon method
failed to show the presence of peristalsis in the stomach. These
contradictory findings are explained by our observations, since we
found peristalsis might be either active or absent during the presence
of pain.

While Carlson frequently did not observe modifications in gastric
peristalsis in the patients with peptic ulcer which he studied, we
observed them in fourteen of the sixteen patients which we studied.
The frequent failure of Carlson to observe modifications of gastric
motor phenomena during the presence of pain may be explained in
two ways: (1) the balloon method permits recording of gastric peri-
stalsis only, and by it small incisurae or pylorospasm are not ascertain-
able; and (2) as Luckhardt and Carlson ^^ note, peristalsis as observed
by the fluoroscope was not accurately recorded by the balloon method.

It has already been noted that we could not always determine the
presence or absence of pylorospasm during the time of occurrence of
pain. For this reason, and since the balloon method does not directly
demonstrate pylorospasm, whether or not spasm of the pylorus always
accompanies pain remains undetermined. However, we have made
one observation which suggests that pylorospasm may be absent during
the occurrence of pain. This observation was made on a subject with
an active ulcer in the first portion of the duodenum. On one occasion
the patient voluntarily complained of severe epigastric pain at a time
when fluoroscopic observation showed barium to be passing through
the pyloric sphincter in an apparently normal amount and manner. The
next peristaltic wave, however, showed the presence of complete pyloro-
spasm, although the pain had ceased.

The essence of our findings regarding the relation of peptic ulcer
pain and motor phenomena of the stomach and pyloric sphincter is
that during the presence of pain abnormalities of these motor phenomena
are usually demonstrable. But merely because of this the conclusion
reached by most observers, that motor phenomena are the cause of the
pain of peptic ulcer, is not necessarily correct. Evidently these observers
have as a basis for this conclusion the reasoning from the analogy that
spasm of voluntary muscle can cause pain. But, on the other hand,
spasm of voluntary muscle is also commonly a phenomenon protecting
against the development of pain. Therefore, the mere association of
the abnormal motor phenomena and pain, which we found in our
observations, does not in itself determine the causal relation of the

11. Luckhardt, A. B.; Phillips. H. T., and Carlson, A. J.: Am. J. Physiol.
50:60. 1919.

10 ARCHIVES OF INTERXAL MEDICINE

two. However, it seems plausible to assume that muscular movements
surrounding an ulcerated area in the stomach and duodenum could cause
pain, provided pain nerves were present. But such an assumption does
not explain why the pain does not usually persist throughout the entire
emptying time of the stomach, nor why pain occurs early after food
ingestion in some patients and late in others. The fact that pain did
not always accompany abnormal gastric motor phenomena or pyloro-
spasm could be explained on the theory that the degree of muscle spasm
was not sufficient to produce pain. This theory could also explain the
observation that pouring tenth normal hydrochloric acid onto a duodenal
ulcer through a duodenal tube produced, without any subjective sensa-
tions, pylorospasm, cessation of gastric peristalsis and duodenal anti-
peristalsis in one patient which we studied. But it is to be emphasized
that there is no means to prove such a theory. From this discussion it is
evident that the fact that the two phenomena, normal or abnormal gastric
or sphincteric motor phenomena and the pain of peptic ulcer, occur
simultaneously does not conclusively demonstrate that motor phenomena
cause the pain.

Hurst ^^ and others have found that distention of the stomach
or intestines by blowing them up with air to a sufficient pressure causes
pain. Because of this finding Hurst proposed the theory that gastric or
intestinal pain was the result of forceful distention of the gut wall due
to violent contraction on material in the lumen of the gut of the section
of the wall immediately above the distended region. Undoubtedly pain
can be caused by the method employed by Hurst, but there is no
experimental proof demonstrating that distention of the magnitude
produced by it occurs as the result of the presence of peptic ulcer.
Furtherm.ore, we have observed, as has Homans, complete cessation of
peristalsis in the stomach during the occurrence of pain; since under
these circumstances no peristalsis occurred, there could not have been
the distention produced as postulated in the theory proposed by Hurst.

This discussion shows that no motor phenomena are peculiar to the
occurrence of the pain of peptic ulcer, that there is at present no accurate
means available for measuring the degree of spasm of the gastric or
sphincteric musculature, and that there is almost no support for the
distention theory proposed by Hurst. In view of these facts it is
evident that there is no incontrovertible proof that the pain of peptic
ulcer is the result of motor disturbances in the stomach and pyloric
sphincter. The most .satisfactory evidence in support of the theory that
such motor disturbances are the cause of the pain of peptic ulcer is
that our observations show that gastric or sphincteric motor disturbances

12. Hurst, A. F.: Sensibility of the Alimentary Tract, London, 1911, p. 47.

REYNOLDS-McCLURE— GASTRIC MOTOR PHENOMENA 11

are almost invariably associated with the pain. But our observations
do not furnish conclusive proof of the truth of this theory, and for
this reason it must be admitted that the causal relation of motor
phenomena to the pain of peptic ulcer remains problematical.

From the cHnical standpoint the most important feature of the work
here presented is considered to be the fact that the usual disappearance
of abnormal motor phenomena occurring simultaneously with the cessa-
tion of pain gives an objective means of judging the effects of
therapeutic measures.

CLINICAL OBSERVATIONS ON THE CAPILLARY
CIRCULATION *

S. O. FREEDLANDER, M.D., and C. H. LENHART, M.D.

CLEVELAND

The direct observation of the capillary circulation in disease has
long been the aim of many workers, interested in a variety of clinical
and experimental problems. It is evident that the rest of the cardio-
vascular system exists only to regulate the blood flow through the
capillaries, for here takes place the exchange of gases necessary for
internal respiration and the exchange of materials necessary for metab-
olism. Any attempt to measure cardiovascular function is an indirect
attempt to measure the efficiency of the capillary circulation. For
example, blood pressure determinations are supposed to give some indi-
cation of the peripheral blood flow, but as will be seen later, blood
pressure is often a poor index of the state of the capillary circulation.
Much work has been done experimentally upon so called capillary
poisons, such as arsenic, etc., stimulating a clinical interest in many of
the acute intoxications, especially those accompanied by skin reactions,
such as occur after arsphenamin injections, diphtheria antitoxin, also
occasionally after the injection of foreign proteins. In acute infections,
such as influenza, the sudden collapse is often attributed to a capillary
intoxication. The disturbances of water balance in conditions like
acute nephritis are by some supposed to be dependent upon an altera-
tion of the permeability of capillary endothelium. In traumatic shock
there may be stasis of blood in the capillaries, as Cannon ^ and his
co-workers have shown. Disturbances of blood flow in arteriosclerosis,
in gangrene of the extremities, in various functional nervous diseases,
such as Raynaud's diseases, in fact, any condition altering the nutri-
tion of a tissue must have some vital relation to the capillary circula-
tion. It is thus evident that a direct view of the circulation in this most
important part of the vascular system might be of some clinical value.

Experimentally, investigation on the capillary circulation began soon
after the microscope came into use. Malpighi observed blood flow in
the mesentery and bladder of the frog in 1686; Leeuwenhoek observed
it in the tail of the fish and in the bat's wing ; Cowper studied it in the
mesentery of the rabbit. During the last century, investigation was
centered particularly upon the contractility of capillaries and its control.

*This work was made possible by the clinical facilities oflfered by the
Cleveland City Hospital.

1. Cannon, et al. : Nature of Wound Shock. Blood in Shock and Hemor-
rhage, J. A. M. A. 70:526 (Feb. 23) 611 (March 2) 1918.

FREEDLAXDER-LEXHART— CAPILLARY CIRCULATION 13

The names of Strickler, Golubew, Tarschanoff and the classical paper
of Roy and Brown ^ in 1879 mark this period. Steinach and Kahn ^ in
1903, Krogh * in 1919, Hooker = in 1920 have contributed to this sub-
ject. An excellent review of the experimental work on capillaries was
made recently by Hooker.*

Most of the animal observations have been made upon translucent
tissues. The difficulties of similar observations in man are obvious.
Ophthalmologists have carefully studied the condition of the vessels
in the cornea and retina by means of various special microscopes.
Augstein ' observed in pannus the development of new corneal vessels
by budding and branching, while Kraupa ^ observed the anastomosis of
papillary veins, and also described the development of granular blood
flow in arteriosclerosis and other conditions ; and Streiff ^ saw changes
in blood flow in vessels at the limbus due to nephritis and other con-
ditions.

In 1874, Hueter,^" by means of reflected light observed the vessels
on the inner border of the lower lip, calling his method cheilo-
angioskopie. While he described stasis due to mechanical pressure,
and experimented with the effects of applying various solutions, his
results were indefinite.

Recently, Weiss," standardized a method for the observation of the
skin capillaries at the ends of the fingers and toes. Many years before,
Spalteholz had shown that whole organs could be made translucent by
reflected light, if they were immersed in a transparent oil. Lombard, ^^
in 1912, working in Von Frey's laboratory applied this principle to the
observation of capillaries at the fingertips in man, and made this obser-

2. Roy and Brown : Blood Pressure and Its Variations in the Arterioles,
Capillaries and Smaller Veins, J. Physiol. 2:323, 1879.

3. Steinach and Kahn : Echte Kontractilitat u. motorische Innervation der
Haut-Kapillaren, Arch. f. d. ges Physiol. 97:105, 1903.

4. Krogh: Studies on Capillariomotor Mechanism, J. Physiol. 53:399, 1920;
52:457, 1919.

5. Hooker: Functional Activity of Capillaries and Venules, Am. J. Physiol.
54:30. 1920.

6. Hooker : Evidence of Functional Activity on the Part of Capillaries and
Venules, Physiol. Rev. 1:112, 1921.

7. Augstein : Gefass-studien an der Hornhaut u. Iris, Ztschr. f. Augenh.
8: 1902.

8. Kraupa : Die Anastomosen an Papillen u. Netzhautvenen, Arch. f.
Augenh. 78: 1915.

9. Streiff: Zur methodischen Untersuchund der Blutzirculation in der Nahe
des Hornhautrandes, Klin. Monatsbl. f. Augenh., p. 395, 1914.

10. Hueter: Chiclangioskopie, Zentralbl. f. d. med.Wissen., pp. 225,241,1879;
Deutsch. Ztschr. f. Chir. 4:105, 1874.

11. Weiss: Beobachtungen u. mikrophotographische Darstellung der Haut-
kapillaren der ledenden Menschen. Deutsch. Arch. f. klin. Med. 3:119, 1916.

12. Lombard : Blood Pressure in the Arterioles, Capillaries and Smaller
Veins of the Human Skin, Am. J. Physiol. 29:355, 1912.

14 ARCHIVES OF IXTERNAL MEDICI XE

vation the basis of a method for measuring capillary pressure.
Recently this method was modified by Hooker." Weiss ^* together with
Jijrgensen ^' and other German workers have made numerous observa-
tions on the capillaries in a variety of clinical conditions. Niekau ^^ by
means of a special instrument extended the observations to other skin
areas. Many records of the capillary picture were made in cardiovas-
cular conditions, nephritis, diabetes, skin diseases, etc., and on this basis
they have constructed tentatively rather definite capillary records corre-
sponding to a variety of diseases. However, as Miiller^'' states, it is
too soon as yet to say what clinical value the method has. Normal
standards for age, sex, climate, etc., have not been made. Many other
difficulties will be pointed out later. A clinical method to be of value
must be susceptible of wide use by a variety of observers under differ-
ent conditions and with fairly constant results. It is with the hope of
stimulating a trial of this method rather than of drawing any very
definite conclusions that our observations are recorded.

Anatomic Basis. — Spalteholz '" showed that the capillaries of skin
are end capillaries, that is, they do not anastamose but form single
terminal loops, each having a distinct arterial and a distinct venous
limb. Each papilla of the skin is supplied by a single capillary loop
which runs at right angles to the skin surface. However, as we
approach areas where the skin ends, such as the base of the finger and
toe nails, the long axis of the capillary tends to become more and more
parallel with the skin, so that in cases with a well developed undis-
turbed cuticle, the loop runs in practically the same plane as the skin
surface. Thus, looking down vertically upon the skin surface, at the
junction of the cuticle and nail, we see the capillary loop in practically
its entire length, while in other localities only the top of loop is seen.
Figures 1 and 2, diagrams from Spalteholz, will make this clear.

The Method. — The finger, one with a well developed cuticle, is
placed on the stage of an ordinary microscope, so that the junction
of cuticle and nail is under the objective. This region of the finger
is coated with any transparent oil (we have used liquid petrolatum).
An electric light is then focused so as to strike the part observed at
an angle of about 45 degrees. The light, a 50-100 watt, is enclosed
in a conical hood with a convex lens inserted at the apex for condensing

13. Hooker and Danzer : Capillary Blood Pressure in Man, Am. J. Pliysiol.
52:136, 1920.

14. Weiss and Miiller : Ueber Beobaclitung der Hautkapillaren u. Hire
klinisclie Bedcutung, Miinchen. med. Wchnschr. 64:609, 1917.

15. Jiirgensen : Microscopic Study of the Capillary Circulation, Dcutsch.
Arch. f. klin. Med. 132:140, 1920.

16. Niekau: Beobaclitungcn init dcm Hautkapillarmikroskop, Deutsch. Arch.
f. klin. Med. 132:301, 1920.

17. Spalteholz: Handatlas der Anatomic des Menschen 3.

FREEDLAXDER-LEX HART— CAPILLARY CIRCULA TIOX

IS

the light. The hood is mounted on a ring stand. We used an ordinary
Bausch and Lomb microscope with a 16 mm. objective and 5 X and
10 X oculars, giving a magnification of from fifty to one hundred
times. Others have used lower magnification. The patient's arm can
be put on pillows and braced by sandbags and to steady the observed
finger, one can hold it lightly or have it inserted in a mould made of
tin or a dental mould. Several minor precautions made the observa-
tion clearer, namely (1) the skin should be clean and dry, (2) it is
best to observe a finger where the cuticle has not been recently cut
or disturbed, (3) at times there are disturbing light reflexes from the

— Epidermis
--Capillaries

(.'\fter .Spalteholz).

skin which can be diminished by changing the angle of the light or
covering part of the convex lens in front of the light. One does not
obtain a clear view in every case, for where the epidermis is thick
and rough, and the cuticle is ragged, a good observation is impossible.
Obviously it can not be used upon colored patients. Futhermore the
patient must cooperate in holding his arm and fingers quiet, for much
pressure upon the finger can not be u.sed in restraint, because the
blood flow will be altered.

Normal Appearance. — Wtih a clear view, just proximal to the junc-
tion of the cuticle and the nail, one sees a row of ten to twenty hairpin
shaped loops, red upon a light orange background. Often more rows
of capillaries are seen, running parallel to the first but with the indi-

16 ARCHIVES OF IXTERXAL MEDICIXE

vidual loops becoming shorter, that is, the more perpendicular to the
skin, the more proximal the row. Observation should be focused
on the most distal row of capillaries. It will be seen that each loop
has a shorter, narrower limb (the arterial limb), and a thicker, longer
limb (the venous limb), and that they are joined by a short connecting
limb usually about the same thickness as the venous limb. Usually,
no flow is at first visible but upon closer observation, in most cases,
a rapid, steady, continuous stream will be seen going from the
arterial to the venous side. The view is sometimes improved by the
use of a green screen. Sometimes the individual blood cells are dis-
cernible, but more often not. In thin skinned individuals the sub-
papillary venules can be seen running at right angles to the loops, and
occasionally the venous limbs can be traced down to this point. The
subpapillary arterioles lie deeper, and are not visible except in infants.
Normally, no pulsation of the field or the individual capillaries is

Arterial -

Fig. 2.— (After Spalteliolz).

seen. The individual loops vary considerably in their contour. The
arterial and venous limbs are fairly straight but in normal individuals
they are often quite tortuous. Some may appear darker in color than
others, probably due to differerfces in depth. There is also some varia-
tion in size.

For a systematic observation, the following i)oints should be
recorded :

1. The number of capillaries in the first row, increased or decreased
over normal.

2. Color of background.

3. Size of capillaries, relative size and thickness of arterial limb,
venous limb, and connecting limb.

4. Contour of capillaries.

5. Degree of capillary filling.

FREEDLAXDER-LESHART—CAPILLARY CIRCULATIOX

17

6. Presence of blood flow, speed and character of flow.

7. Pulsation of individual capillary or of field.

The interpretation of any single case may be wrought with many
difficulties. No standards have been established for age, sex, race,
climate, etc. It is a method not susceptible of accurate quantitative
measurements, so that the personal equation of the observer enters
into the conclusions very greatly. Furthermore, the observations are
limited to skin capillaries, which are larger and because of their
heat and water regulating function, capable of greater change due to
external conditions than are capillaries elsewhere. We also know that

Fig. 3.— .'\pp:

capillary systems vary in form, depending upon function, thus diflfer-
ing in lung, intestine, pancreas, kidney, spleen, etc." For these rea-
sons, the skin capillary picture will never be a short cut diagnostic
procedure, but may be at its best only of some diagnostic aid when
taken with the rest of the clinical picture.

Variations Due to Age and Sex. — In observing over 200 cases, per-
sons ranging in age from 1 day to 70 years, very few variations
could be detected due to age alone. The skin vessels can be seen
very clearly in young infants due to their thin epidermis. About
half of the infants showed very small capillary loops with a diffuse

18. Nagel: Handbuch d. Physiologic 1:760.

18 ARCHIVES OF I XT ERA' AL MEDICI XE

network of larger vessels, such as one sees in a frog's web, while
in other cases there were long thin loops, rather widely separated,
with arterial, venous and connecting limbs of the same caliber. The
arterial and venous limbs were widely separated and very long and
were traceable to the underlying subpapillary arterioles and venules.
Occasionally, short buds were seen branching ofT of the limbs, sug-
gesting the beginning of other loops. Probably, the large loops are
not real capillaries, but only the subpapillary vessels. The blood flow
in infants is steady and rapid. Why some infants show capillary
loops and others do not is inexplicable to us. At the age of six
months most of the cases observed showed very little difference from
the adult condition. In older individuals there is a tendency for the
capillaries to become more tortuous, but tortuosity is frequently seen
in young people without any demonstrable cause. No differences
attributable to sex were nbserved.

Vasomotor Reactions. — It has long been a disputed point as to
whether variations occur in the caliber of capillaries independent of
changes in the arterioles and venules. An excellent review of the
experimental work on this subject has recently been made by Hooker "
and space will not permit us to repeat it here. Suffice it to say
that overwhelming proof has not yet been assembled in favor of the
independent contractility of capillaries. We have tried to observe
the effects of mechanical stimulation as is done in the "tache" test
of Marie, which was used clinically by Miiller,^" and Cotton, Slede
and Lewis.-" However, the field was so clouded by the stroke of the
needle that no unequivocal observations could be made.

* 19. Miiller: Deutsch. Ztsclir. f. Nervenh. 48:413, 1913.
20. Cotton, Slade, Lewis: Contractile Power of Capillaries, Heart 6:227, 1915.

FREEDLAXDER-LEXH ART— CAPILLARY CIRCULATION 19

Weiss claims to have seen the arterial end of a capillary contract
while observing a man with "vasomotor" spasm in the arm. He also
quotes C. E. Weiss as having seen a spasm of the smaller retinal
vessels (capillaries?) in a patient with a transient amaurosis.

Stewart,-^ by measuring the blood flow in the hand, showed that
reflex stimuli such as heat and cold, when applied to the other hand,
markedly altered the blood flow to the part, presumably by the reflex
action causing constriction and dilatation of the arterioles. We have
repeated these experiments by observing the capillaries of the finger in
one hand while immersing the other hand in either cold or warm water.
Cold caused a momentary increase in the velocity of the blood flow
which gave way soon to a marked slowing, amounting sometimes to
complete stasis. The capillaries appeared slightly wider and more
full of blood. The appearance of increase in width was probably
largely due to a slowing of the current, thus decreasing the axial
stream, rather than to an active dilatation. This stasis was not per-
manent, but there was an intermittent alteration of the speed of the
blood flow until gradually at the end of some minutes the flow became
almost normal again. With heat the flow became much more rapid but
the capillaries did not change in contour.

In explaining the reaction to cold one can conceive of the slow
capillary flow being caused in any one of three ways. The velocity of
the stream could be decreased by arterial constriction diminishing the
intake; a contraction of the venules increasing the resistance to the
outflow; and finally, a dilatation of the capillaries themselves. Arterial
constriction by itself, however, could not cause the capillaries to be as
full as they were in these cases ; venous constriction alone could retard
the flow, and in the absence of a decreased intake, as from arterial con-
striction, the venous pressure would not have to be very high in order
to cause almost complete stasis. Briscoe,^^ investigated the capillary
and venous pressures in patients with irritable hearts. Many of these
patients habitually had cold cyanotic hands and others were very
susceptible to cold, their hands becoming blue and cold upon short
exposure. In the first group of cases it was found that the venous pres-
sure was slightly raised, while the capillary pressure was markedly
elevated. In the second group the vasomotor reflex experiments were
tried and upon putting one hand in cold water while observing the
other hand there was a marked rise in capillary pressure and a slight
rise in venous pressure, and the venules on the back of the hand were
observed to decrease in size. These phenomena were supposed to be

21. Stewart: Studies on the Circulation in Man, Heart 3:76, 1912.

22. Briscoe: Observations on Capillary and Venous Pressure, Heart 7:35,
1918.

20 ARCHIVES OF IXTERNAL MEDICINE

due to simultaneous contraction of arterioles and venules. In our
experiments capillar}' dilatation by itself could hardly be the cause of
the slowed flow because the capillaries were only slightly dilated, but at
the same time they were definitely redder, that is, more full of blood.
Consequently, the most plausible explanation of reflex capillary stasis
is a simultaneous contraction of arterioles and venules. The applica-
tion of this to the cyanotic mucous membranes, ear tips, finger tips, etc.,
occurring upon exposure to cold is apparent, as it is also to the local
asphyxia in Raynaud's disease.

Occasionally, as Jiirgensen ^^ stated, in the course of an observation
some of the capillaries suddenly became obliterated and then a moment
later filled up again while the other capillaries in the field did not
change. Many explanations for this have been offered. Jiirgensen
thinks that there are precapillary arterial-venous anastomoses which
are under control of the central nervous system and this sudden blotting
out of the capillaries is due to the shuting off of the blood through
these vessels which have been suddenly opened and closed by reflex
action. We have no evidence which bears on this point. Variations
in the rate of blood flow without any change in the capillary contour
were observed in normal individuals, but. were more frequent in
patients with cardiovascular disease. These changes were probably
dependent on a variation in tone of the supplying arterioles, because
they appeared simultaneously in contiguous groups of capillaries, they
were more frequent in cases with hypertonus, and they were not
accompanied by any demonstrable change in the contour of the
capillaries.

Shock and Hemorrhage. — It will be necessary to review a few of
the recent experimental findings in shock in order to form a basis
for the interpretation of the capillary observations. For a long time
it has been assumed, and more recently it has been proven by Gasser
and Erlanger-^ and also by Lee,-' that there is a reduction in the
effective blood volume in traumatic shock. The question of the "lost"
blood or "exemia" has been vigorously investigated. Inasmuch as
repeated experiments have eliminated the arteries and veins, atten-
tion was directed to the capillary area as the blood reservoir.

During the war, Cannon ^ and his co-workers showed that in
patients with shock the red blood count of the capillary blood was much
higher than that of the venous blood. Dale and Laidlaw ^° by the injec-

23. Jiirgensen : Mikrokapillar Beobachtungen u. Puis der kleinsten Gefasse,
Ztschr. f. klin. Med. 86:410, 1918.

24. Gasser and Krianger: Plasma Volume and Alkaline Reserve in Shock,
Am. J. Physiol. 50:104, 1919.

25. Lee : Field Observations on Blood Volume in Shock and Hemorrhage,
Am. J. M. Sc. 158:570. 1919.

26. Dale and Laidlaw: Histamin Shock, J. Physiol. 52:355, 1918.

FREEDLAXDER-LEXHART— CAPILLARY CIRCULATIOX 21

tion of histamin produced a condition very similar to traumatic shock,
and by exhaustive experiments they seem to have proven that the
important factor in this condition is an endothehal intoxication with
capillary stasis and increased permeability, causing a marked exudation
of fluid into the tissues. Following this work, Bayliss -' and others
have attempted to prove that shock is chemical in its nature, that is, that
some product of traumatized or poorly metabolized tissue causes an
endothelial intoxication with the production of a shock-like condition.
Aub and \Vu,-^ in a few experiments, by measuring the blood gas in
arterial and venous bloods, tended to show that there is stasis in the
peripheral circulation in cases of experimental shock produced by
trauma. However, the result of all this work is difficult to evaluate,
for although it is evident that there is a reduction in the effective blood
volume, and capillary stasis, it has not been definitely proved that this
is the initiating and not a secondary phenomenon. Much controversy
has ranged over the condition of the vasomotor center in shock.
Erlanger, Gesell and Gasser -' in studies on secondary shock produced
by exposure and manipulation of intestines, partial occlusion of the
vena cava and partial occlusion of the thoracic aorta, showed that the
vasomotor center retains its tone during the development of shock. At
necropsy, there was intense capillar)- congestion, especially in the
intestinal area. These authors ^" believed that the slow flow induced
by the vasoconstriction caused a clumping of corpuscles in the capil-
laries and veneules, thus choking and dilating these vessels causing
transudation into the tissues and the diminution of the effective blood
volume. These phenomena could not be limited to the portal area
because shock could be produced in eviscerated animals. Seelig and
Joseph ^^ also proved that the vasomotor center was tonic until late in
shock. Gesell,^- in experiments on volume flow through the submaxil-
lary gland, showed that a relatively small reduction in blood volume, as
by a small hemorrhage, caused a relatively large reduction in volume
flow. This was due to vasoconstriction. A simultaneous constriction of
arterioles and venules could cause a marked capillary stasis. It became

27. Bayliss : Intravenous Injection in Wound Shock. 1918.

28. Aub and Wu : Studies in Experimental Traumatic Shock, \m. J. Physiol.
54:416, 1920.

29. Erlanger, Gesell and Gasser: Studies in Secondary Traumatic Shock,
Am. J. Physiol. 49:89, 1919.

30. Erlanger and Gasser: Treatment of Traumatic Shock, Ann. Svirg. 68:
389. 1919.

31. Seelig and Joseph: The Vasoconstrictor Center During the Develop-
ment of Shock, J. Lab. & Clin. M. 1:283, 1916.

32. Gesell: Factors Controlling Volume Flow of Blood, Am. J. Physiol.
47:468. 1919.

33. Gesell: Studies in Secondary Traumatic Shock. .\m. J. Physiol. 49:90,
1920.

22 ARCHIVES OF IXTERXAL MEDICIXE

evident during the war that hemorrhage plays a part in producing
traumatic shock, or in hastening it, and this may be due to the vaso-
constriction induced.

Clinically, traumatic shock is recognized by the group of symptoms,
consisting of low blood pressure — grayish pallor — mental stupor —
rapid, small pulse — rapid, shallow respiration and usually subnormal
temperature. During the war, Cowell ^^ called shock primary, if these
symptoms followed immediately on the trauma, arid secondary, if they
developed gradually, without primary shock, or if the primary shock
were recovered from to some extent and then followed by the more
permanent condition characterized by the symptoms.

REPORT OF CASES

These three cases would probably fall under the head of primary
traumatic shock :

Case 1. — E. L.. male, aged 23 years, admitted Jan. 8, 1921. About one
hour before admission the patient had been struck by a train.

Physical Examination. — Temperature, 38 C. ; pulse, 140, small volume; res-
piration, 30, shallow ; blood pressure, systolic, 80 ; diastolic, not obtainable.
Patient unconscious, grayish pallor. External wounds on scalp and body. Pupils :
contracted, reacted slightly to light, other reflexes showed nothing abnormal.
Capillary examination: (1) Increased in number; (2; capillaries full, and
dark red in color; (3) contour — connecting and venous limbs relatively wide,
as compared to arterial limb; (4) background, light orange; (5) blood flow-
markedly slowed and stream somewhat segmented; (6) no pulsation of capil-
laries or of field.

Case 2. — J. C, aged 20 years, admitted April 15. 1921. Crushed between
freight cars, about one-half hour before admission.

Physical Examination. — Temperature, 38 C. ; pulse, 140, small volume; res-
piration, 38 ; blood pressure, systolic, 85 ; diastolic, 50. Patient conscious, pale.
General rigidity and tenderness of abdomen, shifting dulness in flank, liver
dulness not obliterated.

Operation revealed a rather large amount of blood in the peritoneal cavity,
caused by the tearing of the mesentery for about six inches of the ileum.
Resection of about one foot of the ileum, with lateral anastomosis. Patient
died the next day.

Capillary examination: (1) Slight increase in number; (2) connecting and
venous limbs relatively full; (3) capillaries, dark red; (4) field, light orange;
(5) flow, slow; (6) no pulsation of capillaries or field.

These cases, clinically, are comparable to cases of primary traumatic
shock. There was a definitely slowed capillary circulation and the
capillaries appeared full. From the discussion in the previous section
upon vaso-motor reactions, it is evident that the capillary stagnation
could be due either to capillary dilatation or to a simultaneous con-
traction of arterioles and venules. Cardiac failure could hardly be a
factor because there were no other signs of cardiac decompensation,

34. Cowell: Initiation of Wound Shock, J. A. M. A. 70:607 (March 2) 1918.

FREEDLAXDER-LESHART— CAPILLARY CIRCULATION 23

such as distension of the jugular veins, enlargement of the liver, oedema,
etc. Unfortunately, capillary and venous red blood counts were not
made on these cases. The next case because of repeated observations
may throw some light on the causative mechanism.

C.\SE 3.— R. W., male, aged 43 years, admitted March 11, 1921. Patient had
shot himself in the abdomen about fifteen minutes previous to admission.

Physical Examination. — Temperature, i7 C. ; pulse, 100, small volume ; res-
piration, 26; blood pressure, systolic, 100; diastolic, 70. Patient conscious, pale.
Abdomen slightly rigid in epigastrium. Gunshot wound through epigastrium.

Capillary examination: (1) Number increased; (2) capillaries very red and
full; (3) contour, nothing remarkable; (4) flow, very slow, almost complete
stasis in some loops; (5) background, dark orange. Red blood counts of the
capillary and venous bloods at this time showed; (1) Capillary red blood
corpuscles, 6,200.000; (2) venous red blood corpuscles, 3,776,000. The exam-
ination was repeated forty-five minutes later, the patient having had morphin,
li grain, fifteen minutes before. Pulse. 80, larger volume than before ; res-
piration, 26, unchanged; blood pressure, systolic, 110, diastolic, 70, Capillary
examination: (1) Capillary number, same; (2) capillaries not so full or so
dark in color; (3) contour about the same; (4) field lighter; (5) flow, faster
than before. Blood counts at this time: capillary, 5,210,000; venous, 5,576,000.

Operation revealed a gunshot wound of the liver with moderate
amount of hemorrhage. Patient seemed in good condition upon leaving
operating room. This patient presented a picture of mild traumatic
shock. There was marked retardation of capillary flow. Yet on the
administration of morphin, and external heat, within forty-five minutes
his systollic blood pressure had risen from 100 to 110, his pulse had
become fuller and his capillary flow had become almost normal. The
slowed peripheral circulation could hardly be due to an endothelial
intoxication causing a capillary relaxation, because of the quick recov-
ery, but was much more likely due to a vasoconstriction which was
relaxed by morphin. However, we are aware of the fact, that we
were only examining a small portion of the skin capillaries, and local
effects due to external conditions, such as temperature have not been
excluded. We can only say that in observing a great many cases, and
without paying attention to external conditions, we have not seen such
marked capillary stagnation except when due to some clearly assignable
cause, such as seen in cardiac decompensation. These observations
present no evidence as to the cause of the vasoconstriction. The low
pressure in the presence of a sound heart and vasoconstriction could
only be explained by a diminution in the efifective blood volume due to
a sequestration of the blood in some part of the circulatory system and
a loss of plasma in transudation through the vessel walls.

The more permanent condition of secondary traumatic shock resem-
bles clinically the so-called surgical shock. Without definite cause,
symptoms appear with a more gradual onset and a tendency to progress
to a fatal ending. It is this type of shock which was studied so exten-
sively during the war. This corresponds also to experimental shock.

24 ARCHIVES OF IXTERXAL MEDICINE

Case 4.— L. R.. female, aged 30 years, admitted Nov. 18, 1920. About twelve
hours prior to admission, patient was seized with a sharp pain in the epi-
gastrium, shortly after which vomiting began, and continued at frequent inter-
vals. Colicky pains had continued. No bowel movement for twenty-four hours.

Physical Examination. — Temperature, 38 C. ; pulse, 84, slow. full. Patient
somewhat pale, tongue dr>-. Abdomen slightly distended, no rigidity, tender-
ness on the left, above the umbilicus. No peristaltic waves visible, increased
peristalsis audible, no evidence of free fluid in abdomen. Repeated enemas
with no results. Laparotomy: left rectus incision revealed almost all of the
small intestines strangulated through an opening in the transverse mesocolon.
Small intestine gangrenous except for about six inches near cecum. Ends of
small intestine were stitched to abdominal wall. Patient was returned to bed
in a condition of shock with a grayish pallor, cold perspiration, pulse 144, small
volume. Temperature, 38 C. ; respiration, 44, shallow ; blood pressure, systolic,
75; diastolic not obtainable.

Capillary examination: (1) Background, purplish hue; (2) increased num-
ber; (3) capillaries showed narrow arterial limbs with wide connecting limb:
venous limbs slightly widened ; (4) capillaries, bluish red ; (5) complete stasis.

In this rather typical case of surgical shock, there was complete
capillary stasis, the connceting limb appeared definitely widened in
relation to rather short and small arterial and venous limbs, and the
capillary blood appeared suboxygenated.

Thus, the evidence of capillary stasis was much more marked than
in the preceding cases. Of course, the point of most interest is whether
the capillary widening was primary or whether it was secondary to a
capillary stasis due to vasoconstriction, and which being prolonged
caused a deficient metabolism of the tissues with the {filing up of carbon
dioxid and other metabolites, and subsequent capillary dilatation. All
that we can maintain in this case of shock is that by direct observation
there was complete stasis in the skin capillaries and the connecting
limb of the capillaries was widened. The globular massing of cor-
puscles mentioned by Erlanger was not seen in any of these cases of
shock.

Collapse in Septicemia. — It is a rather frequent clinical observation
that some patients suffering from a general setpicemia go rather sud-
denly into a shocklike condition several hours before death. We have
had the opportunity of observing a number of these cases soon after
the onset of collapse. They presented the general picture of shock,
low blood pressure, grayish pallor, at times cyanotic, semiconscious,
perspiration, skin cold and clammy, low temperature, small fast pulse,
shallow respiration.

Case S. — Puerperal Septicemia, General Peritonitis After Abortion. — E. B.,
aged 30 years. On admission: Temperature, 41 C; pulse, 140; respiration, 32.
About twelve hours afterward the patient suddenly went into collapse with
the general picture as above except that cyanosis was rather marked. Tem-
perature, 37.4 C. ; pulse, 160; respiration, 40; blood pressure, systolic, 75;
diastolic ?.

Capillaries: (1) Increased in numlicr ; (2) background, light; (3) capil-
laries full of blood, dark red; (4) venous and connecting limbs relatively wider

FREEDLAXDER-LES HART— CAPILLARY CIRCULA TIOX

25

than arterial limb; (5) flow, very slow and intermittent with complete stasis
in some capillaries. Venous red blood cells, 4,800,000; capillarj^ red blood cells,
6.400,000. Capillary blood did not flow freely and was very dark in color.
Patient died about three hours after this observation.

Case 6. — Puerperal Septicemia, General Peritonitis Follozving Abortion. —
M. B., aged 22 years. General picture of collapse : Temperature, 37 C. ; pulse,
132; blood pressure, systolic 75; diastolic, 50.

Capillaries: (1) Background, light; (2) about normal number; (3) con-
tour, nothing remarkable; (4) filled, but in granular and segmented fashion,
the cells being clumped together with clear spaces between; (5) absolute stasis.
Venous red blood cells, 4,124.000; capillary red blood cells, 4,544,000.

These two cases of septicemia had capillary stasis, without showing
other evidence of cardiac decompensation. There are so many factors
to consider in these cases, that at this time we can only submit the
observations without oiYering any theoretic basis for them.

Fig. 5.— Shock.

Several cases of uncomplicated hemorrhage have been observed,
all of them being cases of incoinplete abortion of which the following
is typical :

Case 7.— S. S., aged TiZ years, admitted April 29, 1921. The patient was
about three months pregnant and several days previous to admission began
to have vaginal bleeding. On the day of admission she had a severe hemor-
rhage and passed some clot?.

Physical Examination. — Temperature, 38 C. ; pulse, 128, small volume ; res-
piration, 30, shallow. Blood pressure: systolic, 60; diastolic, 25. Patient was
very pale, conscious and restless. Bleeding from vagina. Examination showed
a partly retained placenta. Hemoglobin, 55 (Talquist) ; venous red blood cells,
1,840.000; capillary red blood cells. 2,108,000.

Capillary examination: (1) Decreased in number; (2) background very
light, almost white; (3) capillaries poorly filled, appear granular due to
clear spaces between blood cells. No dilatation; (4) capillary flow very slow
and intermittent with varying speed.

The capillary was characterized by the granular appearance of the stream
due to the anemia ; and the slow capillary circulation was probably due to
vaso-constriction resulting from the diminution in blood volume. The stasis,
however, was not extreme as was evidenced by the red blood counts.

26 ARCHIVES OF IXTERXAL MEDICIXE

As was mentioned above, Gesell showed that a comparative small
loss of blood volume as by hemorrhage could cause a large diminution
in volume flow, because of vasoconstriction. Robertson and Bock,^° in
observation on the blood volume in soldiers who had had a hemorrhage,
found that the diminution in blood volume was greater than the amount
of the hemorrhage warranted. The red blood corpuscle count and
hemogloblin determinations in capillary and venous blood showed that
this was due to capillary stasis.

Cardiac Disturbances. — Decompensation : Weiss " and Jiirgensen ^°
made direct observations of the capillary circulation in many varieties
of heart disease, and in all stages of decompensation. We have been
able to confirm most of their findings. In well marked decompensation,
the capillaries are increased in number, the arterial limb is small and
narrow while the connecting limb and especially the venous limb are

ig. 6.— Cardiac

wide and full. The flow is very slow. Sometimes there is complete
stasis, and the blood flow is not steady, but the corpuscles appear rolled
together, with clear spaces between. If the spaces are frequent, thus
making the groups of cells small, the stream is called granular, and if
the spaces are farther apart the stream is called segmented. Occa-
sionally in complete stasis there is an intermittent retrograde move-
ment from the venous toward the arterial limb. The background iS
very dark, probably due to stasis in the underlying venules. In mild
cases of decompensation all that one may see are slightly widened
venous and capillary limbs with slight slowing of the current and a
granular or segmented stream. The German investigators claim that

35. Robertson and Bock :
29:136, 1919.

Blood Volume in Wounded Soldiers, J. Exper.

FREEDLAXDER-LEXHART-CAPILLARY CIRCULATIOX 27

signs of stasis may be seen in the capillaries before other clinical signs
are present, for example, after acute infections which are liable to pro-
duce cardiac complications, thej' claim that the capillary picture of
decompensation may precede other symptoms. We have not observed
any such cases. However, we have observed several cases of recovery
from a cardiac breakdown, which clinically seemed to be compensated,
but in which capillary- circulation was slowed, the stream was granular
or segmented, and the venous limb was full and wide. It appeared to us
that the complete picture of stasis in mitral disease appeared earlier, was
more marked, and persisted longer than in aortic disease. One of the
most marked cases observed was a man with tuberculous pericarditis
with a large effusion. A direct view of the capillaries has seemed
useful to us in a few cardionephritics. Especially when they present
the well developed picture of edema, shortness of breath, etc., it is often
difficult to decide readily whether the kidney or the heart is more at
fault. For example, one case was observed, that of a man about 50
years of age, who had general anasarca, ascites and hydrothorax and
was markedly short of breath. The capillaries showed only a slight
slowing and no marked fulness of the venous limb. It was apparent
that this man's clinical picture could not be entirely due to cardiac fail-
ure. Weiss ^^ thought that he could measure cardiac function by putting
a blood-pressure cuff on the arm and then observing the capillaries. His
method was to raise the pressure in the cuff until the flow stopped,
then to lower the pressure, and observe the pomt at which the stream
started. He claimed that ifi normal individuals the stream started at a
point from 5 to 10 m. below the stytolic pressure, while in cases with
diminished cardiac "power," the flow would not start until the pressure
was much lower. As Jiirgensen points out, this method as a test of
cardiac function has very little basis theoretically, and we have found
it of no practical value.

Capillary Pulse: It is well recognized that a capillary pulse is a'
constant finding in aortic insufficiency, and that it is frequently
observed in aortitis, arteriosclerosis, exophthalmic goiter, febrile dis-
turbances, etc. However, clinically, it has not been studied inten-
sively. Herz '■ by means of onygraphic pressure records together with
radial pulse records concluded that what was called a capillary pulse
was really a pulsation of the smaller arterioles and could be changed or
made to disappear in many cases by influences such as changes of tem-
perature. However, Glassner,'' using the same methods, concluded

36. Weiss : Eine neue Methode zur Sufficienz-priifung des Kreislaufs,
Ztschr. f. exper. Path. u. Therap. 19: 1918.

37. Herz: Wien. Klin. 6:165, 1896.

38. Glassner: Klinische Untersuchungen iiber Kapillarplus, Deutsch. Arch,
f. klin. Med. 97:83, 1909.

28 ARCHIVES OF IXTERXAL MEDICINE

that a capillary pulse was either central or peripheral in origin, central
in aortic insufficiency, but peripheral, due to vascular changes, in other
cases, such as arterio-sclerosis, nephritis, exophthalmic goiter, etc. As
Jiirgensen points out, in observing the capillaries in aortic insufficiency,
the whole field seems to move in two planes ; one at right angles to the
microscope and the other in the same plane as the microscope; that is,
the field moves horizontally and vertically. In the few cases which we
observed, it also appeared as if the flow in the capillaries was actually
pulsatile, although, due to the vertical movement, it is difficult to'
observe the blood flow continuously. Observations were made on cases
of exophthalmic goiter, arteriosclerosis and neurasthenia, in which a
capillary pulse was present, but the pulsation was only in the horizontal
plane, and the blood stream in the capillaries was not pulsatile. So

_

..«—

::^

^ ■

A

1 f %
! (1

0.'

ft f

0

r

Y^

tX

Fig.

-Anemia from hemorrhage.

this was definitely an arteriole pulsation and could be called a pseudo-
capillary pulse. Jiirgensen saw capillary pulsations in cases of syphilitic
aortitis, but unfortunately all the patients with this disease, available
to us, were colored.

Chronic Interstitial Nephritis and Hypertension. — Cases of hyper-
tension, either with or without other signs of nephritis, are often
puzzling because of the varying picture which they may present. Given
two patients with the same blood pressure, one may be perfectly com-
fortable and the other may present all sorts of symptoms, such as
aphasia, anaurosis, headaches, dizziness, etc. Various focal symptoms
have been explained upon the basis of small hemorrhages, or arterial
spasm. Again, in some cases, the blood pressure may be reduced with-
out producing any symptoms, while in other cases a slight reduction in
the pressure is attended by marked disturbances.

FREEDLASDER-LESHART— CAPILLARY CIRCULATION 29

The capillary picture may possibly throw some light on these phe-
nomena :

Case 8.— A. F., aged 41 years, female, admitted Feb. 14. 1921. For several
months the patient had dizzy spells and headache, and black spots before the
eyes. More recently she had frequent vomiting spells. For several days she
had slight continuous vaginal bleeding. The day before admission she had
a convulsion.

Physical Examination. — There was evidence of cardiac hypertrophy, but no
edema. Pulse was 100 and small volume. The eye-grounds showed a marked
albuminuric retinitis. Blood pressure: systolic, 240; diastolic, 160. Urine:
albumin -|- 4- -}-, many casts.

Capillary examination: (1) Slight increase in number; (2) background,
light; (3) capillaries poorly filled, appear granular. All limbs thin and of
about the same width. Rather marked tortuosity; (4) flow, intermittent, stream
being very swift for a moment, then it would suddenly stop completely.

When observed the next day the systolic blood pressure was 202, and the
capillary picture was unchanged. An amylnitrite pearl was given and the
systolic pressure dropped to 170, the capillary flow became steady, and there
was a marked pseudo-capillary pulsation. However, in the course of five
minutes the pressure was back to its previous level and the capillary flow
resumed its original character.

It was seen that the capillaries were poorly filled and the flow was
intermittent. This must have been due to arterial hypertonus, because
first the capillaries were poorly filled, as one would expect with arteriole
contraction, and second, the flow became better under amyl nitrite. In
confirmation of this point, Kraus found a lowering of the capillary
pressure along with a rise of the arterial pressure in many cases of
hypertonus. Evidently, many of the patient's symptoms were due to
poor capillary circulation, and in such a case reduction of the blood
pressure would seem to be rational therapy. We have observed other
cases of hypertension in which the flow was steady and continuous, and
these patients had few symptoms.

In chronic interstitial nephritis the capillaries usually have long thin
arterial and venous limbs with a slightly wider connecting limb. The
flow depends on the degree of arterial tonus.

Arteriosclerosis.— In marked cases of arteriosclerosis the arterial
and venous limbs are elongated, thin and tortuous and show a very char-
acteristic formation of small loops ofif of the main limbs. These little
loops look like buds and, by some, are supposed to represent an attempt
at the formation of anastomosis or of new capillaries. Those loops do
not appear in all cases of arteriosclerosis, and in our observations bear
little relation to the amount of arteriosclerosis in the palpable arteries.
The most marked cases of "budding" which we have seen, were in two
cases of club fingers, with little demonstrable arteriosclerosis. However,
other cases of club fingers did not show the "buds." Their significance
is not clear to us.

30 ARCHIVES OF IXTERXAL MEDICIXE

Capillaritis. — The terms endothelial intoxication and capillary
poison, have 'been used lately in relation to a variety of phenomena.
Weiss ^^ gives a rather definite picture for this capillary disturbance
which he calls capillaritis. The capillary has short, narrow, arterial
and venous limbs and a widely distended connecting limb. He described
this in cases of acute nephritis, especially in children, in scarlet fever,
occasionally before eruption. We have observed no cases of scarlet
fever and only one case of acute nephritis, this in an adult whose capil-
laries showed no marked change. However, it was thought one should
be able to obtain some evidence of capillary intoxication in cases of
mercury and arsenic poisoning. In one case of fatal mercuric chlorid
poisoning, with complete anuria, no capillary change was observed.
Several cases were observed, showing salivation and gingivitis due to
mercurial therapy, and with one exception there was nothing of note
in the capillaries. This patient, age 20 years, had received a great deal

imim

■x'k

•^' jt^ "W/

.^

Fig. 8.— Ar

of mercury by injection and inunction. He was markedly salivated
and his gums were very swollen and sore. He also complained of
anorexia and bachache. The capillaries appeared increased in number,
very red, with short arterial and venous limbs, and relatively broad
connecting limb. Blood flow was steady and moderately fast. Pre-
sumably, this is what Weiss describes as a capillaritis. However, the
same picture was observed in three other cases, one, a patient with
Banti's disease, and the others apparently normal individuals. Two
cases of arsphenamin intoxication were observed, one an acute reaction
with an eruption, fever, and edema ; the other, a chronic case of jaun-
dice, nausea and vomiting, and neither showed any capillary change.

39. Weiss: Beobachtung iiber Veranderung der Haut-kapillaren liei Exan-
them, Miinchen. med. Wchnschr. 65: 1918. Das Verhalten der Haut-kapillaren
be! akuter Nephritis, Miinchen. med. Wchnschr. 63:925, 1916.

FREEDLAXDER-LESHART-CAPILLARY CIRCULATION 31

Other Diseases. — Unfortunately, we have not had the opportunity
of observing many conditions which might show something of interest ;
vasomotor disturbances, such as Raynaud's disease, erythromelalgia,
disturbances in various kinds of paralysis, etc. Niekau,^' by means of
a special microscope designed by Miiller observed the skin in locations
other than the finger tips. He studied various skin diseases, such as
psoriasis, also some of the acute exanthemata.

Some work has been done on capillary flow in diseases of infancy,
such as exudative diathesis. Observations would be of interest in the
acute intoxications in which Merriot " has shown by blood flow mea-
surements that the peripheral circulation is slowed. Thus, there are
many clinical branches in which a direct examination of the capillaries
might be of interest.

CONCLUSION

While the observations in this series of cases might lead one to
draw certain general conclusions as to circulatory condition in shock, in
cardiac conditions, etc., one must be very careful in his deductions.
The method is not a quantitative one, and only a few of the capillaries
of a most labile part of the circulatory system are being observed.
Consequently the reports of a large number of cases by diiiferent
observers must be awaited before we can evaluate the method. How-
ever, because of the corroboration of the clinical picture and of certain
experimental work, one has some faith in observations of a capillary
stasis in shock, the capillary picture in cardiac decompensation, the
poor capillary flow in some cases of hypertension, and the different pic-
tures of a capillary pulse.

It is hoped that other workers will be interested in making similar
observations.

SUMMARY

1. Observations on the capillary circulation in many clinical con-
ditions were made by Lombard's method.

2. The effect of reflex vasomotor stimuli on the capillary circulation
were observed, corroborating the blood flow experiments of Stewart.

3. In three cases of primary traumatic shock and one case of sur-
gical shock, stasis was observed in the skin capillaries.

4. Capillary stasis was observed in several cases of septicemia,
which had a sudden collapse.

5. Capillary stasis occurs in cardiac decompensation and may be of
some value in differentiating the preponderant factor in cardiorenal
diseases.

40. Merriot : Some Phases of the Pathology of Nutrition in Infancv, .Am.
J. Dis. Child. 20:461 (Nov.) 1920.

32 ARCHIVES OF IXTERXAL MEDICIXE

6. In some cases of hypertension, there is a granular, intermittent
capillary flow.

7. A true capillary pulse can be differentiated from a pseudocapil-
lary pulse.

8. Observations may be of value in diseases where an endothelial
intoxication or capillaritis is suspected, but no conclusive observations on
such cases were made in this series.*^

Our thanks are due to Dr. L. J. Karnosh for the drawings. The sketches
are serai-diagrammatic camera lucida drawings with a magnification of approxi-
mately X 70. In some cases they represent the composite picture of several cases.

41. The attention of the authors has just been directed to the fact that a
sort of polemic has been going on in the German and Austrian journals between
Weiss and Schur " regarding priority in the application of Lombard's method
to the clinical observation of capillaries. We know nothing of the merits of
the controversy.

42, Schur: Ueber KapiIlar-beot)achtung, Wien. klin. Wchnschr. 33:928. 1920.

THE PROTEIN REQUIREMENT IN TUBERCULOSIS*

WILLIAM S. McCANN, M.D.
(With the technical assistance of Estelle Magill.)

NEW YORK

In a previous coniiminication/ which was concerned chiefly with
the total energy transformations in tuberculosis, a few experiments
were given dealing with the nitrogen minimum in this disease. It was
thought to be desirable to extend these observations so as to include
data, not only regarding the minimal level of protein metabolism, but
as far as possible to throw light on the optimal quantity with which
to supply patients sufiFering from pulmonary tuberculosis.

Because of its intimate bearing on the problem of finding the
optimal protein requirement for such patients, the knowledge acquired
concerning the basal metabolism was supplemented by experiments
on the effect of protein food upon the heat production. Further
experiments were made - in which the action of all the foodstuffs
was studied, not only on the metabolism, but on the pulmonary ventila-
tion as well.

Studies of the protein metabolism in this disease, made up to
1903, have been compiled by Ott.^ Up to that tiine no attempts
seem to have been made to find the protein minimum in tuberculosis.

From 1888 to the present time there has been a great develop-
ment of our knowledge of the size of the normal quota of protein
for wear and tear. The general method for determining this has
been to provide the experimental subject with a diet poor in protein
(from 1 to 3 gm. nitrogen), but with an adequate supply of non-
nitrogenous foods. Under these conditions, in normal men, the pro-
tein metabolism is reduced to the minimum which is necessary to
cover the daily wear and tear of the protein containing organs and
tissues. For ready reference the results of the more important of
these experiments have been summarized in Table 8. More recently
investigations of a similar character have been made to determine to

* From the Russell Sage Institute of Pathology, in affiliation willi thu
Second Medical (Cornell) Division of Bellevue Hospital, New York, and from
the Pathological Department, Bellevue Hospital.

1. McCann, W. S., and Barr, D. P.: The Metabolism in Tuberculosis, .Arch.
Int. Med. 26:663 (Nov.) 1920.

2. McCann, W. S. : The Effect of the Ingestion of Foodstuffs on the
Respiratory Exchange in Pulmonary Tuberculosis, ."Xrch. Int. Med. To be
published.

3. Oft, A. : Die chemische Pathologic der Tubcrculosc. Berlin, 1903.

34 ARCHIVES OF IXTERXAL MEDICI XE

what extent this wear and tear quota is affected in various diseases.
A number of these have been summarized in Table 9.

Considerable discussion has arisen in the literature as to the exist-
ence or nonexistence of a "toxic destruction" of protein. For instance,
if the size of the wear and tear quota is found to be increased some
authors maintain that the increase is due to the action of toxins in
disease. Others hold that the protein metabolism is increased as part
of a generally increased metabolism, and that in spite of this nitrogen
equilibrium may be established if an adequate diet is given. Reference
to the experiments of Kocher * will show that the wear and tear
quota in typhoid is much increased. However, the experiments of
Schaffer and Coleman ^ show beautifully that if adequate amounts
of carbohydrate, especially, and fat are given in typhoid fever the
body may be protected from protein loss. This question has been well
reviewed by Rolland,'"' who finds that there is no toxic destruction of
protein in the sense that the body can not be protected from pro-
tein loss if an adequate diet be given, although she admits the pos-
sibility that it may occur to a slight degree in cases with temperature
about 40 C. Rolland studied four cases of tuberculosis, one of whom
she was able to keep in nitrogen balance with a diet furnishing 47
calories per kilogram and 10.7 gm. nitrogen. In another case 1.6
gm. nitrogen were added to the body daily with 49 calories per
kilogram and 9.6 gm. nitrogen in the diet. Two of her patients were
in negative nitrogen balance, but in one of these an adequate number
of calories was furnished. Kocher* also st;idied some patients with
tuberculosis in two and three day experiments, of which the data
are too meager to justify any conclusions.

A great deal of attention has been directed towards elevation of
body temperature as a possible factor in the increased protein metabo-
lism of acute infections. May " has shown that the carbohydrate
metabolism is greatly increased in fever. A deficiency of glycogen
rapidly occurs, and the body proteins are called upon to take a large
share in the increased energy production. That this state of febrile
inanition plays a great part in the increased protein destruction has
been shown by the results of Schaffer and Coleman,^ to which reference
has already been made. Bearing on the influence of body tenipera-

4. Kocher, R. A.: Ueber die Grosse des Kiweisszerfalls bei Fieber und bei
Arbeitsleitstiing. Deutsch. Arch. f. klin. Med. 115:82. 1914.

5. Schaffer, P., and Coleman, W. : Protein Metabolism in Tvphoid Fever,
Arch. Int. Med. 4:538 (Nov.) 1909.

6. Holland, A.. : Zur Frage des toxogenen Eivveisszerfalls im Fieber des
Menchen, Deutsch. Arch. f. klin. Med. 107:440, 1912.

7. May. R. : cit. v. Noorden's Handb. d. Path. d. Stoffwechsels. Berlin,
p. 591, 1906.

McCAXX—PROTEIX REOUIREMEXT IX TUBERCULOSIS 35

ture upon the protein metabolism are the vahiable experiments of
Graham and Poulton,*^ who elevated body temperature artificially by
means of baths. These authors found that a rise of body tempera-
ture to 104.3 F., (40.2 C.) does not by itself cause any breakdown
of the protein of the body. If a diet of very high caloric value
is taken, containing a large excess of carbohydrate but a minimal
quantity of protein, nonnitrogenous substances supply the whole of
the increased energ}- production set up by the high temperature.

For the best recent summaries of our present knowledge of the
protein metabolism reference may be made to those of Lusk," Cath-
cart " and Van Slyke.'^

TECIINIC

The technic of the management of the metabolism ward, in which
the patients were kept, has been described in a paper by Gephart
and DuBois.'^ In regard to the manner of weighing food, recording it,
calculating its composition and caloric value, and the collection of
the excreta the plan of Gephart and DuBois has been strictly followed.

The preparation of diets which could be completely taken and
retained by sick patients, especially the tuberculous, required the
greatest skill and painstaking care, for which we are greatly indebted
to the chief nurse, Miss Estelle Magill, and her three assistants,
Miss Elsa Forter, Miss Doris Cutler, and Miss Florine Nelson.

The data regarding diets given in Tables 1 to 4 represent only
food which was actually ingested by the patients. The chief source
of error which may exist is in the loss of food in the vomitus. A
note has been made in each case in which this occurred.

In determinations of the respiratory metabolism made on our
patients the apparatus used was a Tissot spirometer, in which expired
air was collected and measured, and from which samples were taken
for analysis. The analyses were done with a modified Henderson-
Haldane apparatus. The exact technic of these determinations is
given in a previous communication.- The first four subjects, in
Table 7, were studied in the respiration calorimeter.' The basal

8. Graham and Poulton : Influence of Temperature Upon Protein Metabolism,
Quart. J. M. 6:82, 1912.

9. Lusk, G. : The Science of Nutrition, Kd. 3, W. B. Saunders Company,
Philadelphia, 1917. Especially Chapter XII, pp. 334-361.

10. Cathcart, E. P.: The Physiology of Protein Metabolism. Monographs
on Biochemistry. Longmans, Green and Co., London, 1921.

11. Van Slyke, Donald 1).: The Chemistry of the Proteins and their
Relation to Disease, Oxford Medicine, Vol. I. O.xford Press, 1920.

12. Gephart, F. C, and DuBois, E. F. : The Organization of a Small
Metabolism ward. Clinical Calorimetry, Third Paper, Arch. Int. Med. 15:829
(July) 1915.

36 ARCHIVES OF INTERNAL MEDICINE

requirement of energy for each patient was calculated from the data
of the metabolism observations, adding 10 per cent, to the observed
heat production to cover the effect of the specific dynamic action
of foodstuflfs.

In separation of the feces carmine was given in perforated cap-
sules. The separations were all made satisfactorily for periods vary-
ing from three to eight days, generally five days. The stools were
preserved with formalin in a refrigerator until the period was com-
plete. They were then mixed and sent to the laboratory where they
were rubbed to homogeneity with alcohol containing 1 per cent, sul-
phuric acid. After drying on a water bath the dry weight was
taken, and the stools ground to a fine powder and preserved in
a tightly stoppered bottle. Nitrogen in the urine and stools was
determined by the Kjeldahl method. These analyses were made by
Miss Frances D. Rule.

CASE HISTORIES

Case 1. — Early Apical Tuberculosis ■with Hemorrhage. — Hubert \\'., a negro
aged 28 years, was admitted Nov. 3, 1920, on account of blood spitting. He had
"felt fine" in the morning of day of admission and had gone to work as
usual. A few minutes after starting work he had a severe fit of coughing
lasting ten minutes, during which he spat up one half cupful of frothy blood.
Coughing thereafter at frequent intervals produced more hemoptysis. With the
cough there was a dull sense of soreness over sternum and precordiuni.

Up to this time his health had been excellent. While in Central America in
1911 he had malaria. Cardiorespiratory history was negative, except for the
fact that he was subject to frequent colds, with headaches and neuralgic pains
about the eyes. For this he habitually took "headache powders." Weight was
constant at its maximum of 135 pounds up to admission, thougli one week after
admission it was found to be 120 pounds.

On admission, examination of the patient was limited to the front of the
chest, in which the findings were only those of dulness, bronchovesicular breath-
ing over the left upper lobe. His fiemoglobin was 85 per cent. ; red blood
corpuscles, 5,152,000; leukocytes 6,400 with a normal dififerential count. The
urine was normal, except for an occasional hyalin cast. Blood Wassermann
reaction was negative. Temperature range, from 98 to 100 F. ; pulse, from
60 to 90; respirations, from 20 to 30. Hemoptysis continued until November 10.
Several sputum examinations were made, but no tubercle bacilli could be
demonstrated.

November 1(1, jihysical examination was negative except for the chest.
Patient was well developed and well nourished. The chest was large and well
formed. There was distinct limitation of motion on the left side. There was
normal resonance, except over the left upper lobe from the third rib in front to
the mid-scapula behind. Over the dull area breathing ranged from broncho-
vesicular to bronchial in quality. Fine rales were heard over a small area
below the outer third of the clavicle. The heart was normal in size and
position. There were no murmurs. The rate and rhythm were regular. Blood
pressure, systolic, 122 mm., diastolic 100 mm.

Roentgenograms Nos. 80.942 and 80,943 showed evidence of thickening of the
pleura of the left lung, with infiltration of the extreme left apex, with hyper-
vascularization of the right upper lobe. Heart shadow normal. Tuberculosis.

McC ANN— PROTEIN REQUIREMENT IN TUBERCULOSIS

37

Patient was observed in the metabolism ward from November 10 to
November 24. Dietary measurements and excreta analyses are tabulated in
Table 1 (See also Fig. 1). '

Nov. 27, 1920: Basal metabolism determination. Tissot spirometer. Age,
28; height, 166.5 cm.; weight, 55.1 kg.; surface area, 1.60 sq. m. ; temperature,
982 F. ; pulse, 80; respirations, 20. Expired volume, liters per hour, S. T. P. D.
= 57.95. Gas analysis— O, per cent. 18.54, 18.58. COj per cent. 2.30, 2.33. R. Q.
0.94, calories per hour = 67.9. Metabolism. 107.5 per cent, average normal.
Twenty-four hour heat production estimated 68 X 24 = 1,632 calories, basal
energj' requirement, 1.632+163 = 1,795 calories.

TABLE 1

-c.

SE OF

Hubert W.

(C.XSE 1)

1

1

2

2

M

e

1

1

o

a
o

1

S
o

1

a
o

a
o

i

1

II

1

s

1^

a
1
1

-

u

a

E-

z

u

o

10

1.977

62.0

89.1

218.4

9.9

16.92

2.39

19.31

—9.4

55.51

182

119

1,836

11

1,950

63.5

90.6

206.5

8.40

10.79

—0.6

155

12

1,751

62.2

90.8

207.7

10.0

10.12

2.39

12.51

-2.5

55.19

(—44)

48

1,520

13

2,538

36.1

121.5

307.2

5.S

6.20

2.30*

8.50

—2.7

743

3t

1,473

2,.^7

37.1

120.9

300.2

5.9

6.20

2.30

8.50

—2.6

712

12t

1,495

-5.4

55.16

(-602-)

1«

2.390

27.7

112.4

300.2

4.4

6.20

2.30

8.50

—4.1

8t

1,358

17

2,390

27.7

112.4

300.2

4.4

6.20

2.30

8.50

—4.1

55.29

t

1,400

IR

2,500

34.3

109.3

327.6

5.5

4.77

2.30-'

7.07

-1.6

705

19

2.601

38.6

110.0

351.3

6.2

6.20

2.30

7.50

806

X.-'W!

35.6

326.3

5.7

4.50

56.41

713

21

2,508

35.6

109.5

326.3

5.7

5.07

2.30

7.37

-1.7

n

2,390

27.7

112.4

300.2

4,4

4.U

2.30

6.41

—2.0

595

298.7

3.78

2.,30

6.08

617

24

.,m

27.7

120.9

300.2

4.4

3.96

2.30

6.26

-1.9

55.11

..4

Comment. — Examination of the data in Table 1 shows that the
urinary nitrogen excretion did not fall as promptly in response to
the reduction in food nitrogen as it did in the subsequent cases.
The cause of the lag is not at once apparent. There was practically
no fever (from 98 to 100 F.) during the two weeks of observation.
There was no evidence of toxicity. The patient felt quite well. It
it not unlikely that the source of the extra nitrogen was clotted blood
retained in the lung and undergoing slow absorption.

The nitrogen loss in the feces was quite large throughout. There
was no diarrhea. The diet was not of a bulky nature such as would
ordinarily lead to poor utilization.

At the end of the period of observation the area of dulness over
the left upper lobe was .somewhat diminished below the clavicle.
Breath sounds had a slightly higher pitch than normal and a some-
what prolonged expiration. No rales were heard.

38

ARCHJVES OF IXTERXAL MEDICINE

Case 2. — Acute Pneumonic Phthisis. — Abraham M., an Arab, aged 21 years,
was admitted Nov. 29, 1920, complaining of cough and pain in the chest. His
ilhiess began two months before admission with a severe "cold," cough persisted
with expectoration. For two weeks prior to admission he had been expectorating
blood.

His previous health was good, except for an attack of pneumonia (date
not given), and an ischiorectal abscess, which he had before coming to
America in 1920. He also had measles and typhoid fever (dates not given).
He had an operation for hemorrhoids in 1917. His appetite was always poor,
but he had no digestive disturbances until present illness. Venereal disease was
denied. He smoked cigarettes excessively. Weight before present illness was
130 pounds. Four of his brothers died of tuberculosis.

55 16
54 15

FOOD N

eODT WEIGHT

Hubert \V.

From Novcmbor 20 to 26 he was on the service of Dr. James Alexander
Miller, on which the following examination was made:

General appearance : Poorly developed and poorly nourished colored .Arabian
male lying quietly in bed with rapid respirations and productive cough. The
skin is hot, dry and shiny, with no visible cyanosis or rashes. Glands are not
enlarged. Extremities show no edema, clubbing, but marked wasting. Reflexes:
Knee jerks sluggish, Oppenheim and Babinski not obtained. The head is of
norrnal contour. Hair, thick and curly. Eyes: Pupils dilated, react normally
lo light and accommodation. There is no nystagmus, strabismus or ptosis.
Ears : No discharge or mastoid tenderness. Mouth : Teeth fairly good. Tongue
coated. Pharynx injected. Neck: No rigidity, no thyroid enlargement. Thorax:
Long and narrow. Retraction of both supraclavicular fossae. Expansion is

McCANX—PROTEIX REQUIREMEXT IX TUBERCULOSIS

39

limited throughout the left side. Fremitus poor throughout and marked dulriess
and bronchovesicual breathing above the left second rib and angle of scapula.
Below these there is bronchial breathing and pectoriloquy. There are numerous
fine and moist rales throughout the left side. The right side is resonant with
harsh breathing and increased whispered voice above the spine of the scapula.
No definite rales were heard. Heart: Apex impulse in fifth left intercostal
space. No enlargement, no thrills or murmurs. The pulmonary second sound
is accentuated. Rate rapid, rhythm regular, .■\bdomen : Scaphoid, no areas of
tenderness or rigidity. Liver and spleen not palpable.

7

FOOD N

M

Tzzmmzznza

21 28 29 30 OtCI Z

Fig. 2.— -Ab. M.

A provisional diagnosis of acute pneumonic phthisis was made. .Many
tubercle bacilli were found in the sputum. Urine normal. Temperature : High
with remissions, from 104 to 100 F. Pulse, from 100 to 120. Wassermann
reaction negative.

November 26: Admitted to metabolism ward (See Table 2 for data regarding
diet and excreta) (Fig. 2).

December 1 ; Passed ascaris lumbricoidcs in stool. Examination shows no
other ova present than those of ascaris.

40

ARCHIVES OF IXTERXAL MEDICIXE

December 3: Leukocyte count, 15,800; eosinophils, 5 per cent.; lymphocytes,
3 per cent. ; large mononuclears, 13 per cent. ; transitionals, 1 per cent., poly-
morphonuclears, 78 per cent.

November 29: Basal metabolism determined. Height, 168 cm. Weight, 40
kg. Body surface area, 1.42 sq. meters. Temperature, 101.6 F. Volume expired.
S. T. P. D.— 57.85 liters. Gas analysis: CO: per cent. 2.17, 2.20. O, per cent.
18.42, 18.40. R. Q. = 0.824, calories per hour, 72.74. Metabolism 130 per cent, of
average normal. Daily heat production (average temperature 101 F.) = 1,744
calories. Estimated basal requirement ^ 1,744 + 174 ^ 1.921 calories per diem.

The patient was observed in the metabolism ward from November
27 to December 6 inclusive. During this time he was extremely ill.
It was with the greatest difficulty that he could be persuaded to take
food. On three days vomiting occurred so that an unknown quantity
of food was lost.

-Abr.\ham M. (Case 2.)

i

1
11

1

II

1

s"
o

1

1

1

1

1

1

E

o

1
i

1

a

jl

1

a

1

1

-

-"■

^

u.

H

^

o

u

■27

2,360

m.x

123.8

257.6

60

10.83

0.70

11.53

—5.5

+440t

71»

96'

Period I

2,497

:«i.8

128.7

280.4

5.9

10.8.^

0.70

11. .53

-5.6

40.25

71-

29

2.497

.%.s

128.7

280.4

.5.9

10.83

0.70

11.53

—5.6

+577

7f

9(K

X»

1.375

H.4

140.0

1,3

9.72

0.70

10.42

—9.1

—545

211*

9H7

Vomited

1

2,497

36.8

128.7

280.4

5.9

11.13

0.70

11.83

-5.9

+577

38.99

227*

997

2

2,947

47.2

148.9

333.9

7,6

9.97

0.80

10.77

—3.2

+1,027

35.15

Period II

2,300

32.4

5.2

0.80

10.23

-5.0

Vomited

4

1,.'?fiO

32.K

70.7

138.6

5.2

10.40

11.20

—6.0

—560

39.39

Vomited

.S

;«.«

87.1

189.4

5 9

10.20

O80

11.00

—5.1

—182

38.70

H

2,184

39.0

119.1

223.4

6.2

10,86

O.SO

11.66

-5.5

+2M

38.6.

• Crea

in-fre

c diet.

1 Basa

1 rcg;

irenieiit estimateri

at 1,9

20 oal

Dries per dieii

The data regarding diet measurements and excreta analyses are
shown in Table 2, and graphically in Figure 2. It will be seen that
the urinary nitrogen figure at its lowest was almost double the food
nitrogen, in spite of the fact that on all but three days the energy
value of the food was in excess of the eslimnted basal requirement.

Of the ten cases in which the nitrogen metabolism has been
studied by us, this one shows the highest protein metabolism. It is
unique in our series, resembling much more closely the patient with
croupous pneumonia studied by Kocher " (Table 9). This patient
liad the physical signs of a pneumonic consolidation, with tubercle
Ijacilli in the sputum. It is not known to what extent oilier organisms
entered into the production of the pneumonia.

McCAXX—PROTEIX REQUIREMENT IX TUBERCULOSIS 41

This patient was the subject of an experiment to determine the
specific dynamic action of fat, and its eflfect upon the pulmonary venti-
lation. The data of this experiment are to be found in a previous
paper.-

Case 3. — Massive Bilateral InAltration of Lungs with Multiple Excavations. —
Fred R., a porter, aged 41 years, was admitted Nov. 23, 1921, complaining that
for three years he had had a chronic cough with progressive loss of weight
and strength. He had never enjoyed good health. In 1917 he began to have a
cough with a little sputum. He felt badly but continued to work fourteen hours
a day up to February, 1920, when he was obliged to quit work because of weak-
ness and shortness of breath. He then had three hemoptyses. He was admitted

99-
98-

k/\\

Fred R.

to the Scaview Hospital in April, 1920, and remained there until September,
1920. After his discharge he went to the country where he stayed until Nov.
15, 1920. Since his return to New York he has had attacks of "heart standing
still," which brought him into the hospital. The past history is chiefly irrelevant.
His best weight was 145 pounds, present weight 105 pounds. Two of his
brothers are known to have tuberculosis.

Physical Examination. — Height, 167 cm., weight, 47 kg. Fairly well devel-
oped, emaciated, lips slightly cyanotic, respirations rapid but with no subjective
distress. The head, eyes, ears, nose, mouth, throat and neck were normal.

Chest : There was marked emaciation with myoidema. There was marked
limitation of the respiratory excursion on the right side, with supraclavicular
retractions and of the lower right intercostal spaces. Both upper lobes were

42 ARCHIVES OF IXTERXAL MEDICINE

dull, especially the right. The whole right chest was dull, front and back. The
left lower chest was resonant, almost tympanitic. Over the whole chest numerous
fine and coarse rales and rubs were heard. The breath sounds over the left
upper lobe were harsh with prolonged expiration. Over the right upper lobe
there were areas of amphoric breathing, elsewhere bronchial, and over the right
lower lobe there was distant bronchial breathing.

Heart: Apex impulse was in the fifth intercostal space 11 cm. to the left of
the midline. Rhythm regular. No murmurs heard. The pulmonary second
sound was accentuated. Pulses synchronous, soft, and rapid. The vessels
were not sclerosed. Blood pressure, systolic 85 mm. The abdomen was
normal, except that the soft smooth liver edge was felt just below the costal
margin.

TABLE 3.— Fred R.

""

?.

r

K

s

i.

^

M

^

C^

5

~

-

g

P.

+'i

V

i-.

^

1

2

-S
5-|

=■

a

o

a"

1

i

1

M
1

s

m

?

S

1

§

s

n

II

U

°S

■&

1

-

r^

■■^

fe

o

u>

^

k;

o

u

w

4.2

2,524

37.9

128.8

286.2

«.o

5.37

1.12

6.49

—0.5

+454

3 9S2

CreatiD-frce

■>H

2,497

3K.K

128.7

280.4

5.9

5.09

1.12

6.21

-0.3

+427

0 957

48.,56

diet

280.4

5..57

HO

2,497

36.H

128.7

2S0.4

5.9

5.79

1.12

6.91

—1.0

+427

20 885

No urine lest

2„<!44

5.87

6.83

Period II

2.491

37.2

112.3

315.7

H.O

K.13

0.96

7.09

—1.1

+421

3

2,502

37.4

112.4

318.2

5.H>

0.96

-0.8

+432

48.08

2„«7

,%.S

6.18

+467

48.04

5

2,537

36.8

127.7

5.9

5.83

0.96

6.79

—0.9

+467

"

g

^7

443 4

4.8

4.8S

6.20

+934

47.65

Period III

3,131

30.4

105.6

493.8

4.9

4,95

1.32

6.1'.7

—1.4

+1,061

48.03

K

3,277

31.0

129.1

475.5

5 0

4 40

1.32

5,72

—0.7

+1,207

47.30

9..fm

4?:?A

4.83

1.32

fi.15

-1.3

+925

48.S0

10

30.5

118.1

422.4

4.9

4.70

1.32

6.02

-1.3

+885

48.35

2,(i47

144.8

9.26

47.46

Period IV

n

(3,239)

9Xn

183.0

282.0

14 9

8,00

48.71

Vomited

n

(9S«)

34.4

50.4

91.8

5.5

48.83

Died

Basal requirement 2.070 per •

Reflexes: .■Ml deep reflexes were exaggerated. There was an exhaustable
bilateral patellar and ankle clonus. Plantar reflexes were normal. The neck
was not stiff, and Kernig's sign was not elicited. Further examination showed
no abnormalities.

The urine was normal. Many tubercle bacilli were found in the sputum.
He was transferred to the metabolism ward November 26 and remained there
until his death on December 14. His diet and excreta are shown in Table 3 and
graphically in Figure 3.

His basal metabolism was determined December 2 and 11, using the Tissot
spirometer. The effect of ingestion of protein food on the heat production
and pulmonary ventilation was studied. These experiments are to be found
in a previous paper."

The basal metabolism was 12 per cent, above the average normal when the
temperature was 98 F. It was 28 per cent, above normal when the temperature
was 100 F., being 65.5 and 75.0 calories per hour, respectively. The elevation

McCANN— PROTEIN REQUIREMENT IN TUBERCULOSIS 43

of metabolism is believed to be due largely to the dyspnea. The basal heat
production is estimated to be between 1,620 and 1,880 calories per diem., ami
the basal requirement 2,068 calories.

Referring to the data in Table 3 it will be seen that, while the
urinary nitrogen fell practically to the level of ingested protein, there
was a persistent loss of nitrogen amounting to about 1 gm. per diem.
Assimiing muscle to be about one fifth protein, this nitrogen loss
would represent the equivalent of about 30 gm. muscle per diem or
about 1 kg. per month.

On the afternoon of the last day of period 3 the temperature rose to 102 F.
On the following day it was planned to raise the food intake to the equivalent
of 4,000 calories, with 90 gm. protein. The patient was unable to take the entire
amount. On the following day vomiting occurred, and the temperature con-
tinued to rise, with a marked increase in the pulse and respirations. The patient
continued to grow worse. Late in the afternoon of December 14 the patient
became very dyspneic, respirations from SO to 60 per minute, pulse from 130
to 140. Death ensued on the night of December 14.

Necropsy showed massive infiltration of the right lung, with huge cavities
but no erosion of vessels. The left lung showed extensive infiltration, with
small cavities, but still contained considerable air. There was no gross evidence
of a superimposed pneumonia.

C.^SE 4.— Infiltration of Both Upper Lobes.— yohn O'C, a porter, aged 49
years, admitted Dec. 3, 1920, complaining of pain in the right chest and lumbar
region, cough and expectoration. His mother died of tuberculosis. He had
been married twenty years, having five children living and well. He had been
well until 1889, when he had some pulmonary trouble for which he was confined
to bed for three months. After that he had a persistent cough. His general
condition remained fairly good up to 1910 when he began to work at night.
His weight which had been 170 pounds dropped to ISO pounds. In 1917 he
developed a migratory arthritis which involved many joints, for which he was
twice admitted to Bellevue Hospital. He had had a nystagmus all his life.
He had many attacks of quinsy sore throat. He had had a persistent cough
since 1889, with occasional streaks of blood in the sputum. No history of
dyspnea, edema or night sweats until present illness. He had had frequent
attacks of alcoholic gastritis. Appetite habitually good. Bowels constipated.
Until present illness patient had nocturia, no dysuria or other urinary symptoms.
Venereal disease was denied. Habits alcoholic. Weight: best, 170 pounds;
150 pounds in 1910; 133 pounds in 1917; present weight 110 pounds.

Present Illness. — Began in the middle of November, 1920, with pain over the
whole front of the right chest and in the right lumbar region. There was
considerable cough and some expectoration. The pain was increased by deep
breathing. He remained at work one week in this condition and then was
forced to spend the afternoons in bed. Growing worse he came to the hospital
for admission.

Physical Examination.— Height, 174.3 cm., weight, 49.7 kg. Patient was a
thin pale man appearing quite ill, lying flat in bed, very emaciated. There were
dilated venules on the nose. The lips and finger tips were quite cyanotic.
Respirations accelerated but not labored. Head: normal. Eyes: extraocular
movements well performed. Patient unable to fix gaze because of a slight
nystagmus, which is brought out strongly by looking to the left but not to
the right, with the slow component to the left. Pupils equal and regular, react
to light and accommodation. Ears and nose normal. Mouth : Teeth carious.

44 ARCHIfES OF I STERNAL MEDICIXE

with many missing in the upper jaw. There was some pyorrhea. Tongue
coated and tremulous. Pharynx congested. Tonsils normal. Larynx not
examined, though voice was husky. Neck : Normal except for spasm of
the right sternocleidomastoid muscle.

Thorax : of the long type with prominent sternum. There were marked
supraclavicular retractions. Respiratory movements were shallow w^ith greater
restriction on the right side.

Lungs: both upper lobes were dull front and back. The left lower lobe
showed little impairment of resonance. The right lower lobe was dull to

19 ?0 Z\ ?Z ti

knnwii am. unit of food los

flat behind, extending almost to midline in front. Over this area there was
diminished fremitus. Breath sounds over both upper lobes were bronchial,
with clicking rales heard during quiet breathing and showers of fine rales during
inspiration after coughing. At the right base the breath sounds are diminished
or absent behind, with distant bronchovesicular breathing laterally. At the left
base there was prolonged low pitched expiration of the emphysematous type.

Heart: Apex beat not seen and hardly felt, .^pex dulness in the fifth
intercostal space 8 cm. to the loft of midline. The right border was 7.5 cm. to

McCAXX—PROTEIX REQUIREMEXT /.V TUBERCULOSIS

the right in the third intercostal space. The rhythm is irregular (extrasystoles),
rate rapid (90). No murmurs were heard. The arteries are somewhat thick-
ened. The pulses equal and synchronous. Blood pressure : systolic, 128 ;
diastolic, 100 mm.

Abdomen : Flat, rather rigid in the right upper quadrant. Liver and spleen
were not felt. No tenderness or masses felt. Further examination revealed
no abnormalities.

On admission the temperature was 101 F., and during the ne.xt five days it
rose gradually to 102 F., in the afternoon. The urine was normal. The sputum
contained many tubercle bacilli. Leukocytes, 9.S00, with a normal differential
count. Roentgen-ray examination (No. 84,275), December 14, showed evidence
of narrowing of the intercostal spaces on the right side with marked thickening
of the pleura of the right lung. There was peribronchial infiltration and
agglutination of the foci in the upper parts of both pulmonic fields. The heart
was slightly retracted to the right. .\ small effusion in the right costophrenic
sinus could not be ruled out.

TABLE 4

—John O'C

1

j

a

a
o

a

a

1

E

1

i

Remarks

1

0

0

1

Calories ± Basal

5

a

S

o

^

S!

SB

S5

z

t

JJ-

o

1

1

—

■=

a

>;

s

1

s

i

s

0
§

g

0

«

1

"^

H

a<

Pu

o

6.

S

&H

H

*=

9

2,505

37.1

10O.3

346.4

5.9

9.40

0.65

10.05

—42

50.71

+ 895 calories

10

2,503

37.3

100.9

6.0

6.44

0.65

7.09

51.23

+ 893 calories

11

2,515

37.3

100.9

347:3

6.0

6.44

0.65

7.09

— l!l

51.41

+ 905 calories

12

2,371

31.8

94.6

331.9

5.1

4.71

0.65

5.36

-0.3

51.06

Vomited part oi food

13

2,257

30.2

97.5

4.8

5.69

0.65

6.34

50.50

Vomited part ol food

14

3.004

37.2

139.2

381.9

6.0

4.00

0.53

49.73

+ 1,394 calories

15

2,982

37.2

143.1

365.4

6.0

5.33

0.53

5!86

+o!i

50.80

Vomited at 11 p. m.

16

3,422

37.8

179.7

389.2

6.0

3.64

0.53

4.17

51.10

+ 1,812 calories

1-

3.427

37.6

161.4

432.3

6.0

3.94

0.53

4.47

51 ..54

+ 1,817 calorie.*

18

3,(M6

28.7

149.7

374.6

4.6

5.02

5.55

50.89

Vomited part of food

19

2,984

32.3

143.7

369.7

5.2

4.26

o!53

4.79

+0:4

Vomited pait of food

20

2,601

64.4

132.1

270.2

10.3

6.58

1.0*

7.6

+2.7

49:33

21

1.776

52.4

105.5

141.4

8.4

6.01

0.8*

6.8

+1.6

49.76

22

62.7

123.6

267.6

10.0

8.95

1.0*

10.0

0

50.00

23

i;823

64.3

63.6

236.1

10.3

49.50

Comment. — The patient was in the metabolism ward from December 9 to
20. The data regarding his diet and analyses of excreta are given in Table 4,
and shown graphically with the temperature chart in Figure 4.

His basal metabolism was determined on two occasions, using the Tissot
spirometer. December 14, with a temperature of 100.8 F., his metabolism was
3 per cent, above the average normal. December 23, it was exactly normal with
a temperature of 1(X).6 F. The heat production was 62 and 61 calories per
hour, respectively, on the two occasions. The basal heat production for twenty-
four hours is 1.464 calories. Adding 10 per cent, for the specific dynamic
action of food the basal requirement is estimated at 1,610 calories.

.^n experiment was done to study the effect of a small protein meal on the
heat production and pulmonary ventilation. The experimental data of this ami
of the basal determinations are given in the preceding communication.'

Referring to the data in Table 4 it will be seen that the lowest
urinary nitrogen excretion was reached December 16. This low figure
was not obtained by reduction of the nitrogen intake to an extremely

46 ARCHIVES OF IXTERX.IL MEDICIXE

low level, but rather by the ingestion of a large quantity of carbohy-
drate and fat. In spite of considerable fever, it was possible to
achieve positive nitrogen balance by the ingestion of from 37 to 38
gm. protein when the total caloric value of the food was more than
twice the estimated basal requirement of energy. The patient, how-
ever showed a marked tendency to digestive disturbances, and vomit-
ing resulted frequently when attempts were made to increase the
diet. The diet was much better taken when 60 or 65 gm. protein
were given, and it will be seen that in this case nitrogen balance
could be achieved by the ingestion of much less nonprotein food.

Case 5. — hiUltration, Right Upper and Middle Lobes, u'ith Caz-itation. —
Frank D., a plumber's helper, aged 17 years, was admitted Jan. 30, 1921, com-
plaining of cough and weakness. Family history rather indefinite. Patient was
raised in an orphanage. He had had measles, chickenpox and mumps. Each
year he had attacks of severe sore throat. After a severe illness he was treated
for one year for "heart trouble." He never had arthritis. He gave no history
of any cardiorespiratory symptoms prior to present illness. Gastro-intestinal,
urinary, and venereal history were negative. After leaving the orphanage he
was employed at night doing very heavy work for two months. He became
very much "run down." His weight fell from 125 to 97 pounds. He \yas out
of work for a month, after which he found that he was too weak to work.
He was then supported by another boy from the orphanage, who could allow
him only $0.40 a day for food. For two months before admission he coughed,
at first slightly, later severely with yellowish sputum. Three weeks before
admission he had become so weak he remained in bed. Two weeks later he
coughed up about one and one half ounces of blood.

Physical Examination. — Height, 16S.5 cm., weight, 43 kg. A thin pale weak
looking boy lying quietly in bed without respiratory embarrassment. No
abnormalities were found other than those in the thorax.

Thorax : The clavicles were prominent, supraclavicular fossae well marked,
especially of the right side. Respiratory excursion more limited on the right side.

Lungs : Both apices were dull, on the right side from the second rib to
midscapula, on the left above the clavicle and spine of scapula. Over both dull
areas the breath sounds were modified, bronchovesicular to bronchial,' with
many crepitant and subcrepitant rales. Resonance and breath sounds over the
lower lobes were normal.

Heart: Apex beat, somewhat diffuse, was seen in the fifth intercostal space
inside the nipple line. The area of dulness extends to the left 7.5 cm. in the
fifth space and to the right 2.5 cm. in the third space. No murmurs were heard.
The rate was rapid. There was a respiratory arrhythmia. The pulmonary
second sound was accentuated. Pulses soft, rapid, equal and synchronous.
Blood pressure, systolic 98 mm., and diastolic 70 mm.

On admission the temperature was 101 F. ; pulse, 100: respirations, 24. The
urine was normal. Tubercle bacilli were present in the sputum. Blood examina-
tion showed: hemoglobin, 79 per cent.; red blood cells, 4,320,000; leukocytes,
10,700 with 31 per cent, lymphocytes.

Roentgen-ray examination (No. 89,021) showed a localized area of peri-
bronchial infiltration at the right upper lobe. There was considerable produc-
tion of fibrous tissue with the presence of a cavity between the first and second
ribs. There were also several foci of peribronchial infiltration in the middle
lobe of the right lung.

The basal metabolism was determined with the Tissot apparatus Feb. 4, 1921.
Height, 165.5 cm., weight, 43.24 kg. Volume expired, S. T. P. D., 363.1 liters per
hour. Gas analvsis : CO: per cent. 3.10; 3.11 : O. per cent. 17.69. 17.72. R. Q. =

.IfcCAXX—PROTEIX REQUIREMEXT IX TUBERCULOSIS 47

0.945. Calories per hour = 58.6. Calories per twenty-four hours ^= 1,406. The
metabolism was 5 per cent, below the average normal for his age. The
basal requirement, allowing 10 per cent, for specific dynamic action, was 1,547
calcries per diem. The subject was not very suitable for respiration experi-

ments because of a persistent tendency to overventilate, which produced
"Auspumpung" of carbon dioxid.

March 5 the metabolism was again determined. The weight was greater,
46.7 kg. Calories per hour 61.0 The metabolism was 4 per cent, below the
average normal. Basal requirement = 1.610 calories.

ARCHIl'ES OF IXTERXAL MEDICJXE

He was observed in the metabolism ward from February 5 to March 7, 1921.
The data regarding his diet and analyses of excreta are shown in Table 5. and
graphically in Figure 5.

TABLE 5.— Frank D.

^

«■

s

■%

g.

N?

f

s

s
c

a

S

S

a

5

+IE

r

B

i
11

O

5
1

1

i

.H

i

o

1

^

^

^

h

»

1.32

19,97

--,1

+422t

=5

5

1,969

93.0

59.4

252.6

14.9

18.65

43.2

6

■ 3,101

93.8

116.2

15.0

11.31

12.63

+1,554

2,863

87.4

133.1

309.1

13.06

-0.4

2,906

89.9

129.8

324.4

14.4

13.95

1.32

15.27

-0.9

+i;359

42.08

9

3,141

93.0

14.9

12.45

1.32
1.94

13.77
16.98

+1.1
+2.8

+1,594
+2,0931

43.06

10

3,640

123.8

172.5

372.8

19.8

15.04

43.28

11

3,377

126.2

133.3

395.2

20.2

16.43

1.94

18.37

+1.8

+1,830

43.77

12

3,455

125.7

166.9

338.5

20.1

17.86

l.(M

19.80

to.3

+1,908

225.0

19.8

+0.1

U

3,573

123.V

,63.6

376.9

19.8

15.62

IM

17.56

+2,026

43.6

15

3,564

63 4

467.8

11.76

Ifi

3,634

51.1

171.9

445.1

8.2

9.81

38.2

149.3

485.1

6.1

5.59

18

3,470

26.6

121.4

519.9

4.3

5.23

44 88

19

3,544

45.7

149.4

480.0

7.3

4.63

45.17

20

3,783

44.8

143.8

651.7

7.2

4,22

3,568

43.8

156.9

470.5

4.20

■46 8

22

3,344
3,654

40.4

143.0

176.6

450.8
444.9

6.5

4.52

1.25

5.53

23

45.9

7.4

4.28

+1.9

45 69

181.9

387.0

+1.5

+1,850

25

3,434

51.0

159.4

425.1

8.2

1.25

5.63

+2.7

+1,824

45.65

2e

3,489

39.8

159.8

448.8

6.4

3.98

1.25

6.23

+1.2

+1,879

45.85

27

3,156

49.1

134.7

415.2

7.9

4.37

1.25

5.62

+2.3

+1,546

28

45.4

168.0

560.0

7.3

3.75

1.82

6.57

+1.7

+2,434*

45.85

45.0

5S0.6

7.2

3.57

1.82

5.39

+1.8

+2,398

2

3,804

43.4

135.4

677.4

6.9

3.60

1.82

6.42

+1.5

+2,194

46.12

3

44.0

133.2

531.6

7.0

4.00

1.82

6.82

+1.2

+ 1,989

46 33

507.9

7.2

3,59

1.82

5.41

+1.8

+1,964

46.66

5

3,272

47.3

132.0

451.1

7.6

1.82

6.45

+1.1

+1,662

6

4,150

44.6

202.3

608.7

7.1

4.22

! 1 '

* Basal requirement estimated at 1,610 calories per diem,
t Basal requirement estimated at 1,547 calories per diem.

The observations were divided into five periods, in four of which the feces
were separated and analyzed. In the first two periods the protein ingested was
about 90 and 125 gm. per diem, respectively, with enough nonprotein calories
to greatly exceed the requirement.

^^^'oTclZ'^L

Average Protein
Daily, Gm.

Average Positive
Nitrogen Balance

3,008
3,632
3,439
3,717

91
124.7
46
45

0.55
1.44
1.9
1.5

In period three the average daily calories ingested was 3,4.30. The food
nitrogen was rapidly reduced to 4.3 gm., the urine nitrogen output falling
gradually to 5.2 gm. At this point the food nitrogen intake was again raised
to about 7 gm., (period 4), the urine nitrogen continuing to diminish.

McCAXX-PROTEIX REQUIREMENT IN TUBERCULOSIS 49

Case 6.— Infiltration of Left Apex and Right Hilum, with Large Cavities.—
George R., a truck driver, aged 35 years, was admitted Jan. 4, 1921, complaining
of fistula in ano. cough, loss of weight, and night sweats. Family history negative.

Past History.— Genera] health not impaired until two months before admission.
He had had fistula in ano for three years. In childhood he had measles only.
Xo history of other infections, operations or injuries. He had never had a
cough before Julv, 1920. and had never had dyspnea or edema. His appetite
was always good and his bowels regular. There was never any pam on

Fig. 6.— George R.

defecation though for three years he had had a fistula discharging pus. In
January, 1920, a second fistula appeared, and in December, 1920, a third. He
had never had anv urinary symptoms. In 1908, he had a chancroid infection
in the left inguinal glands. He had recently become a heavy drinker in spite
of prohibition. His weight rose from ISO to 172 pounds in 1918, but fell
again during the summer of 1920 to 144 pounds. Present weight 126 pounds.
Present Illness.-Ue began to cough in July, 1920. At first the cough was
In December, 1920, he began to notice small clots of

dry, later productive,
blood in the sputum. He

)f afternoon fever, but about

ARCHIVES OF IXTERXAL MEDICINE

TABLE 6.— Gf.orge R. (Case 6)

u

1

i

_

s

fH

0

i.^

M

■g

i

g

a

Si^

^

J

a

a

3

i

0

4

S3

s

1-5

1

0

■S

0

0

la

0

ia

1

i

1

1

1

5

1

.1

5

1

1

5

7'.

¥

1

5

2,608

62.6

112.5

294.0

10.0

11.20

....

+724

57.1

6

2,597

67.8

122.2

288.6

10.8

10.22

+813

57.1

7

2,581

66.1

120.5

300.0

9.0

9.16

+797

58.1

8

2.760

51.3

111.3

369.6

8.2

8.00

....

+976

57.4

9

3,375

44.4

134.9

472.7

7.1

7.08

....

+1,591

10

2,910

37.6

126.5

385.4

6.0

6.02

+1,126

57.0

11

2.968

31.6

131.3

394.5

5.1

5.54

+i:i84

57.3

12

3,094

25.2

129.8

435.1

4.0

4.88

+1,310

57.1

13

3,012

19.6

128.6

423.4

3.1

4.48

+1,228

57.1

14

3,038

42.4

129.9

403.8

6.8

4.86

).S4

5.70

+1.1

+1,254

,t6.8

15

2,922

46.3

135.4

359.3

7.4

5.57

).&4

6.41

+1.0

+1,138

57.4

16

2,928

40.8

131.7

374.6

6.5

5.89

).84

6.73

—0.2

+1,144

17

2,919

41.5

129.1

377.6

6.6

5.69

).84

+0.1

+1,135

57.2

18

2,877

45.5

129.5

362.7

7.3

5.42

).84

6:26

+1.0

+1,093

57.5

19

2,906

45.7

134.2

358.8

7.3

).84

7.19

+0.1

+1,122

57.6

20

2.933

44.6

128.4

379.5

5:05

).84

5.89

+1.2

+1,149

57.9

21

4,013

94.8

180.1

475.3

15.2

7.24

8.8

+6.4

+2,229

22

3,916

94.4

198.6

410.2

15.1

7.85

9.4

+5.7

+2,132

59.8

23

3,695

173.8

413.2

15.0

10.37

•

11.9

+3.1

+ 1.911

24

3.556

goio

179.0

15.4

9.50

11.0

+4.4

+1.T72

58.T

25

3,631

94.2

174.3

396:0

15.1

9.27

•

10.8

+4.3

+1.S47

57.4

26

3,744

96.7

166.0

439.9

15.5

10.05

11.6

+3.9

+1,960

58.1

27

3,674

177.6

15.0

11.61

13.1

+1.9

+1,890

iiO.5

28

3,918

208.0

14.3

9.26

10.7

+3.6

+2 134

.-8.C

29

3,659

94:0

180.3

389:6

15.0

9.59

»

11.1

+3.9

+ 1.875

30

3,935

125.6

201.6

376.8

20.1

12.74

14.7

+5.4

+2,151

3;824

125.4

179.5

400.2

13.05

15.1

+6.0

+2,W0

59.8

3.795

127.1

182.2

385.4

20:3

14.08 1

•

16.1

+4.2

+2,011

60.6

3,976

123.8

204.1

383.0

19.8

12.53

14.5

+5.»

+2,192

59.9

4;042

125.7

189.8

429.8

12.73

14.7

+6.4

+2,258

61.4

4.249

125.6

223.7

403.3

20:1

14.56

•

16.6

+3.5

+2.465

10.8

4,169

125.2

227.3

376.0

20.0

15.56

17.6

+2.4

+2,385

61.0

3;897

124.3

191.5

391.8

20.0

15.03

17.0

+3.0

+2.113

3,934

131.6

191.5

393.4

21.4

15.57

•

17.7

+2.150

61.2

2,775

123.4

131.5

255.1

19.7

18.01

20.0

—0:3

+991

61.3

3,892

124.4

212.6

342.7

19.9

17.16

19.2

+0.7

+2,108

61.1

3.603

124.5

175.6

356.2

19.9

16.06

18.1

+1.8

+1819

61.6

3,901

127.1

184.6

405.5

20.3

14.77 1

•

16.8

+3.5

+2,117

61.4

3,714

125.4

161.9

413.1

20.1

15.58

17.6

+2.5

+1,930

61.7

4.043

123.5

196.7

401.5

19.8

16.32

•

18.3

+1.5

+H^

3.567

125.9

162.9

374.2

20.1

15.30

17.3

+2.8

+1,783

62.6

3:781

123.9

168.2

416.8

19.7

15.69

17.7

+2.0

+1,997

02.4

3.896

90.6

174.6

463.7

14.5

13.00

14.5

0

+2,112

3.818

89.1

145.0

513.1

14.3

10.98

•

12.4

+1.9

+2,034

62.8

3,511

50.4

173.6

412.2

8.1

8.58

9.4

—1.3

+1,727

03.4

3.729

62.4

137.3

545.5

8.4

5.93

6.8

+1.6

+1,W5

62.3

20

3.399

49.5

151.2

436.7

7.9

5.90

•

6.7

+1.2

+1,615

21

3:780

48.9

160.7

508.6

7.8

4.98

6.8

+2.0

+1,996

62.4

22

3.268

49.7

166.1

370.4

8.0

5.59

6.4

+1.6

+1,484

23

3,626

49.9

192.4

398.0

8.0

5.46

*

6.3

+H

+i'i^

63.4

24

3,589

48.0

169.1

443.6

7.7

5.2s

6.0

+1.7

+1,805

• 63.1

25

3,428

47.7

146.1

456.3

7.6

5.22

6.0

+1.6

+1,644

62.7

26

3,525

43.6

176.9

417.1

7.0

4.56

•

6.8

+1-1

+1.741

63.1

27

3:527

53.1

104.1

434.8

8.5

5.60

*

6.6

+2.0

+1,743

McCA.W—FROTEIX REQUIREMENT IX TUBERCULOSIS 51

July. 1920, night sweats began and had persisted. Appetite was still good, but
patient was weak and short of breath on exertion.

Physical Examination. — Height. 163.5 cm.; weight, 57 kg. Patient was a
young man with a full red face, sitting in bed, coughing occasionally, not acutely
ill. He was fairly well nourished, with a rather flabby panniculus adiposus.

Head : Eyes, ears, and nose were normal. Mouth : Most of teeth present,
moderate pyorrhea and dental caries. Pharynx and tonsils were normal. Neck
normal. Thorax : Well formed and symmetrical. Respiratory excursions were
equally limited.

Lungs : Both upper lobes were dull above the clavicles and behind as far
as the midscapula. Anteriorly resonance was fairly good. Breath sounds were
modified in the right axilla and along the border of the pectoral muscles there
was bronchovesicular breathing and fine rales, and over the rest of the front there
were moist and sonorous rales. Over both apices behind fine rales were heard
on quiet breathing, chiefly, on the left side. These rales were exaggerated by
coughing. At the left apex just below the tip a squeaking post-tussive sound
was heard. Breath sounds were bronchial or bronchovesicular with showers
of fine rales.

Heart: Apex in fifth intercostal space 10 cm. to left of midline. Cardiac
dulness extended to the left II cm. in the fifth space, and to the right 3 cm. in the
third space. The sounds were of good quality, with no murmurs. Rhythm
regular and rate slow. The pulmonary second sound was accentuated. Pulses
soft and synchronous. Arteries normal. Blood pressure: systolic, 98;
diastolic, 60.

Further examination revealed no other abnormalities, with the exception
of the three fistulas which were situated close to the anus.

The sputum contained many tubercle bacilH. The urine was normal. Fluoro-
scopic examination showed pulmonary fields of equal size, diminished in
illumination of both upper halves. There was a dense area running out from
the hilus on the right side to the ribs. Above this area there was an intensely
illuminated circular area which was thought not to be an antrum because lung
markings could be seen through it. At the left apex there was seen to be an
antrum, w'hich was brought out by coughing, which was undoubtedly the
source of the post-tussive sound heard in that region.

January 12 : leukocyte count 10,000 with normal differential count.

January 18 : leukocytes 15,000 with normal differential count ; red blood
cells, 4,200,000 ; hemoglobin, 75 per cent.

The patient was in the metabolism ward from January 5 to February 28.
The diet record and the analyses of excreta are given in Table 6 and shown
graphically in Figure 6.

The basal metabolism was determined January 26, and it was found to be
6 per cent, above the average normal, with a heat production of 67.6 calories
per hour. The twenty-four hour basal heat production is estimated, therefore,
to be 1,622 calorics. If one adds 10 per cent, to this to allow for the specific
djtiamic of food the basal requirement equals 1,784 calories per diem.

An experiment was done to determine the effect of ingesting 350 gm.
meat. The data for these observations is contained in a previous communication.'

Comment. — The data in Table 6 show that from January 5 to 13 the food
nitrogen was reduced about 1 gm. daily, while during the same period the total
caloric value of the diet was increased from 2,500 to about 3,000 calories. Under
these circumstances the urine nitrogen decreased parallel with the food nitrogen
until the latter reached 6.0 gm. per diem, .^fter this point was reached the
urinary nitrogen decreased more slowly until with a minimum ingestion of 3.1
gm. food nitrogen the urinary nitrogen was 4.48 gm.

52 ARCHIVES OF IXTERXAL MEDICIXE

For the remaining periods of observation the following summary is given :

Jan. 14-20
Jan. 21-29
Jan. 30-Feb. ID
Feb. 19-27

Average Daily Average Daily Pro- .\verage Daily

Calories in Food tein in Food, Gm. Nitrogen Balance, Gin.

2,932 43.8 + 0.67

341

It will be seen, therefore, that 43.8 gm. protein may be fairly taken as
the minimum for this patient if a total of 2,500 calories per diem are ingested.
It is quite clear that in this case there was no advantage to be had in giving
more than 94 gm. protein per diem. Also in order to make large gains in
nitrogen it was necessary to give a diet of which the energj- value was about
twice the basal energy requirement.

Entirely satisfactory gains in nitrogen and weight occurred during the last
period, February 19 to 27. In this period 192 gm. was the maximum fat ingested,
so that it was necessary to give between 370 and 545 gm. carbohydrate. From the
standpoint of sparing the ventilation of the lungs this diet would not be satis-
factory, because the increased breathing volume after taking this amount of
carbohydrate would be greater than the effect of a diet containing more protein
and less carbohydrate.

At the end of the period of observation the patient was transferred to the
tuberculosis service. He had gained 14.8 kg. in weight, but in spite of the
improved nutrition no improvement could be noted in his clinical condition.

DISCUSSION OF RESULTS

The Nitrogen Minimum. — Table 7 is a summary in which are
arranged the data for each subject for the day on which the lowest
urinary nitrogen excretion was noted. The first four subjects were
studied in 1919-1920, and the complete data regarding them are to be
found in the previous communication of McCann and Barr.^ The
total length of the observation in days is given. The number of the
day selected for Table 7 is given for easier reference to the original
data in preceding tables.

T.^BLE 7. — SuMM.\RY OF Nitrogen Minimum Expf.riments in Tuberculosis

•r;

?

a

1

4

^

Nome of
Subject

"1

E

.8

1

1

1

1
S5

1

1

i

It

So

-J

^2B
III

¥b

11

11

-:

1

1

I

^

u

=

.1

il

"

^

^

6,

'^

CliarlosG.'

24

IB

2,091

40.1

W.9

2.8

4.3

(1.0)

(5.31

(-•'5)

0.054

0.083

43

1.1

(Jeorge P.-

12

10

2,3SO

41.0

5S.1

2.9

3.9

(1.0)

(-2.01

0.0.W

0,067

70

1.2

1.914

(1.01

0,070

Joseph E.'

Hi

14

2,493

41.4

60.4

2,5

(1.01

(3.51

(-0.2)

0.05,1

0.041

54

1.3

IB

2,493

42.0

3.3

3.9

(1.01

(4.9)

(-1.6)

0.056

0.066

54

1.3

2,312

5fi.4

4.3

3.8

2.3

6.1

1.3

Fred B.

18

47,3

John O'C.

l.l

K

3,422

Bfifl

51.1

6,0

3,6

0.53

4.2

+1.8

0.117

0.071

47

2.1

Ab. M.

in

2,300

B5.3

0.8

10.2

-5.0

0.14,8

0.267

15

1.2

Frank D.

•so

2fl

.3,fi(M

S2.ri

46.1

«.»

8.«

1.R

5.4

+1.5

0.160

0078

62

2.4

3.574

7«.B

46.7

7,'/

3,»

1.8

5.4

+1.8

0.1.->4

0.077

58

George R.

54

"

3,012

52.7

57.1

3.1

4.5

0.9

5.4

-2.3

0.054

0.079

58

1.7

Coses of MeCann and Ifarr, 1920.

McCAXX—PROTEl.X REQllREMEXT IX TUBERCrLOSIS 53

In the first five cases and in the last case shown in Table 7
the nitrogen minimum was attained by reducing the nitrogen intake
to a low level (from 2.8 to 4.3 gm. N), at the same time giving a
diet, of which the caloric value varied from 1.1 to 1.7 times the
basal requirement of the patient. Wlien these six cases are compared
with the experiments on normals, which are compiled in Table 8,
it will be seen that the urinary nitrogen excretion, in terms of grams
per diem, fell to approximately the same level as in most of the
normal subjects. A diminution of urinary nitrogen elimination
occurred with about the same promptness in both series, following a
reduction in food nitrogen. This can best be seen by consulting the
graphic figures.

\\'hile in grams per diem the urinary nitrogen excretion of the
tuberculous patients approaches closely that of normals, when one
estimates the amount excreted per kilogram of body weight it is
seen to be much higher for the tuberculous series. However, it was
not to be expected that the wear and tear quota should vary directly
with total body mass. It is rather more probable that it varies with
the total metabolism.

In the remaining four cases shown in Table 7 the protein in the
diet was only moderately reduced (to the equivalent of 5 or 7 gm.
nitrogen daily). The attempt was made rather to determine to what
extent the nitrogen excretion could be reduced by the protein sparing
action of a diet rich in nonnitrogenous foodstuiTs. Positive nitrogen
balance was achieved in two of these cases by the ingestion of 6.0
and 6.9 gm. nitrogen, respectively (from 37.5 to 43.1 gm. protein),
when the diet furnished from 2.1 to 2.4 times the basal requirement of
energy. It will be noted that in the case of George R. (Table 6),
that with a diet furnishing 2,932 calories and 43.8 gm. protein an
average of 0.67 gm. nitrogen was added to the body daily for a seven
day period. Subject Ab. M. showed persistent large losses of nitro-
gen. It was impossible to give him adequate amounts of food because
of digestive disturbances. He had physical signs of « pneumonic
consolidation, and it is not known to what extent organisms other
than the tubercle bacillus had entered into the production of the
pneumonia.

There is evidence that the quality of the protein supplied is an
important factor in determining the amount which will cover a given
wear and tear quota. Thomas =' studied the biologic value of various
proteins. First, he reached his nitrogen minimum by taking a nitrogen-
poor diet. He then found the relative amounts of various proteins

21. Thomas, K. : Ueber die biologi.sche Werthigkeit der Stickstoffsubstancen
in Verschiedenen Nahrungsmittein, .\rcli. f. Physiol, 219, 1909.

54 ARCHIVES OF IXTERXAL MEDICIXE

required to cover this known wear and tear quota. A few of
Thomas' results are given for purposes of illustration. Each 100
gm. of the following proteins will cover the equivalent of the follow-
ing number of grams of flesh :

O.x-flesh 104.74 gm.

Milk • 99.71 gm.

Fish 94.46 gm.

Potato 78.89 gm.

Flour 39.56 gm.

Maize 29.52 gm.

In our three cases in which positive nitrogen balance was obtained
with a low protein diet the source of animal protein was chiefly from
milk, lean beef, and chicken, with occasionally a little cheese and
eggs. In the case of John O'C, 52 per cent, of the protein was
from animal sources ; with Frank D., 42 per cent. ; and with George
R., 47 per cent.

In Tables 7 and 8 the percentage of total calories in the form
of carbohydrate is given. In the tuberculous series this varies from
15 to 70 per cent., while in the normal series it varies from 38
to 100 per cent, of the total calories. In the case of Ab. M., who
showed the greatest nitrogen excretion, only 15 per cent, of the
calories were from carbohydrate. In this instance the table may be
misleading, for if one refers to Table 2 it is apparent that the nitrogen
loss was greater on several days on which a much higher percentage
of carbohydrate calories was taken.

There has been a great deal of work done to determine the rela-
tive efficiency of carbohydrate and fat as protein sparers. Kayser "'^
found that he could not replace carbohydrate in a diet with an isody-
namic amount of fat without a rise in the output of nitrogen. Voit
and Korkunoflf^^ showed that when protein is fed with fat the
fall in protein catabolism, although it reaches a level well below that
for protein alone, is not so marked as in the case of protein plus
carbohydrate.' However, Tallquist " in Rubner's laboratory found
that, keeping the same carbon: nitrogen ratio, the relative propor-
tions of fat and carbohydrate in the diet could undergo considerable
variation without disturbing the nitrogen balance greatly. Lander-
gren '■''^ made a clear demonstration of the difference in sparing action

22. Kayser: Ucber die Eiweisssparende Kraft des Fettes vergleichen mit
derjenigen des Kohlenhydrats. DuBois-Reymond Arch. f. Physiol., 371, 1893.

23. Voit, E., and Korkunoff: Ueber die geringste zur Erhaltimg des Sticks-
toffgleichgewiclites nocthige Menge von Eiweiss, Ztschr. f. Biol. 32:58, 1895.

24. Tallquist: Arch. f. Hyg. 41:177, 1902.

25. Landcrgren : Ueber die Eiweissumsetzung des Mcnschen, Skand. Arch,
f. Physiol. 14:112, 1903.

McCANX— PROTEIN REQUIREMEXT IX TUBERCULOSIS 55

of the two foodstuffs. There was a steady fall in the nitrogen out-
put on an exclusively carbohydrate diet, and when the diet was changed
to one exclusively of fat there was a steady and progressive rise.
Finally Zeller ^® fed both dogs and men with varying quantities
of fat and carbohydrate. He came to the conclusion that from 70
to 90 per cent, of the carbohydrate in a diet could be replaced by an
isodynamic amount of fat without the nitrogen minimum, reached by
exclusive carbohydrate feeding, being materially affected. The gen-

TABLE 8. — Summary of Nitrogen Minimum Experiments on Normals

^

g

c

M

55. t

i

H

If

R

fi

F

R

i

II

f

Author and
Subject

=- 8

t

S

i5

n

a
2

K

So

S

V,i

¥

¥

^

1

1

B

1

1

z

^1
1^

£

e
^

Hirschftid

II 1888

8 |fr«

3,462

47.0

73.0

7.46

5.76

1.65

7.41

+0.05

0.102

0.079

(IS)

Hempercr

8

IT

(Mi
42

b,m

2,441

77.0
42.0

65.3
58.9

2.43

2.51

1.02
1.33

3.63
3.17

+1.75

0.081
0.041

0.038
0.030

38
67

(14)

Siven 1898

-0.74

(15)

Landergren

4

4

3,374

45.2

V3.4

0.82

().'/.'>

4.(.1

-3.69

0.011

0.051

95

(16)

4 4

3.163

37.8 ! 79.1

2..'.

3.»f.

1.47

5,42

0.032

0.030

IV

2,920

43.0 1 62.4

2.05

3.04

1.02

-2.01

44

V

4 4

3,089

38.4 1 77.3

2.4

4.21)

1.33

5..W

—2.13

0.031

0.054

43

VI

—

2,745

73.4

2.2

4.95

6.23

0.030

0.067

53

(MadseD)

Av.

VII-15

19

per.

3,796

53.0

n.5

3.62'

3.41

....

+0.21

0.051

0.048

61

(17)

Kochcr

R. A. K.

10

5

5,089

64.0

79.2

1.01

4.0M

—3.07

0.013

0.037

1(10

J. G. F.

lU

4

5,089

72.0

70.4

1.01

2.89

1.13

4.02

-3.01

0.014

0.041

100

(4)

• Utillzable nitrogrn

13. Hirschleld: Vird

14. Hempcrer: Zeit-<

15. Siven: Skand. '
10. Lanclorgren: Sk.;'
17. Hindhede: SkauO -u

eral correctness of Zeller's conclusions seems to be borne out by the
results in Table 7 and by the experiments of Landergren shown in
Table 8.

Attention should next be turned to the observations carried out on
Frank D., and George R. during periods of forced feeding. In the
case of Frank D. a very striking result was obtained, which was
summarized under the case history. The gain of the body in nitrogen
was greater on the low protein diets of periods 4 and 5. This was
not due to any apparent diminution in the activity of the tuberculous

26. Zeller: Einfluss von Fett unci Kohlenhydrat bei Eiwei-sshunger auf die
Stickstoffausscheidung, Arch. f. Physiol. 2:3, 1914.

56 ARCHIVES OF IXTERXAL MEDICIXE

process during the latter periods, as evidenced by the physical signs or
by the temperature record (Fig. 6). It should be pointed out that
the subject was only 17 years old, incompletely developed, and that
he had suffered from inanition before coming under observation. How-
ever, the periods of high protein feeding preceded the low periods, so
that the body had the greater opportunity to store nitrogen first.
Recently, von Hoesslin -' in Germany has studied groups of individuals
who were suffering from severe inanition. He noted an abnormally
high nitrogen retention during periods of relatively low protein and
caloric intake, and that this retention was but little influenced by the
caloric value of the food.

In the case of George R., who had been better nourished previously,
it was seen that although a slight gain in nitrogen was made when
43.8 gm. protein were ingested, the retained nitrogen increased as
the protein ingestion increased. The maximum retention was noted
when there was 90 gm. protein in the diet. Less nitrogen was stored
when the protein intake was increased to 125 gm. daily.

Extensive and very valuable studies have been made by Bardswell
and Goodbody -- on the effect of large diets on patients with pulmonary
tuberculosis. Some of their patients received 270 gm. protein with
about 5,000 calories, others more moderate diets with 160 gm. protein
and 3,400 calories. The clinical results with the more moderate diet were
satisfactory, in fact "the patients made much less satisfactory all-round
progress on the very large diets. . . ." Failure of appetite, marked
digestive and intestinal derangements were noted. "\\'hen the amount
of proteid in the diet was much increased, it resulted in an increased
excretion of nitrogen out of all proportion to the increased amount
retained in the body." The percentage of nitrogen excreted as urea
fell. There was an increase in the amounts of aromatic sulphates
excreted. Generally the large gains in weight were not permanent, but
disappeared on return to a normal diet. Dyspnea was frequently com-
jjlained of by patients undergoing forced feeding. This we know ' is
ihe direct result of the effect of the ingestion of food on the respiratory
exchange.

In determining, therefore, the optimal quantity of protein for tuber-
culous patients one must strike a balance somewhere between the two
extremes. While some patients can be made to gain nitrogen and
weight on diets containing from 37.5 to 45 gm. protein this requires a
large allowance of nonprotein food, especially of carbohydrate, in

27. Von Hoesslin, H. : Klinische Eigcntiimlichkeitcn iind Ernalirung bci
schwerer Inanition, Arch. {. Hyg. 88:147. 1918.

28. Bardswell, Noel, and Chapman. J. K. : In "Diets in Tubcrciildsis." they
give a complete summary of this work, Oxford Press, 1908.

McCAW— PROTEIN REQUIREMENT IN TUBERCULOSIS 57

amounts sufficient to bring the total calories up to from 3,500 to 4,000
per diem. When one compares the effect on the pulmonary ventilation
produced by the amount of protein, which might be fed at one meal in
an ordinary diet (from 30 to 40 gm.), with the effect of the amount of
carbohydrate to be fed at one meal with a forced diet, the effect of the
ordinary amount of protein becomes insignificant. Our results have
shown that nitrogen balance may be established at from 10 to 15 gm.
nitrogen (62 to 93 gm. protein), with a rather moderate number of

TABLE 9. — Summary of Nitrogen Minimum Experiments in Various Diseases

Author:

1.

1
1

i

h

s

p

H

H

i

s

5

^ ii

Subject:

-s

if

a

O

O

O

s>

5

i-

Disease

s

^

a"

Z

^

h

■fl

^

5

so

u

i

1

1

1

S

1

1

Psoriasis •

1,666

?

54.5

4.3»
4.29
5.16

1.88
2.99
2.47

1.97
1.93
1.89

3.85
4.92
4.36

+0.54
-0.63
-t-0.80

0.075
0.079
0.095

0.032
0.055
0.045

(18)

7

7

9 rv

■•

Kochcr

(1)

Paratyphoid

3,213 ] 56.0

■A.V.

14.52

O.HK

15.40

—13.2

0.038

0.230

T.. 38.1 C 16

i(»

4.666 ! 78.0

S.5

5.82

0.8S

6.70

-,S.2

16

Erysipelas

0.033

0.240

J. S., T.. 38 C. .. 10

4,280 75.0

53.8

1.88

E.M.,T.,39.6C. 6

3,360 73.0

4«.(l

iM

16.02

0 044

0.350

R.M.,T.. 39-400. 6

4,050 ....

61 .0

«,(ll

9.3S

»a

10.29

—8.28

0.033

0.166

R.M..T.,nornial ..

fi

4,230 ....

60.0

1.05

0.94

4.98

Lobar pneu-

Acutc polyarth-

ritis

E.S., T., 40C. ..

11

7

2,600

11.48

(».!«

12.40

—11.28

0.033

0.342

E. S.. T. normal

11

2,840

33.5

1.12

3.52

0.92

4.44

—3.32

0.033

0.105

Basedow's disease

16

"

2,480

60.7

3.13

3.19

1.15

4.34

—1.21

0.062

0.063

(19)

Syphilis, secon-

dary—nia 8

5

2,406 1 43.0 : 56.0

1.6

0.9

4.5

—2.9

0.029

0.064

(20)

cnit. Dis., Oct., Nov., 1913.
8.
des StickstolTwcchscIs in der Friihperiode der

calories in the diet (2,500). We have no evidence of any advantage
of ingesting an excess over 90 gm. protein. If a maximum of 90 gm.
per diem were given the stimulating effect of the protein taken at any
one meal would be negligible. It .should be possible to give most patients
at least 150 gm. fat, which would require only 150 gm. carbohydrate to
bring the total energy value up to 2,500 calories, .'^uch a diet could
he taken at the least expense of respiratory function.

58 ARCHIVES OF INTERNAL MEDICINE

SUMMARY AND CONCLUSIONS

1. In nine of the ten tuberculous patients studied the minimal
urinary nitrogen excretion observed was between 2.5 and 45 gm.
per diem, and between 0.041 and 0.093 gm. per kilogram per diem. In
one case the lowest excretion of urinary nitrogen was 9.4 ghi. per
diem, or 0.267 gm. per kilogram. In the nine cases with a minimal
nitrogen excretion the diet given had an energy value from 1.1 to 2.4
times the basal energy requirements of the subjects, and furnished 39
to 70 per cent, of the calories in the form of carbohydrate.

2. In some cases it is possible to maintain nitrogen balance, and
even to retain nitrogen, when from i7 to 44 gni. protein are
ingested, of which about one half is from animal sources. The
attainment of nitrogen equilibrium with such a small amount of protein
is dependent upon the ingestion of large amounts of carbohydrate and
fat, sufficient to make the total caloric value of the diet from 1.7 to 2.4
times the basal energy requirement.

It is probable that the failure to establish nitrogen balance on such
low protein diets is due to failure or inability of the subject to ingest
sufficiently large quantities of carbohydrate and fat, rather than that
it is due to an inherently large wear and tear quota in tuberculosis.

3. Positive nitrogen balance in bed patients may be attained by the
ingestion of from 60 to 90 gm. protein when the diet contains carbohy-
drate and fat, with a total caloric value of less than 1.7 times the basal
requirement. The evidence indicates that the optimal quantity of
protein for patients who are confined to bed with pulmonary tubercu-
losis, lies between the limits of 60 to 90 gm. per diem, when the caloric
value of the diet is about 2,500 calories. Additional carbohydrate and
fat calories must be furnished when patients are allowed to exercise.

OBSERVATIONS ON THE USE OF QUINIDIN IN
AURICULAR FIBRILLATION *

J. A. E. EYSTER, M.D., and G. E. FAHR. M.D.

A great deal of interest has been aroused by the recent reports on
the favorable action of quinidin in auricular fibrillation. Since Frey's '
first publication in 1918, the results from the treatment of over one
hundred cases have appeared, nearly half of which have been reported
by Frey.- The literature is given in the recent article by Levy,^ who
also reports the results on four cases. The purpose of this report is to
record the experience in the use of this drug in two cases selected from
a number under observation in this clinic. The first of these cases is
reported to bring out a point which in our opinion has not been empha-
sized sufficiently, namely, the danger in the use of this drug in certain
cases. The other case illustrates the strikingly favorable action of
quinidin in the rather rare condition in which auricular fibrillation is
uncomplicated by advanced valvular or myocardial damage and in
which no serious break in compensation has occurred.

ABSTRACT OF CASE REPORTS

C.\SE 1. — Mrs. P. F. S., aged 47, white, admitted to hospital, .\pril 5. 1921,
complaining of shortness of breath and palpitation. From the history, it was
evident that from moderate to severe cardiac decompensation with irregular
pulse had been present for the preceding seven months. Previous to this time
there had been no serious decompensation. The diagnosis was chronic endo-
carditis, mitral incompetence, cardiac decompensation, auricular fibrillation.
Examination revealed cyanosis, dyspnea, hydrothorax, ascites and edema of
the extremities. On admission the ventricular rate was 124; radial rate, 110.
Venous pressure was 23 cm. of water. Urine output in the first twenty-four
hours was 7 ounces. As a result of rest and treatment all symptoms improved,
the ventricular rate and pulse deficit decreased, the venous pressure fell and
urine output increased.

April 19, the apical rate was 80; radial, 76; venous pressure, 8 cm. Urine
output was 36 ounces. Cardiac area as determined by the teleroentgenogram
showed definite reduction over that present on admission. The condition con-
tinued to improve, and by May 7, ascites, hydrothorax and nearly all trace
of peripheral edema had disappeared. The patient was up in a chair for sev-
eral hours daily and was allowed moderate exercise. Auricular fibrillation
persisted, however, and numerous electrocardiograms taken during residence
in the hospital failed to reveal a normal cardiac rhythm at any time.

* From the Bradley Memorial Hospital, Department of Clinical Medicine of
the University of Wisconsin.

1. Frey, W. : Ueber Vorhofflimmern beim Menschen und seine Beseilungung
durch Chinidin, Berl. klin. Wchnschr. 55:417, 1918.

2. Frey, W. : Chinidin zur Bekampfung der absolulen Herrunregehnjissig-
keit. Deutsch. Arch. f. klin. Med. 136:70, 1921.

3. Levy, R. L. : Restoration of the Normal Cardiac Mechanism in .\uricular
I'ibrillation by Quinidin, Arch. Int. Med. 76:1289 (May 7) 1921.

60 ARCHIVES OF IXTERXAL MEDICI XE

May 17, qiiinidin sulphate was given in doses of 0.25 gm. every fourth
hour for four doses. After three doses (0.75 gm.) auricular tachycardia
developed with a rate of 180. After an hour or so, periods of auricular
fibrillation alternated with the periods of rapid regular heart action. Palpita-
tion, dyspnea and cyanosis developed. Venous pressure rose to 12 cm. Urine
fell to 26 ounces. During the night, a few hours after the last dose of quinidin,
heart action became more stable. On the following morning, quinidin admin-
istration was begun again with the same dosage as on the previous day.
Palpitation, cyanosis and dyspnea again appeared. The urine was 21 ounces,
the venous pressure 11 cm. Electrocardiograms showed periods of tachycardia
interspersed with periods of auricular fibrillation. The symptoms again sub-
sided with the cessation of quinidin administration the following night. The
following morning (May 19; quinidin administration with the same dosage
was begun again. A short time after the first dose, extreme distress developed.
There was dyspnea (respiratory rate 36), marked cyanosis, palpitation, pain
in side and cough with expectoration of blood streaked sputum. Examination
revealed right sided hydrothorax, pulmonary edema, ascites, tympanites, and
slight edema of the extremities. Venous pressure rose to 18 cm. Electro-
cardiograms revealed the same type of rhythm as on the two previous days.
The cardiac area was increased (teleroentgenogram). Due to the critical con-
dition of the patient and the absence of any favorable action on the cardiac
rhythm, quinidin administration was stopped after the third dose on this
day, and digitalization with general treatment for cardiac decompensation
established. A total of 2.75 gm. quinidin sulphate had been administered. On
the following two days the condition was critical. The venous pressure had
risen to 21 cm., the urine had fallen to 6 ounces in twenty-four hours. Con-
tinuous auricular fibrillation persisted with an average apical rate of 136 and
radial of 82. By the next day (May 22) the general condition had begun to
improve. Venous pressure had fallen to 16 cm. and urine increased to 16 ounces.
Stead}' improvement continued and June 3, the patient was discharged from the
hospital free from any symptoms or signs of decompensation. .Auricular fib-
rillation, however, was still present, with an apical rate of 72 and with a very
small pulse deficit. On discharge the venous pressure averaged 7 cm. and the
urine was approximately equal to the fluid intake.

Case 2. — L. J. R., white, male, university student, aged 23. came under
observation for the first time Aug. 7, 1919. complaining of shortness of breath,
palpitation and irregular heart action. This condition developed on the tliird
day of a rather severe influenzal attack in November, 1918. No heart irregu-
larity or palpitation was noted previous to this time. Following recovery
from the attack of influenza, the cardiac irregularity persisted. Shortness of
breath was present on very slight exertion and at times paroxysms of dyspnea
with severe cardiac palpitation occurred, even when the patient was at rest.
Examination revealed marked irregularity of the heart with considerable pulse
deficiency. There was, however, no evidence of venous stasis in lungs, liver
or other organs, and no adventitious heart sounds were heard. The cardiac
area (roentgenogram) was increased approximately 25 per cent. The enlarge-
ment was symmetrical. Electrocardiograms revealed auricular fibrillation. The
past medical history was unimportant and furnished no evidence of cardiac
involvement previous to the attack of influenza.

Under restricted activity and dietetic control, the condition improved some-
what. The patient was again examined Jan. 29. 1920. General health liad
been fair for the past few months, and he had been able to pursue his uni-
versity work on a restricted schedule. There had been no symptoms of severe
cardiac incompetence. Moderate exercise could be taken without extreme
shortness of breath. Examination revealed a cardiac area somewhat smaller
than previously observed (approximately 20 per cent, above normal), totally
irregular cardiac rhythm, no adventitious sounds and no evidence of venous
stasis.

EY ST ER-FAHR— AURICULAR FIBRJLLATIOX 61

The patient entered the hospital June 28, 1921, for treatment with quinidin.
The irregular heart action had persisted without apparent interruption from
its initiation thirty-one months previously. For the past year he has had to
give up practically all work, due to frequent periods of shortness of breath
and general weakness. Rather severe attacks of palpitation and dyspnea, last-
ing from fifteen to twenty minutes, have occurred every few days and have
increased in frequency recently. There has been, however, no serious cardiac
incompetence at any time. For the past eighteen months he has had abdominal
symptoms which were stated by the physician in attendance to be due to
spastic colitis.

Examination revealed practically the same condition as found at the two
previous examinations, thirty-one and eighteen months ago. The cardiac area
was found to be less than IS per cent, above normal. No adventitious heart
sounds were heard. The patient was kept at rest in bed for five days without
medication. Repeated electrocardiograms were made during this time, all
showing uninterrupted auricular fibrillation. The apical rate averaged 115,
the radial 94. Venous pressure varied between 8 and 9 cm. of water in the
dorsal prone position. Urine output was normal.

Jjily 4, the patient was given a single dose of 0.25 gm. quinidin sulphate by
mouth. About two hours later an electrocardiogram revealed periods of tachy-
cardia interspersed with periods of fibrillation. On the next day (July 5),
he was given 4 doses of 0.25 gm. quinidin at three hour intervals. There
developed brief periods of rapid regular heart action with moderate distress
(palpitation and shortness of breath). Xo other signs of circulatory deficiency
were present. The apical rate averaged 120. the radial 110. Venous pressure
was unchanged. Urine output was normal.

July 6. six doses of 025 gm. quinidin at two hour intervals were ordered.
Before the first dose, rather frequent series of rapid regular beats were present,
interspersed with periods of fibrillation. At 4 p. m., following the fourth dose
of quinidin. the patient became uncomfortable, due to palpitation and short-
ness of breath. The pulse became exceedingly irregular in force and rhythm.
Electrocardiograms made at this time showed continuous fibrillation with a
ventricular rate of 120. The subjective distress increased somewhat until
about 6 p. m., when the patient fell asleep. On awakening in about half an
hour, all distress had disappeared and cardiac action was found to be entirely
regular in force and rhythm. Electrocardiograms revealed a normal sino-
auricular rhythm with a rate of 66, with normal P-R interval and without
evidence of preponderance of either ventricle.

The patient was discharged from the hospital July 14. Subsequent to its
initiation, eight days previously, normal heart action had persisted without
interruption. For the first two days following the return of normal rhythm,
a single dose of 0.25 gm. quinidin sulphate was given daily, subsequently a
single dose every other day, to be continued after leaving the hospital. All
symptoms disappeared and examination failed to reveal any abnormality with
the exception of slight symmetrical cardiac hypertrophy (15 per cent.). The
pulse rate averaged approximately 60 during this period. The venous pres-
sure (dorsal prone position) fell to 5 cm. shortly after the establishment of
normal cardiac rhythm and remained at this point.

COMMENT
No one who ha.s carefully followed the actiim of (|uiiiidin on the
heart in auricular fibrillation can be but impressed with the i)0werful
action of this drug on the cardiac mechanism in this condition. Clinical
experience has shown that its action is sufficient to restore the normal
mechanism at least temporarily, in aj^proximately half of the cases of
auricular fibrillation in man. Much of this experience is too recent.

62 ARCHU'ES OF JXTERXAL MEDICIXE

however, to state in how large a percentage permanent success can be
obtained. Since in most cases conditions tending to produce auricular
tibrillation are probably still present even after the normal rhythm is
restored, subsequent administration of the drug may be necessary to
avoid recurrence. Additional experience will be necessary before the
details of this will be clear. In most cases, perhaps in all, disturbances
of rhythm occur during the transition stage between auricular fibrilla-
tion and sino-auricular rhythm. The most characteristic and frequent
of these transition rhythms is rapid regular heart action (auricular
tachycardia, "auricular flutter") occurring either alone or in periods
interspersed with periods of fibrillation. These intermediary stages may
occur even when the normal rhythm is not subsequently restored, as
in the first case presented here. It is apparently the result of these
stages of transition in which the dangers of the treatment lie. While
acutely developing auricular fibrillation undoubtedly causes consider-
able mechanical deficiency of the heart (Eyster and Swarthout *) and
is probably not infrequently the immediate cause of cardiac decompen-
sation, the heart may compensate for this as it does for valve injury,
particularly when it is assisted by the protective influence on ventricular
stimulation of digitalization. That the removal of this compensated
auricular fibrillation under the action of quinidin in producing transi-
tion rhythm may destroy clinical cardiac compensation, is illustrated by
the first case reported. Possibly also the contractility of the ventricular
muscle is reduced by the drug. The case again becomes critically ill,
and if restoration of the normal sino-auricular rhythm fails, as it
apparently so frequently does in the older and more severe forms of
chronic heart disease, the best that can be hoped for is a tedious
restoration of compensation with another period of cardiac failure
with its attendant permanent damage to be charged to the quinidin
treatment. On the other hand, when auricular fibrillation is unassociated
with valvular or severe myocardial damage and with no history of
severe circulatory failure, the cardiac reserve is able to carry the
circulation through the periods of "transition rhythm" with only
transitory circulatory deficiency.

With a therapeutic agent as definite in its action as quinidin of use in
a clinical condition as common and important as auricular fibrillation,
it is inevitable that great interest in and widespread use of the remedy
will follow. We have attempted to point out a definite danger in its
use and the necessity of the utmost care in its administration to at least
the more severe forms of chronic heart disease. In such cases it should
be used only, first (as has been pointed out by Frey) after compensa-

4. Eyster, J. A. E., and Swarthout, E. : Experimental Determination of
the Influence of Abnormal Cardiac Rhythms on the Mechanical Efficiency of
the Heart, Arch. Int. Med. 25:317 (March) 1920.

EVSTER-FAIIR— AURICULAR FIBRILLATIOX 63

tion is as thoroughly established as possible ; and second, only when the
patient can be under almost continuous observation, when frequent
electrocardiograms are obtained, and when repeated physical examina-
tions are made for signs of breaking compensation. For the present,
and until more experience is gained as to the mode of action, contra-
indications, the size of dose and the frequency of administration for
the best results, the treatment of the more severe cases should be car-
ried out only in a hospital under strict regulation and observation. The
patient should be prepared for a break in compensation by rest in bed
and limitation of food and liquids during the period of administration.

TUBERCULOSIS OF THE HEART

WITH THE REPORT OF TWO CASES *
EDWARD WEISS, M.D.

PHILADELPHIA

III a Study of 7,219 necropsies Norris ^ found 1,780 tuberculous
cases and among these eighty-two cases of tuberculous pericarditis.
In five cases the heart muscle was involved. This indicates the relative
infrequency of the conditions being reported. A point of added interest
is the unusual degree of involvement in the first case.

REPORT OF CASES

Case 1. — History.— The patient, a colored male, aged 25, was admitted to
the Jefferson Hospital, on the service of Dr. H. A. Hare, Dec. 27, 1920, com-
plaining of pain in the upper half of the abdomen and chest. His family his-
tory was negative for tuberculosis. He had gonorrhea in 1909 and syphilis
in 1913, for which he received no treatment. In October, 1918, he had an
attack of influenza and since then a persistent cough at night. In July, 1920,
he developed abdominal pain which grew progressively worse and about Decem-
ber, 1920, his abdomen began to enlarge. His cough became more severe ; he
was dyspneic and complained of pain in his chest. He stated that he had lost
40 pounds in the past year.

Examination. — On examination there were evidences of a pleural effusion at
the right base and immense enlargement of the heart — both confirmed by
roentgen ray. His abdomen was slightly distended and tender and the liver
could be palpated at the level of the umbilicus in the right midclavicular line.
There was some fluid in the left tunica vaginalis. Blood count showed a leu-
kopenia and moderate secondary anemia ; his blood Wassermann was posi-
tive; his sputum was negative for tubercle bacilli. He died Feb. 21, 1921,
and necropsy was performed the same day.

\ccropsy Report. — The body was that of an adult colored male, weighing
about 170 pounds. The heart and pericardium were immensely enlarged, weigh-
ing 1,580 gm., the pericardial cavity being obliterated by large, dense, yel-
lowish nodules which invaded the heart muscle, auricles and ventricles to a
similar degree. The nodules were continuous with the mediastinal and peri-
bronchial lymph nodes which were also large, yellowish and firm. No tuber-
culosis of the lungs could be established but the pleurae averaged about 0.4 cm.
in thickness and this thickening was especially marked at the right base which
contained an encapsulated effusion of about 700 c.c. of brownish-red serum.
The spleen, right suprarenal, kidneys and liver showed a number of firm, yel-
lowish nodules scattered throughout. At the juncture of the ileum and cecum
was a large ulcer, 2.S cm. in diameter, with base of reddish granulations and
firm, undermined edges. This, with the large, firm and yellowish mesenteric
and retroperitoneal nodes aided in the gross diagnosis of tuberculosis.

The microscopic study revealed a fibrocaseous tuberculosis of the follow-
ing organs and tissues : pericardium and myocardium, pleurae, spleen, right
adrenal, kidneys, liver, mediastinal, mesenteric and retroperitoneal lymph nodes.

* Read before the Pathological Section of the National Tuberculosis .Asso-
ciation. New York City, June 16, 1921.

* From the Department of Pathology. Jefferson Medical College.

1. Norris, C;. W. : Tuberculous Pericarditis Hased on a Study of 7,219
Autopsies in Philadelphia Hospitals, Univ. Penn. M. Bull. 17:155. 1904.

JVEISS— HEART TUBERCULOSIS

65

No typical tubercles were seen and very few giant cells. Smears from the
fresh material and stained sections of pericardium, pleura and retroperitoneal
nodes were studied for tubercle bacilli but failed to reveal the organism.
Guinea-pig inoculation, however, produced a diffuse tuberculous lymphadenitis
and tuberculosis of the spleen. Smears from the caseous nodes showed many
tubercle bacilli but attempts to culture the organism failed.

C.^SE 2. — S. W., a colored male, aged 25, dishwasher by occupation, was
admitted to the Department for Diseases of the Chest of the Jefferson Hospital,
March 15, 1921." He complained of a dull pain in the back and sternal region
and of a moderate cough productive of a large amount of mucoid sputum which
was occasionally blood-streaked.

Family History. — Xegativo for tuberculosis.

Fig. 1. — Case 1. Tuberculous pericarditis and myocarditis. Note thi
invasion of heart muscle by the immense, tuberculous nodules of the peri
cardium, shown best by insert in left hand corner.

Personal History. — He had a chancre in 1913 and gonorrhea in 1914. lie
stated that he had been in poor health ever since his discharge from tlic army
in 1918. At that time he noticed a cough. Shortly after the onset he became
short of breath on slight exertion and felt weak. From time to time he had
a vague pain in the front of the chest. Increasing disability caused him to
give up his work in the latter part of January, 1921. He entered a Phila-
delphia hospital and was later transferred to the Jefferson Chest Hospital.

am indebted to Dr. F.. H. Funk for the clinical notes of this case.

66

ARCHIVES OF IXTERNAL MEDICINE

Physical Examination.— This showed an emaciated, colored, adult male. His
pupils reacted normally. Throat was red. tonsils swollen. The cervical lymph
nodes were distinctly palpable with a small mass of enlarged glands above
the clavicle on the "left side. Chest was long, narrow and flat. Expansion
was generally limited. Vocal fremitus was increased throughout the right side
of the chest and the percussion note was generally impaired. Breath sounds
were distinctly audible with a blowing characteristic over the entire chest except
the right base posteriorly where the intensity was diminished. Many squeak-
ing sounds were heard with numerous crackling rales throughout, especially

Fig. 2. — Case 2. Tuberculous pericarditis, showing (a) enlarged, tuber-
culous nodes at the base of the heart; (b) thickened, parietal pericardium, and
(c) tuberculous nodules of visceral pericardium.

on the right side. Whispering pectoriloquy was noted in the third interspace
on the right side, anteriorly.

Heart : There was marked precordial pulsation with an increase to the
left in the lateral diameter of cardiac dulness. Heart action was rapid but
regular ; heart sounds were distinctly heard ; no murmurs were present. Sputum
examinations were repeatedly negative for tubercle bacilli. Urine contained a
trace of albumin and an occasional hyaline cast. Temperature, 102 F. ; pulse.
130; respirations, 40. Patient died, April 25, 1921, and a necropsy was made
the following day.

WEISS— HEART TUBERCULOSIS

67

Xccrofsy Rcfiort.— The important findings were as follows: The anterior
cervical lymph nodes were distinctly palpable and those on the left side just
above the clavicle were especially enlarged. There were several enlarged, firm,
yellowish nodes adherent to the posterior surface of the sternum. AW of the
mediastinal nodes were similarly enlarged, clustered about and pressing on the
trachea and larger bronchi. These nodes varied in size from less than 1 cm.
to 3 or 4 cm. in diameter. Some had caseous centers.

The pericardial cavity was distended with about 700 c.c. of blood-tinged
serous fluid. There was likewise a dense, yellowish, fibrinous exudate present.

ulosis of peribronchial lymph nodes with practically

The visceral pericardium of the left ventricle was distinctly thickened, measur-
ing about 0.3 cm., and consisted of a zone of dense, yellowish tissue made up
of conglomerate nodules which grew larger near the base of the ventricle at
the reflection of the pericardium. These nodules, unlike those of the first case,
showed no tendency to invade the heart muscle. The lymph nodes at the base
of the heart were of the same type as the mediastinal nodes and were con-
tinuous with thom, pressing on the great vessels. These nodes were likewise
firmly adherent to the pericardium.

68 ARCHH'ES OF INTERNAL MEDICINE

Left Lung: The left pleural cavity contained about 500 c.c. blood-tinged
serous fluid and the lung was compressed. On section the pulmonary tissue
was dark red in color and only partially crepitant. About 6 cm. from the
apex, on the anterolateral surface of the lung, was a conglomerate caseous
tubercle, a little less than a centimeter in diameter. There was no further
evidence of tuberculosis in the lung structure; the nodes at the hilus, how-
ever, were greatly enlarged, similar to and continuous with the mediastinal
nodes.

Right Lung : It was more voluminous than the left but on section presented
nearly the same appearance. Near the hilus. a little area of pulmonary tis-
sue, about 2 cm. in diameter, adjacent to an enlarged and caseous lymph node,
showed tuberculous infiltration.

None of the other organs showed any evidence of tuberculosis but the
mesenteric and retroperitoneal lymph nodes were moderately enlarged, firm
and yellowish.

A few tubercle bacilli were found in smears from the caseous medi-
astinal nodes.

Microscopic study revealed a typical caseous tuberculosis of the structures
mentioned. The pericardium showed the lesion especially well ; unlike the
first case, there were many typical tubercles present with caseous centers and
numerous giant cells. The tuberculous process, as mentioned in the gross
description, exhibited no tendency to invade the myocardium.

Guinea-pigs were inoculated with emulsified caseous nodules from the base
of the heart. The animals, killed six weeks later, showed tuberculosis of
the abdominal lymph nodes and spleen. A number of tubercle bacilli were
recovered in smears but cultures both on Petroff's and Dorset's mediums failed
to develop the organism.

COMMENT

In 1902 Anders ' summarized seventy-one cases of tuberculosis of
the myocardium and added a case of his own. He classified the condi-
tion as occurring in three forms: (1) large tubercles; (2) miliary
tubercles (less common) and (3) diffuse form or tuberculous infiltra-
titon (rare). He stated that cardiac tuberculosis was most frequent
in early life, 40 per cent, of bis cases occurring under \5 vears. In
a large proportion of cases, the lungs and bronchial glands were
tuberculous. In twenty-nine out of thirty-one cases in which the
glandular condition was noted, tuberculosis was present and this he
lich'eved tf) be the seat of the disease. This extended from the bronchial
tn the nu'diaslinnl nodes and thence directly to the liearl or (|uite
iiimnionly liy way of the pericarcHum.

Cases very similar t(i the ])resent ones have been repcirtcd by F.llis,*
Raviart and Caudron,' Bcnda and Geissler," Toldt,' {'assamonti,*

3. Anders, T. M.: Tuberculosis of the Myocardium, 1. .\. M. A. 39:1081
(Nov. 1) 1902.

4. Elli."!. A. G.: Heart Showing Chronic Tulierculosis ..l the IVricardium.
with Involvement of the Myocardium, Proc. Path. Soc, Phila.. June. 1903.

5. Raviart and Caudron : A Case of Tuberculosis of the Myocardium, Echo
med. du Nord 8:529, 1904.

6. Bcnda, C, and Geissler : New Cases of Tuberculosis of the Heart and
Blood Vessels. Deutsch. med. Wchnschr. 31:1169 (July 20) 1905.

7. Toldt, G. : A Case of Tuberculosis of the Myocardium, Rev de med.
26:101, 1906.

8. Passamonti. M. : A Case of Tuberculosis of the Myocardium. I'olicliiiic.i.
Rome 14:88, 1907.

WEISS-HEART TUBERCULOSIS 69

Beifeld,'-' Fraga,'" and more recently by .\damson," Doernier,*- and
Binder.'^

Ellis' case presented, in addition to tuberculous pericarditis and
myocarditis, tuberculosis of the mediastinal nodes and lungs and a
miliary tuberculosis of liver, spleen, pancreas and kidneys. Only after
a study of many blocks from the heart were a few tubercle bacilli
seen in one section. Beif eld's case showed a tuberculous scar at the
apex of the left lung and a caseous tuberculosis of the mediastinal,
tracheal, bronchial and left lower cervical nodes. There was a tuber-
culous pericarditis with total obliteration of the pericardial sac. The
liver and spleen contained conglomerate tubercles. Microscopic exam-
ination demonstrated typical tuberculous tissue but up to the time of
his report search for tubercle bacilli in the tissues was unsuccessful.
He concluded that infection traveled from the left apex to the media-
stinal nodes and then by contiguity or lymphatic extension or both, to the
pericardium and by contact infection to the heart muscle. In Adamson's
case no evidence of tuberculosis was found in the lungs or bronchial
lymph nodes but he felt that the process probably started in the nodes
at the base of the heart.

It is generally conceded that the mediastinal nodes act as the focus
of infection for the pericardium and myocardium in this condition. Toldt
however, reports a case of tuberculosis of the heart (limited to the
right auricle) with no sign of pericardial involvement. He found but
little evidence of tuberculosis disease in the tracheobronchial nodes
but did note a healed lesion at the apices of both lungs. In commenting
on the condition he mentions that Fuchs in a report of fifty-three cases
of tuberculosis of the myocardium, found twelve cases of isolated
auricular tuberculosis. Of these the right auricle alone was aflfected
in nine; both auricles in two; and the left auricle only in one. Of the
twelve cases eight showed tuberculous pericarditis, in two the peri-
cardium was unafTected and in two the condition was not stated.

Passamonti also reports a ca.se of tuberculosis of the myocardium
with no other tuberculous lesion except a nodule in the lower lobe
of the left lung. He did not feel that the usual propagation of infection
(glandular) had occurred in his case. In a discussion of llic condition

9. Beifeld, A. F. : Tuberculosis of the Myocardium, with the Report of a
Case, Tr. Chicago Path. Soc. 8:104. 1909-1912.

10 Fraga, C. : Concerning a Case of Cardio-Tuberculous Cirrhosis, Brazil-
med. 31:398. 1917.

11. Adamson, W. W.: A Case of Tuberculosis of the Myocardium. J. Path.
& Bacteriol. 23:.?99 (Dec.) 1920.

12. Doermer, VV. : A Case of Conglomerate Tuberculosis of the heart. Diss.,
Jena. 1918.

13. Binder. A.: Tumor-Like Tuberculosis of the Heart, Zentralbl. f. inn.
Med. 41:462, 1920.

;n .IRCHIJ-ES OF IXTERXAL MEDICIXE

he mentions that the finding is practically always a postmortem one,
and from a study of the literature brings out the fact that there are
no characteristic signs of tuberculosis of the heart. He states that
attacks of dyspnea, cyanosis, arrhythmia and enlargement of the cardiac
area have been observed but adds that hypertrophy and dilatation
are not always noted. Raviart and Caudron record a case in a white
male, aged 31, with vague symptoms of oppression and pain in the
epigastrium, and no definite physical signs. Necropsy revealed a
tracheobronchial glandular tuberculosis and tuberculosis of the heart
(confined to the right auricle). Toldt's case, mentioned above, occurred
in a woman of 58 who had been ill for eighteen months, but, with
"no symptoms of localization." Examination showed deviation of
the apex of the heart, arrhythmia and muffled apical meurmurs — no
other physical signs.

Anders made note of the following symptoms : palpitation, feeble
heart sounds, pericardial distress, diffuse pulsation, tumultuous and
rapid heart action, fetal and gallop rhythym, rarely murmurs, sudden
and recurring syncope, dyspnea, cyanosis, unconsciousness, general
edema and sudden death. But he states, "in all this list there is nothing
specific, no single symptom or combination of symptoms that is not
seen in functional and organic disease of the heart other than tuber-
culosis. Often the patient dies without a symptom that would attract
special attention to the heart.'" Eisenmeyer is quoted by Anders as
stating that a diagnosis may occasionally be made by the presence in a
victim of general tuberculosis of sudden, severe collapse, quickly
passing; and the detection of weak endocardial murmurs varying in
phase and intensity.

One point not mentioned in the cases cited, is the apparently,
relatively frequent occurrence of this unusual form of tuberculosis in
the colored race. Four cases, Ellis', Beifeld's and the two making up
the present report, occurred in young, adult, colored males. For this
reason it seems reasonable to suggest that when evidence of cervical
and mediastinal glandular disease e.xists in a young colored man with
indefinite symptoms referable to the circulatory system ami |i(issil)ly
signs of cardiac enlargement — tuberculosis of the heart .should certainly
be considered. In this connection Biefeld's case merits particular
mention. A colored male, aged 21. had dyspnea and cyanosis; cough;
rapid, regular pulse of good quality; and an evening rise of temperature.
Examination showed a heart enlarged to right and left. No murmurs
were present but a systolic retraction was noted, in ])lace of the ajiex
beat. .\ clinical diagnosis was made of adherent ])ericar(liuni, tuber-
culous in n,-ilure; the necropsy, as .nhove described, confinned this
diagnosis.

UEISS— HEART TUBERCULOSIS 71

SUMMARY

The brief clinical reports and necropsy findings of two colored, yoiuig,
adult males are recorded. Symptoms and physical signs were vague;
necropsy demonstrated an immense, fibrocaseous tuberculous involve-
ment of the lymph nodes of the thorax, with extension to the peri-
cardium and in the one case to the heart muscle, causing tremendous
enlargement of the organ. Reference to the literature indicates that
the heart involvement is almost always secondary to the disease of the
mediastinal lymph nodes. The practical limitation of the process to
the lymph nodes, the curious reaction of the tissues and the difficulties
in the diagnosis of such conditions are some of the problems offered to
pathologist and clinician by these case reports.

I am indebted to Dr. H. .\. Hare and Dr. Thomas McCrae for permission
to report these cases.

OBSERVATIONS FOLLOWING INTRAVENOUS INJEC-
TIONS OF HYPERTONIC SALT SOLUTIONS
IN CASES OF NEUROSYPHILIS*

JAMES WYNN, M.D.

BOSTON
INTRODUCTORY

^^'eed and McKibben ' and others ^ have shown that the adminis-
tration of hypertonic salt solutions in the cat causes a marked and
prolonged fall in cerebrospinal fluid pressure. By an ingenious method
they were further able to show that with this fall in pressure a consid-
erable amount of subarachnoid fluid was dislocated into the nervous
system ; that the fluid "passed along the perivasculars into the substance
of the nervous system, reaching the interfibrous spaces in the white
matter and the pericellular spaces in the gray." The method leading
to these conclusions consisted in allowing a few cubic centimeters of
iron-ammonium citrate and sodium ferrocyanid to run into the sub-
arachnoid space as the cerebrospinal fluid pressure (after intravenous
injection of hypertonic salt solution) reached zero or was rapidly
falling; then fixation of the central nervous system in liquor formal-
dehyd acidified with 5 per cent, hydrochloric acid precipitated Prussian
blue (readily demonstrable microscopically) at points to which the
injected citrate and ferrocyanid had penetrated. Foley and Putnam ^
later showed that similar falls in cerebrospinal fluid pressure (pre-
sumably with the same dislocation of fluid) could be obtained in cats
by administering approximately similar doses of salt per duodenum or
per rectum.

In view of these observations, Foley raised the question as to
whether or not intraspinal injections of arsphenamized serum in man
might be more efifectively distributed if followed by the administration
of salt either by mouth, rectum or vein. In patients there is no way
of determining how extensive the distribution of intraspinal injections
actually is. Effective distribution can only be inferred from favorable
clinical and serologic results. In the cat, when cerebrospinal fluid
pressure (after intravenous injection of salt solution) is reduced to
zero or to a negative figure, an intraspinally injected substance can be

♦From the Medical Service of the Peter Bent Brisham Hospital.

1. Weed and McKibben: Pressure Changes in the Cerebrospinal Fluid
Following Intravenous Injections of Solutions of Various Concentrations,
Am. J. Physiol. 48:512 (May) 1919.

2. Foley and Putnam : F.flfect of Salt Ingestion on Cerebrospinal Fluid
Pressure and Brain Volume, .Xm. J. Physiol. 53:464 (Oct.) 1920.

U-yXX—XEUROSYPHILlS 73

shown to be dislocated into the substance of the nervous system.
Consequently, in man it would seem reasonable to suppose that the
fluid content of the subarachnoid space passes similarly into the brain
substance after intravenous salt injections, if cerebrospinal fluid
pressure falls in man as in cats. Such a displacement of injected serum
would seem especially to be desired in cases of neurosyphilis with few
or no posterior root symptoms, i.e. those cases ^ with cerebrospinal
fluids showing pleocytosis, increased globulin and strongly positive
Wassermann reaction, but with purely subjective evidences of distur-
bance and with reflexes for the most part intact.

The first problems were to determine the optimum salt dosage and
means of administration, and to ascertain whether intravenous injection
of hypertonic salt solutions cause a fall in cerebrospinal fluid pressure
in man as in the cat. Foley and Morris, attempting alimentary
administration of salt in several preliminary observations, were soon
convinced of its impracticability; almost without exception doses of
from 15 to 30 gm. (in capsules) were promptly rejected, whether
given orally or rectally. Intravenous solutions of varying hypertonicity
were later tried by me and 15 per cent, solutions were finally adopted
as the optimum. It was soon found that patients could tolerate in
this concentration with no very alarming symptoms 410 mg. salt per
kg. of body weight. Slight variations in dosage above and below this
figure produced such negligible change in efTect that it was eventually
decided to use as a routine salt dosage 200 c. c. of a 15 per cent, solution.

With the assistance of several cooperative patients it was possible
to get actual graphic records in man of the changes in cerebrospinal
fluid pressure during and for about thirty minutes after the intravenous
injection of the routine salt solution. The procedure was as follows:
Lumbar puncture was performed with a small bore needle (patient
in the usual position on the right side) and a glass capillary manometer
containing normal salt solution was attached, with the loss of as little
cerebrospinal fluid as possible, usually only a drop or two. Five or ten
minutes were allowed to elapse until a reliable normal could be read;
no readings were taken unless cardiac and respiratory fluctuations in
pressure evidenced the technical integrity of the apparatus. The
intravenous salt solution was then given through a No. 19 needle by
gravity, the patient having before lumbar puncture had his arm pre-
pared, thus avoiding even the slightest change in position and consequent
disturbance of the manometer during venipuncture."

3. Only nine of the sixteen cases here reviewed are of this group; the
other seven were cases of tabes, the only other cases available for the salt
treatment at the time.

4. In three of these determinations Dr. F. E. B. Foley rendered, through
his suggestions and cooperation, most valuable assistance.

74 ARCHIVES OF IXTERXAL MEDICIXE

In six cases studied thus there was a constant rise in pressure of
from 30 to 50 mm. during the fifteen minutes occupied in giving the
salt solution. A moment or two before the end of the injection the
pressure would start down, and by thirty minutes after the end of the
injection it would usually sink from 80 to 100 mm. below the original
level. By this time the patients were invariably so uncomfortable
that the lumbar puncture needle had to be withdrawn, which, of course,
made further pressure observations impossible. However, the close
correspondence in these initial pressure variations with those in the
cat would make it reasonable to suppose that the depression increases
till zero and possibly even negative levels are reached, as in the case
of the cat.

Fig. 1. — Vertical figures: pressure in mm. of cerebrospinal fluid. Horizontal
figures: time. Dotted line: period during which 200 ex. of l.S per cent, salt
solution was given intravenously.

Normal jjressures in the different cases often varied from 30 to 40
mm., but the relative changes after injection of salt solution have heni
remarkably constant. Figure 1 represents the course in one case ([uite
typical of the group.

In view of these demonstrated pressure changes in man,'' fluid
displacement into the substance of the nervous system seemed an
hypothesis reasonable enough to warrant the systematic use of salt

5. Three of these pressure dctcrtuinations were carried out with Pr. F. K, B.
Foley, and since charts of these will he puhlislied by liim elsewhere, only one
curve is reproduced here,

lirXX—XE UROSYPIIIUS 75

solutions in chosen cases of neurosyphilis, in an effort to render more
effective arsphenamized serum distribution through the subarachnoid
space.

PROPOSED CLINICAL PROCEDURE

It was decided that sixteen patients with neurosyphilis receiving
intraspinal treatment ° should receive 200 c. c. of 15 per cent, salt
solution intravenously during the hour following intraspinal treatment,
this to be repeated with each succeeding serum treatment, given at
fortnightly intervals, till six salt injections had been given. Serologic
and cytologic changes in the cerebrospinal fluid thus could be followed
closely ; with the return of the patients two months after the last
treatment for follow-up examination and lumbar puncture, it would
be fwssible to note any variation from the usual serological and clinical
course of the- cases.

.\CTUAL CLINICAL COURSE OF THE P.\T1ENTS GIVEN SALT
SOLUTION

The actual procedure with the cases varied considerably from that
planned, as will be apparent from a glance at Table 1, in which the
results are tabulated. Only one patient in sixteen received all six
proposed salt injections, three received five, ftve four, and the rest
less. The reason for this deviation from schedule is apparent from
Table 1. Five patients failed to return for completion of the course
and were lost from observation. Six absolutely objected to completing
the salt injections because of extreme distress produced by them on
previous occasions. In five cases the Wasserniann reaction and cell
count became negative before the giving of all six injections of salt
solution.

The column headed "subjective effect of salt," Table 1, shows that
almost every patient experienced more or less marked discomfort from
the hypertonic solutions, in six cases this discomfort being severe
enough to necessitate abandoning subsequent injections. The symptoms
associated with giving salt were surprisingly constant in all the cases ;
shortly after the start, marked facial flushing and a sensation of heat,
involving first the face and then in succession neck, arms, trunk, legs

6. In this clinic the indications for intraspinal treatment are essentially
those set forth by Fordyce' and others." t a >t v

7 Fordvce: The Treatment of Syphilis of the Nervous System, J. A. M. .\.
63:552 (.\'ug. 15) 1914; Intraspinal Therapy in Neurosyphilis, .^m. J. Syph.
3:337 (July) 1919. _^ . , ^ ^ ,

8 Swift and Ellis: The Treatment of Syphilitic Affections of the Central
Nervous System with Especial Reference to the Use of Intraspmous Injections,
\rch Int. Med. 12:331 (Sept.) 1913: A Study of the Spirochaeticidal Action
of the Serum of Patients Treated with Salvarsan, J. Exper. M. 18:435 (Oct.)
1913.

¥

Ifi

— —

' " " !2 » « - ' -

-

=1
ill

Is

- 1 -

III

iiiil il 1 1 1

1

11

^ O

1

B.

1

1

1

1

i

1

00

a -^

o

|2coj;>goga.g«i^-=g«g

S

r

2

1
1

1

+ 1 + 1 1 1 1 + I+ + + +I
^+'+1 1 ' I+I++++

+ + + + + + + + 1 + + + +

=;+ + -t + + + + + ' + + + +
«3 2 § 3 S S 2 2 2 S d 2 S

+

2

+
+

d

iS 2

+ +
2 2

z

o

H

o

+ + + + + + + +1 1 + + + +
+ + + + + + + + + + + + +

=3+ + + + + + + + + + + + +

^2§S§23§2§S222

+

2
2

+ +
2 2
+ +
i S

d d

i

s

1 i ilil»i 1 i ! 1 f 1 1

1 S "'l^g^'w '^ 1 1 1 1 "^ ^

1
1

1 1
< <

^0«^C,., g3-.C.CO^,=

,

^ »

^

— — — •= — — "

,

-«• ■»

si
11

3o

III!

s

a

<!<!<!<a,<!<!<<<;<!<a,<!

<!

-« <

£

(l,(l,^ft,fl<C4(l4|l<fl,(l,(l,(lH<!

«!

■< <

^

n,!i<(i,a,a,!i,^a,cn-<:<!<<

<

< ■<

Ui

3,^(i,fi,(i<pMa<PL,(i<<<;-<!<!

<!

< <

OS

<!<i;<!<!<!<!<!<!<a<<«<l(l<

<

h 1^

III

o-|-2"- = "|"|g

S

a 8

1

lliiillll j ; i ;

IllllllliJiii

5

i i

5 5

JVYXX—XEUROSyPHIUS 77

and feet ; then intense dryness of the throat and at the end of the
injection intense "boring" occipital or frontal headache, usually of
only from ten to twenty minutes' duration, but relieved at the time
by nothing except morphin.'-*

Comparison of cell count and Wassermann reaction improvement
(Table 1) in patients receiving (a) intraspinal treatments alone and
(b) intraspinal treatments plus salt solution intravenously would seem
to indicate that the salt is of no therapeutic value."

Furthermore, clinical evidences suggested that such use of hypertonic
salt solutions might be positively harmful, for almost without exception
there were aggravation of root pain and, in many instances, the
occurrence of alimentary upsets persisting for four or five days during
the intervals between treatments.

In view of these facts, it would seem that despite a sound theoret-
ical basis, the use of hypertonic salt solution intravenously has no place
in augmenting the intraspinal treatment of neurosyphilis. It may
validly be objected that these observations are not conclusive, extending,
as they do, rarely over nine months. However, it did not seem justifiable
to push the work further in the face of no marked serological and
cytological improvement and actual symptomatic retrogression.

EFFECT CF S.\LT INJECTIONS ON BLOOD PRESSURE, BLOOD AND
URINE CHLORID

In Table 3 are tabulated the effects of intravenous injection of
salt solutions on blood pressure, whole blood chlorid, and urine chlorid.
The relation of these changes in a single case is shown in Figure 2.
From Table 3 it is apparent that in this group immediately following the
routine salt injections, there was from 51 to 61 per cent, increase in the
whole blood chlorid, that an hour later the increase was from 20 to
35 per cent. Twelve hours after injection there was still from 14 to
22 per cent, elevation in blood chlorid ; at that time approximately half
of the injected salt had been excreted in the urine. In three cases (Nos.
5, 6 and 8, Table 3), because of previous disagreeable reaction to the
routine dose of salt, smaller injections were given, with correspondingly
less marked and prolonged elevation in blood chlorid. In one of these
cases practically all of the injected .salt had been excreted in the urine
in twenty-two hours. In the other two cases excretion was not quite
so rapid. In view of the marked and prolonged elevation in blood

9. The rate of giving the salt injections is shown in Figure 1.
10. Table 2 illustrates the course in one case which is quite typical of the
series. It will be noted that the most striking change in the Wassermann
reaction occurred on the day of the first salt injection which, of course,
discredits the salt as being the effective agent in the subsequent unusually
rapid improvement in the fluid.

78 ARCHIVES OF INTERNAL MEDICINE

chlorid in the cases receiving the routine salt injection, the triviality and
transiency of the blood pressure changes is rather striking. It is
worthy of special note that several of the determinations were made on
the same patients at fortnightly intervals, and that in one such case
the blood pressure levels were slightly lower during the second deter-
mination than during the first: Nos. 1 and 2 (Table 3) are consecutive
determinations (eighteen day interval) on the same patient, and the
average of the blood pressures during the first was 112 6S, during
the last 101/58." These facts are of interest in view of the recent
work on the relationship of hypertension and elevations in blood chlorid.

T.'\BLF. 2.— Clinic.\l Course of One Case
Injected

/HicH S.ALT Solution

Treatment

Remarks

Reaexes

Spinal Fluid

Intra-
venous

(Diar-
senol in

Gm.)

Intraspinal

(50% diarsen-

olized Serum,

C.c. and

Diarsenol,

Gm.)

Date*

Cells Glob-
ulin

Wasser-

3/10/20
3/1-/20
4/ 7/20

4/14/20
4/21/20

0.3
0.4
0.3

0.4
0.3

0^4
0.3
0.4
0.4
0.5
0.4
0.3
0.3

o!3

0.3

20 + O.0OO2
20 + O.0O03

20 + 0.0003

20 + 0.0003

20 + 0.0004

20 + 0.0004
20 + 0.0004

20 + 0.0002
20 + O.0flO2

20 + 0.0003
20 + O.00O3
20 ->- 0.0002

No reaction

Out o£ city
since last
treatment

Symrtom
free

Negative
Negative

Negative
Negative
Negative

20
18

9
9

1

-1-

±

±

0.2 c.c. + +
0.4 c.c. + +

0.6 c.c. +-(-

5/12-20
5/19/20
6/ 2/20
6/17/20
6/29/20
7/13/20
7/26/20
8/10/SO

8/24/20
9/10/20
9/25/20

Rest period
12/17 /20t

Rest period
3/ ]/2i:

Rest period

08cc. -I--I-

Negative 7
Negative 15

10

200 c.c. 15%
salt solution

Salt aslbove
Salt as above

Negative

Negative
Negative
Negative

8

7
7
5

3

1.4 c.c. -

2 0 c.c

* The irregularity in treatment intervals is due to the fact that the patient was out (
city mucli of the time on business.

+ Control lumbar puncture. Symptom tree.
X Control lumbar puncture. Feels well.

With ten of the salt injections leukocyte counts were made imme-
diately before and after injection, an<ithe fresh and stained smears were
examined at the same time. Red cells showed no morphologic evidence

11. The same Rcncral blood pressure tendency, though less striking, char-
acterizes the consecutive determinations, Nos. 4, 5, and 6 (Table 3) and this
despite the fact that the blood chlorid never got t>ack to the original normal,
once salt injections were started ; the average of pressures in No. 6, though
higher than in \o. 5. is still lower than in the original determination. No. 4.

iryxx—XEUROsypHiLis

79

of injury; there were minor inconstant variations in the differential
count, and in about half the cases a slight depression in the total white
count. Xo effort was made to study variations in blood picture
exhaustively.

SUMM.VRV .VXD COXCLUSIOXS

1. Intravenous injections of 200 c.c. of 15 per cent, salt solution
were given to six patients with neurosyphilis, with resulting disagreeable
but not alarming symptoms. In these cases the cerebrospinal fluid
pressure was found to rise sharply and then to fall, reaching a point
about 100 mm. below the original level by thirty minutes after the end
of the salt injection.

TlguTt 3.

Fig. 2. — Upper left margin figures and top curve : whole blood chlorid in
gm. per liter. Lower left margin figures and lower two curves: systolic and
diastolic blood pressures. (During the three hours prior to injection, 200 c.c.
urine were voided, containing 0..S1 gm. salt; during the twenty hours after
injection, 1,270 c.c. were voided containing 16.4 gm. salt.)

2. Salt injections were given according to a definite routine over a
period of months, augmenting intraspinal treatment in a group of
patients with neurosyphilis. There was no serologic or cytologic
improvement over the usual course with intraspinal treatment alone,
and symptoms were distinctly aggravated. Such injections hence
would seem to have no therapeutic value in this group of neurosyphilis
cases.

A.

00

s

=

z«

"

"

"

"

"

"

"

i

a

1^1

o

a

3

g

s

a

s

^

o

a

c< n

,,

s

S

-=!

i^6

s

S

^

°

S

^

1

1

u

p

«

"^

g

S

s

S

s

s

5!

•3

n

z^

s

o

S

m

•X

„ „

"

.s^S

„

«

m

CO

CQ

0.

^

M

S w

1

i

§

s

3

i

1

1

1

m

II

m

s|s

o I o

sh

s 1°

§1 =

2|s

sl^

§|s

3

S.o^

51-

s|i5

s|s

S|s

S|s

S|s

§1^

sl^

gl5

1

§|g

§ |s

s|.^

§ls

l|s

g|s

sN

§|s

fe

■S„g

1"

l|°

s|s

Sle

§|3

a|s

S|s

sis

sis

S Si3

p

£Ka

u

:.a

¥■

Sl'^

i|s

sis

3|e

§|8

g|8

S|2

§|8

§11

m

.a

^

-

^•3 •

l|s

% \'&

§|s

s|s

g|s

lis

§1'^

S|s

<

1

Soil

SI

s|

Si

si

5f

s|

5f

sf

5f

Is^"

%

«K

Si.

§g

Si

!f

sg

Si

si

Sg

1
to

<

sf

si

if

si

if

si

ii

if

£

D3

<

»

■*

■•

"

"

■•

■•

jfllj

2

i

s

=;

6

s

i

i

1

i

1

1

g

§

g

-

J.

„

-

.

s

-

«

IVYXX—XEUROSYPHILIS 81

3. In a short series of cases whole blood chlorids were determined
before and at intervals after salt injections, the output of salt in the
urine was ascertained, and blood pressures were followed ; immediately
after the intravenous injection of 200 c. c. of 15 per cent, salt solution
the average whole blood chlorid elevation above normal (i.e., the first
determination) was 57 per cent.; one hour later, 26 per cent.; twelve
hours later, 18 per cent. In from seventeen to twenty-two hours, about
half of the injected salt had been excreted in the urine. \'ariations
in blood pressure were within physiologic limits.

A METABOLIC STUDY OF PROGRESSIVE PSEUDOHY-
PERTROPHIC MUSCULAR DYSTROPHY AND
OTHER MUSCULAR ATROPHIES*

R. B. GIBSON, Ph.D.; FRANCIS T. MARTIN, B.S.,

AND

MARY VAN RENNSELAER BUELL, Ph.D.

IO\V.\ CITY

We have recently had the opportunity to study the metaboHsm of
nine cases of pseudohypertrophic muscular dystrophy in different
stages of advancement. These cases will be reported from a clinical
standpoint with about twenty more by Dr. R. V. Funsten. We also
include some metabolic observations on other types of atrophic mus-
cular involvement.

Endocrine disturbance in progressive pseudohypertrophic muscular
dystrophy is indicated by (I) the hereditary character of the condi-
tion, (2) the metabolic abnormalities, (3) the occasional recovery
at puberty when glandular readjustments occur, (4) reported improve-
ment following endocrine therapy in some cases, and (5) the develop-
ment of the disease in polyglandular dystrophies, notably in association
with dystrophia adiposogenitalis. Necropsy findings with special ref-
erence to the ductless glands are urgently needed to elucidate further
the pathogenesis of the condition.

The symptomatology' and the pathologic changes have been well
reviewed by Timme ^ and by Janney, Goodhart, and Isaacson.^
Involvement of the pineal gland has been suggested by Timme from
a study of the literature and from roentgen-ray examinations in
three cases from the same family ; pineal shadows were evident
in two cases only of the series of nine studied by Janney, Goodhart,
and Isaacson. These last named investigators, as does McCrudden,^
attribute the condition to endocrine dysfunction affecting carbohy-
drate metabolism.

According to McCrudden, there is hypoglycemia, no increased heat
formation, a rapid fall in blood sugar during the first twenty-four
hours of starvation indicative of a diminished glycogen reserve, increas-

*From the Chemical Research Laboratory of the Department of Theory
and Practice of Medicine and Clinical Medicine, in cooperation with the
Department of Home Economics and the Graduate College, the State Lni-
versily of Iowa.

1. Timme, W. : Arch. Int. Med. 19:79 (Jan.) 1917.

2. Janney, N. W.; Goodhart, S. P.. and Isaacson, V. I.: Arch. Int. Med.
21:188 (Feb.) 1918.

3. McCrudden, F. H.: Arch. Int. Med. 21:256 (Feb.) 1918; J. A. M. A.
70:1216 (April 27) 1918.

GIBSOX-MARTIX-BUELL— MUSCULAR DYSTROPHY 83

ing "fatty degeneration," an increased respiratory quotient, and lipeniia.
Since a prompt and marked rise in blood sugar follows the adminis-
tration of epinephrin, the impaired glycogenesis is associated with
damage to the suprarenals rather than to the liver.

One of our patients (Case 5) died of influenzal pneumonia and
a necropsy was obtained. This was a polyglandular condition, Froh-
lich's syndrome complicating the picture of progressive pseudohyper-
trophic muscular dystrophy. Through oversight, the pineal gland
was not preserved for microscopic examination. The following descrip-
tion is abstracted from Dr. E. M. Medlar's records.

On gross examination, the suprarenals were apparently larger, though
thinner than normal, with a very small medullary substance and a thin
cortical layer. The thyroid was slightly hypertrophied and there was con-
siderable colloid. The parathyroids (four, the size of small peas) were red-
dish in color; on section, they seemed to be made up of small round bodies.
The pituitary contained a cyst filled with a clear fluid ; the cyst was about
0.5 cm. in diameter; the remaining tissue was flattened and less in amount
than normal.

Microscopic examination revealed that the bodies taken to be parathyroids
consisted of acini filled with colloid material; "if these bodies are accessory
thyroids, the parathyroids were so small that they could not be detected in
gross." Sections of the pituitary showed a small cyst in the pituitary lobe,
with an apparent marked increase in the secretory cells (and secretion) of
the anterior lobe. Thyroid, thymus and pancreas exhibited little or no change,
except that for the last two there was an ingrowth of fat tissue. The
adrenals were congested, and there were areas of necrosis with infiltration of
leukocytes in the cortex.

The muscle tissue showed the fatty infiltration and atrophy as described for
pseudohypertrophic muscular dystrophy. It is of incidental interest that the
respiratory muscles were less atrophied than the others, indicating a selective
conservation of this group.

Because of the muscular changes, a study of the creatinin elimi-
nation in the muscular atrophies was made by Spriggs * ; he observed
a lowered creatinin excretion in pseudohypertrophic muscular dystro-
phy, in myotonia congenita, and in myasthenia gravis. Levene and
Kristeller ^ determined the creatinin and creatin elimination in several
types of muscular involvement, including five cases of muscular dys-
trophy, on high and low protein intakes ; cases with considerable
loss of muscular function gave a much diminished creatinin output and
a creatinuria. In conditions of muscular atrophy or dystrophy prac-
tically 90 per cent, of ingested creatin (in beef) reappeared in the urine.
McCrudden and Sargent " found the daily creatinin excretion to
be normal and constant (1.5 gm. or 22.6 mg. per kilograin of body
weight) and a creatinuria (about 0.5 gm. daily) in a male, aged idi

4. Spriggs, E. I.: Biochem. Ztschr. 2:206, 1907.

5. Levene, P. A., and Kristeller, L. : Am. J. Physiol. 24:45, 1907.

6. McCrudden. F. H., and Sargent, C. S.: Arch. Int. Med. 17:465 (April)
1916; 21:252 (Feb.) 1918.

84 ARCHIVES OF INTERNAL MEDICINE

years. There was a low ammonia. Total nitrogen, uric acid, and
calcium and magnesium elimination were normal, though the ratio
calcium : magnesium was a little high. Blood findings were : creatinin,
1.43 mg. ; creatin, 3.86 mg. ; nonprotein nitrogen, 28.9 mg. ; uric acid,
2.3 mg. ; glucose, 0.075 per cent.; and cholesterol, from 0.05 to 0.144
per cent. They state that the metabolic picture is distinctly different
in progressive muscular atrophy (two cases) in which there is crea-
tinuria and normal blood sugar, and in myasthenia gravis (two cases)
in which there is hypoglycemia without creatinuria.

The nine cases reported by Janney, Goodhart, and Isaacson were
in patients in advanced stages of the disease, and from 11 to 48
years of age; one female case is included. They found a marked
decrease in the urinary creatinin, creatinuria, hypoglycemia, normal
blood urea, low blood creatinin, and normal ( ?) blood creatin. There
was a retention of calcium and magnesium when the calcium content
of the diet was adequate. Of importance is the delayed utilization of
glucose for five cases when ingested (1.75 gm. glucose per kilogram of
body weight).

Brock and Kay ' have presented three adult cases of endocrin-
opathies associated with unusual manifestations in the muscular system.
One was a case of dystrophia adiposogenitalis with a recent limited
muscular dystrophy. This case showed a creatinuria, hypoglycemia,
and a delayed glycogenesis without glycosuria following glucose inges-
tion. A. Gibson ^ has reported a case of familial "muscular infantil-
ism" in an adult male. There was creatinuria, though histologic
examination of the muscle showed but little dejiarture from the
normal. Blood sugar findings resembled the results we have obtained
for our two adult cases of pseudohypertrophic muscular dystrophy.

Pemberton '■' has reported a somewhat lowered creatinin elimination
and a negative calcium balance in myasthenia gravis, and a very low
creatinin and slightly negative calcium balance in myotonia atrophica.
Diller and Rosenbloom ^^ obtained similar results for a case of myas-
thenia gravis ; there was no creatinuria. Rosenbloom and Cohoe ^^
found a normal nitrogen partition in myotonia congenita (Thomp-
sen's disease) ; there was a negative calcium balance. Three studies
of the metabolism in amyotonia congenita (Oppenheim's disease) in
children are found in the modern literature ; Gittings and Pemberton '^
reported a very low creatinin excretion in a boy 21 months old, with

7. Brock, S., and Kay. W. E. : Arch. Int. Med. 27:1 (Jan.) 1921.

8. Gibson, A.: Arch. Int. Med. 27:338 (March) 1921.

9. Pemberton. R. : Am. J. M. Sc. 139:816. 1910; 141:253, 1911.

10. Diller. T.. and Rosenbloom, J.: Am. J. M. Sc. 143:65. 1914.

11. Rosenbloom, J., and Cohoe. B. A.: Arch. Int. Med. 14:263 (Aug.) 1914.

12. Gittings, J. C, and Pemberton. R.: Am. J. M. Sc. 144:732. 1912.

GIBSOX-MARTIX-BUELL— MUSCULAR DYSTROPHY 85

no disturbance of the calcium elimination : Powis and Raper " obtained
low creatinin and high creatin figures for a girl 4 years of age,
ingested creatin being promptly eliminated, and calcium and potas-
sium retention being normal ; and Ziegler and Pearce ^* found low
creatinin and high creatin values, normal uric acid, an increased
undetermined nitrogen and neutral sulphur, blood glucose of 0.14
per cent., and blood creatin and creatin of 1.5 and 5.45 mg. respectively.

The nine cases of progressive pseudohypertrophic muscular dys-
trophy included in this report comprise: Two cases in young boys
in which the atrophy was moderately advanced. Three cases in older
boys (one with a coincident dystrophia adiposogenitalis) in which
the muscular atrophy had progressed until movements were practi-
cally limited to the muscles of the thorax, distal upper extremities,
neck, and face ; and two male adult cases, the condition developing
after puberty and progressing slowly, with muscular pseudohyper-
trophy, lordosis, and general weakness. Of particular interest frnm
a comparative metabolic standpoint is the fact that two of the patients
in the series were brothers, one with a moderately advanced case and
the other presenting marked atrophic changes. One case of myasthenia
gravis, one of muscular atrophy following acute anterior poliomyelitis,
and one of myositis ossificans are also included. These three cases
were adult males also.

Each case was transferred to our special metabolism unit for study.
The diet was nonpurin and noncreatin, the constituents being kept
essentially the same daily. The protein intake was set at a figure
slightly above the physiologic requirement only, inasmuch as the litera-
ture on creatinuria indicates that an increased protein metabolism may
be accompanied by a greater creatin elimination, a fact which we have
verified in some of our cases. ^^ Blood sugar determinations were made
according to the Benedict modification of the Lewis-Benedict method :
other analyses were made by the current procedures. Blood samples
for Case 6 could only be obtained from the jugular vein. It was impos-
sible to do a satisfactory venepuncture in the myositis ossificans case,
and in Case 5. though micro-sugar determinations were made for the
latter with 0.04 c. mm. of blood, centrifuging the precipitates, and
using similar Sahli hemoglobinometer tubes as a dilution colorimeter.

Levene and Kristeller have pointed out that the creatinin elimina-
tion in muscular atrophy cases does not have the same constancy that
obtains in the normal individual. From data obtained in our patients,
we find this to be decidedly the case, the twenty-four liour collection

13. Powis, F., and Raper, H. S.: Quart. J. M. 10:7, 1916.

14. Ziegler, M. R.. and Pearce, N. O. : J. Biol. Chem. 42:581, 1920.

15. Gibson, R. B., and Martin, Francis T.: J. Biol Chem. 41: xxxvi, 1920.

86 ARCH WES OF IXTERXAL MEDICIXE

of the urinary specimens being absolutely assured. However, we have
found it possible to stabilize the creatinin excretion under constant con-
ditions of diet over periods of several days.

A summarj- of the ages, weights, creatinin coefficients, and blood
findings is given in Table 1. Brief clinical descriptions and the nitro-
gen partition in tabular form are presented also for the individual
cases.

TABLE 1. — Creatinin and Creatin Coefficients, .\nd Blood An.\lyses of
Twelve Cases

s:

^

T^ is °^ °^ ^ 31

^1

a 5 £= ou "^ o- .&5 oa oa o oS, oa g;s
<&qEhSs o mm mnmW

1. E.T 5 18.0 5.2 9.0 48 0.070 O.098 0.75 10.3 0.61

•2. D. R 8 21.8 5.0 11.5 32 0.071 0.067 0.45 5.7 0.S9

3. J. Schw. Adm. 1.. 5 18.2 2.5 4.3 30 0.080 0.174 0.91 7.6 0.59 7.5

Adm. 2.. 6 20.0 2.6 5.6 D.0C7 0.101

4. B. Stol. Adm. 1... 9 20.0 2.1 3.9 90 0.112 0.101 O.fiO 7.2 0.67 0.29

0.119 0.115 1.06 7.5 0.57 0.28

5. O. Schaf 15 90.0 1.0 3.4 To 0.089 0.1T8

6. Wm Schw.Adm.l 15 21.3 1.0 3.7 SO 0.069 0.160 0.76 5.7 0.61 7.6

Adm.2 15 .... 1.6 7.5 .. 0.117 0.198 0.53 9.7 0.50 0.18

7. W. Stil 12 30.6 1.7 5.0 100 0.125 0.147 0.50 7.0

0.082 0.124 0.50 15.6 0.95

8.B.MCD 22 55.0 7.1 7.7 40 0.133 0.157 .... 15.1 10.1 ...

0.116 0.143 2.45 12.9

1.9 30.0 0.62 0.22 11.8 ...

9. 0. Sm 38 50.0 4.2 5.7 ... 0.101 0.123 1.0 12.1 0.51 0.27 12.6 ...

10. Geo. S 48 79.0 7.1 7.1 ... 0.105 0.74 5.2

0.125 0.185 1.4 7.7

11. L. W 25 63.0 7.6 8.3 60 0.112 0.111 1.25 6.8 0.72 0.22

12. Geo. F £6 54.0 5.3 5.S

Cases 1 to 9. inclusive, were pseuiiohypertropliic muscular dystrophy patients; Case 10
had myasthenia gravis; Case 11 muscular atrophy due to anterior poliomyelitis, and Case 12
myositis ossificans.

In the cases in the earlier stage of progressive pseudohypertrophic
muscular dystrophy (Cases 1 and 2), there was creatinuria; this is a
normal condition in childhood (Rose "). Shaffer '' gives the creatinin
coefficient for normal children as between 3.3 and 6.5, and for adults
from 8.1 to 11.0. We regard creatinin coefficients of 5.2 and 5.0 for
Cases 1 and 2 as a low normal. Ingested creatin (0.5 gm.) was
destroyed only in part, 48 and 32 per cent., respectively, being recov-
ered. Krause's " normal cases, given 0.3 gm. creatin by mouth, elimi-
nated the following percentages ; girl of 6 years, 56 ; boy of 8 years, 43 ;
girl of 11 years, 31. The recovery of ingested creatin in our two cases
is not to be regarded as abnormal. A rather high uric acid nitrogen
for Case 2 was observed. Total nitrogen figures were quite low, and
positive nitrogen balances are indicated. There was hypoglycemia in

16. Rose. \V. C: J. Biol. Chem. 10:265, 1911.

17. Shaffer, P. A.: Am. J. Physiol. 23:1, 1908.

18. Krause, R. A.: Quart. T. Phvsiol. 8:87, 1913.

GIBSOX-MARTIX-BUELL— MUSCULAR DYSTROPHY 87

both cases. Diminished glycogenesis (Janney) and high blood creatin
obtained for Case 1 ; normal blood creatin figures for children are not
yet established satisfactorily, though Veeder and Johnston ^^ found
from 2.2 to 7.2 mg. with a mean for the nonextreme cases of 3.8.

Still lower creatinin coefficients (2.5 and 2.1) were obtained for
Cases 3 and 4; in these boys, the muscular changes were more pro-
nounced than for the previous group. There was creatinuria in both
cases. Hypoglycemia and diminished glycogenesis were found for
Case 3, but not for Case 4. Case 4 was clinically the more advanced,
and this is reflected by the Jower creatinin coefficient and the much
greater recovery of ingested creatin.

Cases 5,. 6, and 7 were older boys of 15, 15, and 12 years, respec-
tively. There were most pronounced atrophic changes of the muscular
system. Strikingly low creatinin excretion and creatinin coefficients
(1.0 to 1.6) were obtained; however, the creatinin coefficient of Case 5
(with Frohlich's syndrome) is based on a body weight of 90 kg. There
was creatinuria, and ingested creatin was largely recovered, completely
so for Case 7. Hypoglycemia was observed, though normal blood
sugars have been obtained also for Cases 6 and 7. Glycogenesis was
reduced; glycosuria followed the glucose ingestion (100 gm.) for
Case 5. High blood creatins have been noted for Cases 6 and 7. The
high blood fat for Case 7 is probably accounted for by an excessive
diet. Case 6 was a brother of Case 3.

The adult pseudohypertrophic muscular dystrophy cases (Cases 8
and 9) present some differences that are noteworthy. The metabolic
disturbance is not so intense. Case 9 presented a greater atrophic
change with a much less creatinin coefficient and a greater creatinuria
than Case 8. In fact, there was little or no creatinuria on some days
for Case 8. P'orty per cent, of ingested creatin (1 gm.) was excreted
by Case 8. For Case 9, there was an increased creatin excretion fol-
lowing creatin ingestion at least over 40 per cent., but the metabolism
was not stabilized, and the result was unsatisfactory. In the normal
male adult, ingested creatin disappears. Normal blood sugar obtained
for Case 8 (0.116 and 0.133) and a low normal figure (0.101) for
Benedict's method was found in Case 9. Both patients showed a some-
what deficient glycogenesis without glycosuria following glucose inges-
tion (100 gm.). High blood creatinin and very high creatin were
observed for Case 8, and high blood creatin for Case 9.

Such figures as we have obtained for blood fat (e.xccpt as noted),
blood cholesterol, hydrogen ion concentration of the blood, and plasma
calcium arc not to be regarded as significant.

19. Veeder. B. S., and Johnston. M. R.: .\m. I. Dis. Child. 12:1.36 (Aug.)
1916.

88 ARCHIVES OF ISTERNAL MEDICIXE

We. found, as did IMcCrudden. no creatinuria in myasthenia gravis
(Case 10). A rather low creatinin coefficient (7.1) was observed.
There was a normal blood sugar rather than a hypoglycemia and a
delayed storage of ingested glucose. Incidentally, combined therapy
with calcium lactate, cafifein citrate, desiccated whole pituitary sub-
stance, and epinephrin injections over a month resulted in no clinical
improvement.

In the case of muscular atrophy resulting from acute anterior
poliomyelitis in an adult (Case 11), there was creatinuria, a somewhat
lowered creatinin coefficint of 7.6, and 58 per cent, of ingested creatin
(1 gm.) was recovered. Normal blood sugar and glycogenesis were
found.

A case of myositis ossificans (Case 12) is included in the present
series. The metabolic picture is characterized by a high uric acid
nitrogen, a low creatinin (coefficient 5.2), and no creatinuria. There
was a daily nitrogen retention of 2.28 gm. and positive balances for
calcium of 1.14 gm. and for magnesium of 0.46 gm. over a five day
period.

Inasmuch as carbohydrate deficiency induces creatinuria (Mendel
and Rose^"), it has been argued that the faulty carbohydrate metab-
olism is responsible for the creatinuria in progressive pseudohyper-
trophic muscular dystrophy. We are inclined to believe that the
atrophic dysfunction and actual diminution of the amount of muscle
tissue are the conditions inducing the creatinuria, just as in the non-
endocrine muscular atrophies where the carbohydrate metabolism is
normal as in Case 11. The lowest creatinin coefficients occur in those
cases in which the greatest muscular atrophy has occurred ; likewise, the
recovery of ingested creatin is greatest in the advanced cases. The
creatinin and creatin figures are not particularly abnormal in the earlier
atrophic changes of the disease, though the endocrine disturbance must
be well established at that time. Again normal blood sugar may be
obtained in the muscle dystrophy cases (Cases 4, 8 and 9; see also
Cases 6 and 7 ) and even normal glycogenesis may be observed (Cases
2 and 4).

From the literature and our own observations, then, the outstanding
metabolic features of progressive pseudohypertrophic muscular dys-
trophy are :

1. Those associated with the atrophic condition of the muscles, and
which are intensified as the atrophy progresses :

1. Lowered creatinin e.xcretion and creatinin cocfticient.

2. Creatinuria.

3. Recovery in the urine of ingested creatin.

4. Creatinemia, though high blood creatins are not a constant finding.

20. Mendel, L. B., and Rose, W. C: J. Biol. Chera. 10:213, 1912.

GIBSOX-MARTIX-BUELL-^MUSCULAR DYSTROPHY 89

2. Disturbances of carbohydrate metabolism of endocrine origin,
usual but not constant findings.

1. Hypoglycemia.

2. Deficient glycogenesis following moderate glucose ingestion, and com-
monly without glycosuria.

The differential diagnosis of progressive pseudohypertrophic mus-
cular dystrophy, myasthenia gravis, and progressive muscular atrophy
may be checked according to McCrudden's suggestion by the metabolic
and blood findings. The characteristic differences to be expected, with
the addition of the glucose tolerance, are summarized as follows:

Creatiniu-ia

Blood Glucose

Glucose Tolerance

Progressive pseuaohypertrophic muscular

Present

Low

Diminished

Myasthenia gravis

.\bsent

Low or normal

Diminished

Progressive muscular atrophy

Present

Noniial

There is little evidence on which to base a satisfactory therapeutic
procedure for progressive pseudohypertrophic muscular dystrophy.
McCrudden reports improvement in his case as the result of pituitary
extract and epinephrin treatment, and a return of the blood sugar to
normal. One patient (Case 1) in this series has apparently recovered,
according to Dr. Steindler, so far as physical examination and strength
tests may be depended on ; we have not had the chance to make a fur-
ther metabolic study. Whether or not the recovery was spontaneous with
the prevention of overexertion and an adequate diet, or due to therapy
with calcium lactate (0.3 gm., t.i.d.) is undetermined ; calcium therapy
was employed because of the suggested part it plays in carbohydrate
metabolism (Underbill and Blather wick =') and the reported favorable
influence in other obscure myopathies. One patient (Case 3) has not
improved on calcium lactate and dried whole pituitary gland. Pituitary
extract (Parke, Davis, and Co., pituitrin, obstetric preparation) injec-
tions given daily for six weeks to one patient (Case 8) had little or
no effect. All of the patients gained in weight and improved some-
what under conditions of hospitalization.

SU.MMARV AND CONCLUSIONS

Seven cases of progressive pseudohypertrophic muscular dystrophy
in boys and two adult male cases have been studied. One case each of
myasthenia gravis, muscular atrophy following acute anterior poliomye-
litis, and myositis ossificans are included in this report.

The progress of the atrophic condition of the muscular system in

progressive pseudohypertrophic muscular dystrophy is indicated meta-

bolically by a diminishing creatinin excretion and creatinin coefficient,

21. Underbill, F. P., and Blatherwick, N. R.: J. Biol. Chcm. 19:119, 1914.

90 ARCHIVES OF IXTERXAL MEDICISE

by the creatimiria. and by the incomplete destruction of ingested
creatin. The creatinin — creatin picture is quite similar to that in mus-
cular atrophy where there is no disturbance of carbohydrate metabolism.

AVhile hypoglycemia and deficient glycogenesis are characteristic for
progressive pseudohypertrophic muscular dystrophy, normal blood
sugar figures may be obtained, and normal glycogenesis may occur
even when hypoglycemia is present.

McCrudden's metabolic differentiation (with the addition of glucose
tolerance findings) of progressive pseudohypertrophic muscular dys-
trophy, myasthenia gravis, and progressive muscular atrophy is essen-
tially confirmed and should be of value in the diagnosis of these
conditions.

In a case of myositis ossificans, we have observed a diminished
creatinin excretion without creatinuria, an increased output of uric
acid, and positive nitrogen, calcium, and magnesium balances.

Apparent recovery over a period of a few months is reported in
an early recognized typical case of progressive pseudohyj^ertrophic
muscular dystrophy. This case was a boy. 5 years of age.

The need of a study of the necropsy findings with special reference
to the ductless glands in pseudohypertrophic muscular dystrophy is
emphasized.

We are indebted to Dr. Arthur Steindler, who has referred several of the
cases in this series to this laboratory for study.

The diets were calculated and prepared under the supervision of Miss
Gertrude Whiteford and Miss Lela E. Booher.

ADDENDUM

Since this paper was written, we ha\e studied another case of
pseudohypertrophic muscular dystrophy in a boy, 10 years of age. The
findings in this case bear out our contentions regarding the creatin and
carbohydrate metabolism. The patient presented the characteristic
picture of the disease; the atrophy had progressed to a stage inter-
mediate between Cases 3 and 4. The clinical history and nitrogen
metabolism are given under Case 13.

The creatinin coefficient was 2.2 and the total creatinin (creatinin
plus creatin) coefficient was 5.6. About 60 per cent, of the 0.5 gm.
creatin given by mouth was recovered for that day. Blood sugar was
0.144 per cent, and two hours after the ingestion of 44 gm. glucose it
was 0.138 per cent., with the hourly urines negative for sugar. There
was, therefore, a hyperglycemia and a normal sugar tolerance. Blood
creatinin was 1.1 mg. and blood creatin was 5.2 mg. per hundred c.c.
Blood cholesterol was 0.22 per cent. The basal metabolic rate was
51.8 calories per square meter of body surface per hour; the expected
normal was 52 calories. The respiratory quotient was 0.85.

GIBSOX-MARTIX-BUELL— MUSCULAR DYSTROPHY 91

REPORT OF C.\SES

Case 1. — Proyrcssive pscudohypcrtropliic muscular dystrophy.

E. T., white, male, 55 years of age, weight 18 kg. The condition was first
noticed a month previous to admission to the hospital. He had always seemed
normal, though easily fatigued. This patient has since discharge apparently
recovered.

There was muscular atrophy as follows : adductors of the thumb ; muscles
of the thenar and hypothenar groups ; muscles of the shoulder girdle ; long
muscles of the back ; gluteal muscles ; thigh muscles, and calf muscles. Histo-
logical examination of the muscle was characteristic for pseudohypertrophic
dystrophy.

The case was transferred to the metabolism unit Oct. Z2. 1919, and put on
a nonpurin and noncreatin diet of 30 gm. protein, 37 gm. fat, and 125 gm.
carbohydrate, the total calories being 1,000.

T.\BLE 2. — Urine Nitrogen Partition (Case 1)

Day

10/23
10/24
10/25
10/26
10/27

TotalX
Gm.
4.03
4.35
3.96
3.50
3.30

Urea K

Gm. %
3.57 88.5
3.89 89.3
3.43 86.5
2.96 83.0
2.64 75.4

Ammonia N

Gm.' %
0.215 5.34
0.210 4.83
0.195 4.92
0.2ed 7.57
0.262 7.49

Uric Acid N

Gm. %
0.065 1.81
0.060 1.38
0.03S 1.47
0.073 2.08
0.058 1.66

Creatinin N

Gm. %
0.098 2.43
0.098 2.14
0.089 2.25
0.003 2.66
0.089 2.54

Great

inX

Remarks

In bed
Active
Active
0.5 gm. creatin

Gm.
0.061
0.067
0.066
0.155
0.055

%
1.53
1.54
1.66
4.33
1.57

Case 2. — Progressive pseudohypertrophic muscular dystrophy.

D. R., white, male, 8 years of age, weight 21.8 kg. The patient first walked
at 3 years of age, but could not turn. He tired easily. The condition was
progressive. He was admitted to this hospital Nov. 17, 1919.

The patient had a distinctly asthenic gait; he walked with the legs abducted.
There was contracture of the tendo Achillis of each leg, and the left foot
showed a tendency to varus. Both calves were unusually developed. The mus-
cles of the arms and back seemed normal.

While in the metabolism unit, he was on a diet of 35 gm protein, 45 gm.
tat, and 115 gm. carbohydrate, with a caloric value of 1,000.

TABLE 3. — Urine Nitrogen Partition (Case 2)

Day

10/19
10 '20
10/21
10/22

Total N

5.49
6.51
4.76
6.37

Urea N

Gm. %

4.59 83.6
4.41 67.7
3.72 78.1

3.60 57.9

Ammonia N
Gm. %
0.454 8.27
0.460 7.07
0.335 7.01
0.461 7.24

Uric Add N
Gm. %
0.095 1.73
0.108 1.65
0.108 2.17
0.121 1.90

Creatinin N
' Gm. ■ % '
0.110 2.00
0.106 1.61
0.114 2.31
0.109 1.71

Creatin N

0.144 2.63
0.143 2.24
0.198 4.00
0.095 1.49

Remarks

In hed
Active
0.5 gm. creatin

Case 3. — Progressive pseudohypertrophic muscular dystrophy.

J. Schw., white, male, 6 years of age, weight 182 kg. The condition was
noted at the time he began to walk, and was characterized by progressive
muscular weakness. At the time of admission, the patient was unable to get
up stairs, and had to brace himself when he got up from the floor.

There was a decided ataxic gait. The calf muscles were much enlarged.
The arms were notably weakened. Lordosis was present.

Three boys by an earlier marriage of the mother had been similarly affected
and died. An older brother of this patient by her present husband is one of
our series (Case 6). One girl by the first marriage and two by the second
are alive and well.

92 ARCHIVES OF IKTERXAL MEDICIXE

This case was under observation in the summer of 1919, and again in the
winter of 1919-1920. Urinary nitrogen partition figures are given for the
second admission. The diet was nonpurin and noncreatin. There were 30 gm.
protein, 40 gm. fat and 150 gm. carbohydrate, with a total of 1.039 calories.

T.\BLE 4.— Urine Nitrogen Partition (Case 3)

Total N,

Bay tjxii.

12/14 4.97

12/15 4.5.';

12/lG 5.39

12/17 4.55

12/18 4.13

12/19 4.16

Uric Acid N

Creatinin S

Gm.

%~

Gm. %

0.091

1.83

0.048 0.97

0.092

2.02

0.051 1.12

0129

2.40

0.048 0.S9

0.136

2.99

0.054 1.19

0.117

2.83

0.053 1.28

0.101

2.43

0.053 1.27

Case 4. — Progressive pseudohypertrophic muscular dystrophy.

R. S., white, male, 9 years of age, weight 20 kg. The patient walked first
at 2.5 years. The previous history of his ilhiess was typical.

The patient walked with a peculiar gait, stumbled a great deal, but could get
about slowly. He would get up from the floor only with difficulty, using
the hands on the thighs ; at times he could not get up unless assisted. The
legs were weak and the ankles turned outward. There was lordosis, the
scapulae were winged, and the calf muscles were pseudohypertrophied.

He was placed on a diet consisting of 32 gm. protein, 45 gm. fat, and ISO
gm. carbohydrate, making a total of 1,165 calories, for a period from Decem-
ber 2 to Dec. 9, 1920. He was in the metabolism unit again during a second
admission (Jan. 14, 1921) for an additional glucose tolerance test only.

TABLE S. — Urine Nitrogen Partition (Case 4)

Day

Total N
Gm.

UreaN

Ammonia N

Uric Aeid N

Creatinin N

Creatin N

R ma

6m.

%

Gm.

%

' Gm.

%

Gm.

%

Gm.

%

12/4

2.23

0.072

3.23

0.042

1.88

0.036

1.61

12/5

2.39

1.82

7«.2

0.143

6.98

0.080

3.35

0.045

1.89

0.039

1.71

12^6

2.00

1.49

74.8

0.105

5.26

0.098

4.91

0.035

1.75

0.029

1.46

12/7

2.31

1.«2

70.1

0.126

5.46

0.112

4.85

0.041

1.77

0.049

2.13

12/8

2.70

1.99

73.7

0.102

3.78

0.125

4.64

0.051

1.89

0.145

5.38

0.5 gm.

creatin

12/9

2.50

1.92

76.8

0.105

4.20

0.089

3.56

0.053

2.12

0.082

3.28

The sulphur

Dartition was determined

on a

composite sample

of the

urines

lor

December 4

and 5.

The

results

are given as

grams

1 ol SO3 per

day

Inorganic S

Ethereal S

Neutral S

Gm.

%

Gm.

%

%

0.608

5.393

77.4

0.072

14.2

0.042

8.3

Case S.— Progressive pseudohypertrophic muscular dystrophy: dystrophia
adiposogenitalis (Frolich's syndrome).

O. Schaf., white, male. 15 years old, height 140 cm., weight 90 kg. The
condition was first noted eleven years previous to admission to the hospital
(Nov. 28, 1919). He had not walked for three years. A sister is similarly
affected. The patient died of influenzal pneumonia. The necropsy findings are
discussed elsewhere in this paper.

The facies were typical for dystrophia adiposogenitalis. The muscles of
the extremities, the thigh group, and the psoas appeared large and flabby.
Fle.xion and extension of the arms, forearms, thighs and feet were little or

GIBSON-MARTIX-B UELL— MUSCULAR D YSTROPH Y

93

nil. The calf muscles were a fourth normal strength. On roentgen-ray exam-
ination, the sella turcica was irregular and indefinitely outlined; the antero-
posterior diameter was 14 mm., the superior-inferior diameter 9 mm.

The patient was given a diet comprising 75 gm. protein, 70 gm. fat and
190 gm. carbohydrate, with a calorific value of 1,700.

TABLE 6. — Urine Nitrogen P.\rtition (Case 5)

Urea X Ammonia N Uric Aeid N Creatinin X Creatin N

12/ 6 7.14 5.03 70.4

12/ 7 7.70 5.17 67.1

12/ 8 7.35 5.48 74.5

12/ 9 8.23 5.87 71.3

Gm.
0.703
0.503

o.en

78.3 0.523 5.0

1.129 1.84 0.166

0.098
0.079
0.060

0.366 6.28

0.269 2.96

0.216 2.38

0.278 3.08

Case 6.— Progressive pseudohypertrophic muscular dystrophy.

Wm. Schw., white, male, 15 years of age, weight 21.3 kg. The condition
was first noticed seven years previous to the first admission to the hospital
(July 8, 1919). The history of the progress of the disease was typical. This
case was an older brother to Case 3.

The patient could not walk. The muscular system was extremely atrophied,
and there were contractures of the feet, knees, and flexion contracture of the
hips. There was no motion in the shoulders, knees and hips, and but slight
motion in the elbows and fingers. The back was considerably curved.

The urinary nitrogen figures below were obtained on the second admission.
The diet for December 13-19 contained 30 gm. protein, 40 gm. fat and ISO gm.
carbohydrate, with an energy value of 1,080 calories ; for the later period,
January 20-22, the patient was given 40 gm. protein, 40 gm. fat, and ISO gm.
carbohydrate, equivalent to 1,120 calories. At this time he weighed 22.7 kg.

TABLE 7.— Urine Nitrogen Partition (Case 6)

Urea

N

Ammonia N

Uric Acid N

Creatinin X

Creatin X

^%^i'-

Day

Gm.

%

Gm.

%

Gm.

%

Gm.

%

Gm.

%

12/13

4.20

2.64

62.9

0.363

8.61

0.109

2.60

0.0^9

0.69

0.122

2.91

12/14

3.99

3.13

78.4

0.440

11.0

0.136

3.41

0.027

0.63

0.111

2.78

2.59

52.9

0.384

7.84

0.125

2.47

0.028

0.57

0.160

3.27

12/18

3.78

2.56

67.7

0.280

7.41

0.122

3.28

0.030

0.79

0.149

3.95

12/19
1/20
1/21

0.293

6.96

0.123

2.93

0.038

0.79

0.153

3.B4

5.04

3.53

70.0

0.321

6.37

0.031

•0.62

0129

2..W

5.51

4.20

76.0

0.265

4.81

0.033

0.60

0.126

2.29

1/22

4.97

3.76

75.6

0.367

7.36

0.032

0.61

0.131

Case 7. — Progressive pseudohypertrophic muscular dystrophy.

W. Stil., white, male, 12 years of age, weight 30.5 kg. The condition was
first noted four years before admission (Sept. 5. 1919). The patient had not
walked for two years. There was no familial history of the disease.

There was general muscular weakness, marked muscular atrophy (espe-
cially of the deltoids), and contractures of the hamstrings, toes and hips. The
calf muscles showed the characteristic pseudohypertrophy.

The patient was observed for three months in the metabolism unit. At the
time of his discharge he weighed 39 kg. The diet given for the metabolic
period tabulated in Table 8 consisted of 32 gm. protein, 75 gm. fat and 200
gm. carbohydrate; the energy value was 1.600 calories.

ARCHII'ES OF IXTERXAL MEDICIXE
TABLE 8.— Urine Xitroge.v Partition (Case 7)

Drea N Ammonia N Uric Acid N Creatim'n N Oreatin 1

Day

■^■^'^/^

G^

%

Gm.

%

Gm.

%

Gm.

~%^

Gm. %

Hem arks

10/12

3.98

■"0

.55?.

0.362

9.10

0096

2.46

0.054

^.xi

0.095 2.41

0.238

7.91

0.067

2.23

0.047

1.56

O.IOO 3.32

S.9R

0.078

1.RS

0.060

0.253 6.01

10/15

3.59

?30

fi4.1

0.288

8.03

0.057

1.59

0.059

1.31

0.127 3.54

69.3

0.402

12.6

0.071

2.28

0.054

1.69

0.100 3.14

ion:

13.2

0.0-3

2.34

0.099 3.17

10,18

3.15

2.30

73.0

0.353

11.2

O.057

1.81

0.053

1.6S

0.101 3.21

Case 8. — Progressive pseudohypertrophic muscular dystrophy.

B. McD., white, male, 22 j'ears of age. weight 53 kg. The condition was
first noted a year previous to entering the hospital. There was increasing
weakness and the arms and shoulders were getting smaller. The patient's
normal weight was 60 kg.

There was moderate lordosis and the trunk was tilted to the left. The
neck muscles were normal as regards strength and motion. The shoulders
were not winged. The biceps and triceps had a tendinous feel. The calf mus-
cles had the firm consistency of pseudohypertrophy. The patient walked with a
waddling gait, but was quite active.

The diet consisted of 70 gm. protein, 90 gm. fat and 250 gm. carbohydrate ;
the energy value was 2,000 calories. On days June 2 and June 3, 1920. the
protein intake was raised to 95 gm.

T.\BLE 9. — Urine Nitrogen Partition (Case 8)

Day

UreaN

Ammonia N

Uric Acid N

Creatinin N

Creatin N

Remarks

Gm.

Gm. %

Gm. %

' Gm. %'

Gm. %

Gm. %

5/30

10.12

7.60 75.1

0.693 6.84

0.289 2.85

0.370 3.66

0.050 049

9.03

7.26 80.3

0.619 6.86

0.217 2.41

0.396 4.39

0.027 0.29

8.22 79.1

0.494 4.75

0.200 1.92

0.363 3.49

0.063 0.61

Rf

13.16

11.21 83.2

0.686 5.21

0.234 1.78

0.390 2.96

0.068 0.44

Protein 95 gm.

14.20

11.94 84.1

0.661 4.66

0.228 1.61

0.399 2.81

0.078 O-.W

Protein 95 gm.

6.34 Sl.l

0.608 7.78

0.212 2.71

0.375 4.80

0.051 0.77

fi'ft

0.556 7.36

0.221 2.92

0.402 8.31

0.139 1.S5

fi/10

7.73

6.40 82.8

0.601 7.78

0.238 3.08

0.428 5.60

0.000 0.00

Case 9. — Progressive pseudohypertrophic muscular dystrophy (juvenile form.)

O. Sm.. white, male, 38 years of age, weight 50 kg. The condition was first
noticed when the patient was 18 years old ; enlargement of the calves and
beginning atrophy of the muscles from the waist down were then noted.

There was general weakness, pronounced lordosis and enlarged calf mus-
cles as compared' with the remaining musculature. The trunk, shoulder and
arm muscles were small.

The patient was transferred to the metabolism unit May 17, 1920. He was
placed on a diet of 79 gm. protein. 100 gm. fat and 220 gm. carbohydrate, and
equivalent to 2.172 calories. The metabolism was not sufficiently stabilized
for a satisfactory creatin ingestion observation. (A part of the data obtained
on this case will be used in another paper, and is not included here.)

TABLE 10.— Urine Nitrogen Partition (Case 9)

Urea N Ammonia N

Day Total N , — ' — , , > ,

Gm. Gm. % Gm. %
5/18 8.61 7.32 85.0 0.771 8.95
5/19 6.63 5.54 83.5 0.546 8.23
5/23 8.40 ,

GIBSOX-MARTIX-BUELL— MUSCULAR DYSTROPHY 95

C.\SE 10. — Myasthenia gravis.

Geo. S., white, male. 51 years of age, weight 80 kg. The patient was
admitted Sept. 4, 1919. complaining of increasing weakness of the arms and
legs over a five year period.

There was bilateral external ophthalmoplegia and bilateral ptosis. Sensa-
tion of all types was normal. There was no ataxia of the upper or lower
extremities. Knee jerk and plantar flexion were equal on both sides. Faradic
stimulation of the anterior tibial group of muscles gave a poor response.
Intermittent faradism induced fatigue of the flexors of the forearm, but the
myasthenic reaction failed for the finger flexors.

The patient was in the metabolism unit for the period September 30 to
October 4 only. The diet then given consisted of 51 gm. protein, 90 gm. fat,
and 300 gm. carbohydrate, or 2,275 calories. A nonmeat and nonsoup diet was
given September 5, and a general diet on September 24.

TABLE 11.— Urine Nitrogen Partition (Case 10)

Total S,
Gm.
12.23
8.79
6.14
5.80

.\mmonla S

Uric Acid X

Creatinin N

Creatin X

Day

9/ 5
9 24
10/1
10 3

Gm.

0.422
0.398

6.87
6.66

Gm.
0.112
0.199
0.101
0.078

%
0.92
2.26
1.64
1.34

Gm.
0.543
0.579
0.564
0.564

%
4.44
6.58
9.18
9.72

0.000
0.000
0000
0.000

%

0.00

o.oo

0.00
0.00

§='Sfet^'^'

Case 11. — Acute anterior poliomyelitis.

L. W., white, male, 25 years of age, weight 63 kg. The patient entered the
hospital Sept. 2. 1920. complaining of inability to walk without aid. The con-
dition developed just previous to admission.

The arms were normal in strength with no tremor or ataxia. The legs
were held in normal position in the bed, and the muscle tonus at rest was
normal. The patient was unable to raise any part of the right leg; he flexed
the left knee SO per cent., but could not raise the foot. Faradic stimulation
showed the least response over the right buttock and thigh. There was a
25 per cent, response on the right calf and a 5 per cent, response on the right
anterior tibial group. Sensation was normal.

While in the metabolism unit, the patient was given 75 gm. protein and from
2.000 to 2,500 calories daily.

T.'KBLE 12. — Urine Nitrogen Partition (Case 11)

Total N,

14.43
12.63
12.35
12.05
12.45

UrcaN
Gm. %
10.72 74.3
9.13 72 3
9.28 75.1
9.46 78.5
10.57 84.9

Ammonia N

Uric Acid N

Creatinin N

Gm. % ■
0.460 3.32
0.478 3.78

0.437 8.61

Creatin N

Gm. %
0.055 0.3S
0.074 0..'»
0.024 0.19

t^ 6:1?

Day
10/3
10/4
10/S
10/6
10/7

'Gm. %'
0.690 4.78
0.667 5.28
0.717 5.81
0.616 6.11
0.787 6.82

Gm. %
0.154 1.07
0.134 1.06
0.123 0.98
0.139 1.15
0.126 1.01

Remarks
1 gm. creatin

Case 12. — Myositis ossificans; chronic endocarditis, mitral stenosis and
cardiac decompensation.

Geo. F., white, male, 26 years of age, weight 54 kg. The patient was
admitted to this hospital April 13. 1920. The duration of his illness was given
as 18 months. It started with rheumatic symptoms.

The muscles of the trunk had a normal feel. The joints, except the hips
and ankles were stiff and swollen, movements being restricted. The deltoid
muscles seemed atrophied ; the biceps and triceps were small. The muscles
of the forearms had a bony feel. The thigh and calf muscles seemed hard and
fibrous throughout.

96 ARCHIVES OF IXTERXAL MEDICIXE

The patient was transferred to the metabolism unit. He was placed on a
nonpurin and noncreatin diet of 75 gm. protein, 72 gm. fat and 243 gm. car-
bohydrate, a total of 1,926 calories.

TABLE 13.— Urine Nitrogex Partition (Case 12)

Vol- Total
ume, Speciflc N,
Day C.c. Gravity Gm.

UreaN

Ammonia N

Uric Acid N
'Gm.' %

Creatinin N
Gm. %

Creatin X

Gm., %

Gm. %

'Gm. %

4/17 575 1.026 8.45
4/18 600 1.030 8.99
4/19 700 1.029 9.41
Balance period April 20-24 inc
Average per day... . 9.62

6.51 77.0 ■
7.20 80.1

7.27 77.2

0.488 5.77
0.567 5.25
0.599 6.37

0.302

3.21
3.45

0.283
0.283

IS

3.01
2.94

0.000 0.00
0.000 0.00
O.OCO 0.00

X Balance

per Day,

Gm.

Ca Balance

for Period,

Gm.

fo^rlS
Gm.

Food

Crine

Stools

Total

12.97
9.62
1.02
10.69

-f2.2S

5.838
0.105
4.597
4.702

+1.136

0.

2.147
197

— 1.6S6

+0.461

Case 13. — Progressive pseudohypertrophic muscular dystrophy.

C. O., white, male, 10 years of age, weight 25 kg. The patient was referred
to the medical service June IS, 1921.

The patient presented the characteristic picture of progressive pseudo-
hypertrophic muscular dystrophy. There was lordosis and the patient walked
with a waddling gait. When placed in a lying position on the floor and
instructed to arise, he rolled on the side and pushed himself to a sitting posi-
tion, but could not get up. He would stand with his feet widely apart in
the characteristic lordotic position. The calf muscles were markedly hyper-
trophied.

TABLE 14.— Urine Nitrogen Partition (Case 13)

Creatinin X

Total N,
Gm.

Day

Gm.

%

6/15

1.52

6/16

4.30

0.055

4.56

0.057

1.25

6/18

4.13

0.065

1.57

> gm. creiitin by mouth

The diet given consisted of 46 gm. protein, 70 gm. fat and 175 gm. carbo-
hydrate, the total calories being 1,564.

THE XITROGEN REQUIREMENT FOR MAIN-
TENANCE IN DIABETES AIELLITUS *

PHIL L. MARSH, AI.D.; L. H. NEWBURGH, M.D.,

AND

L. E. HOLLY, M.D.

ANN ARBOR, MICH.

The general acceptance of the principle of restriction of the total
caloric intake in the dietetic management of diabetes mellitus in con-
trast to the older principle of overfeeding has increased the importance
of an accurate knowledge of the minimum amount of protein that will
maintain nitrogenous equilibrium in the diabetic patient. The ultimate
effect of long continued gradual loss of body nitrogen is not known,
but it seems probable that such a condition is very undesirable. The
subject whose nitrogen excretion is constantly higher than his nitrogen
ingestion is certainly suffering for want of one of the most important
of the tissue repairing elements, and' a diet so arranged as to induce
this negative balance, even though not lethal, must produce a severe
grade of inanition.

The following experiments were undertaken in an effort to deter-
mine the minimum protein ingestion that will safely maintain nitrogen
balance in patients with diabetes mellitus.

1. The usual method of determining the state of the nitrogen equi-
librium was followed, namely, the balancing of the nitrogen in the food
against the nitrogen in the urine and stools. Because of the difficulties
in the way of collecting specimens from women, male patients were
used. Their ages vary from 18 to 80 years. The severity of the dis-
ease also varied, as can be seen from the protocols, but over half were
of the more severe types. In general, we chose the more intelligent
patients as the ones most likely to cooperate with us by strict adherence
to diet and in the collection of specimens, and care was taken to impress
each of them with the importance of this cooperation.

The diets were arranged by an expert dietitian, and their delivery
to the patients was carefully supervised by an unusually competent
nurse. Copies of the diet lists were delivered to us daily, and were
frequently checked against the ward reports. Uneaten portions were
measured and deducted. The nitrogen content was determined by
dividing the protein by 6.25, and the protein was estimated by use of

♦From the Department of Internal Merlicine. Medical School, University
if Michigan.

98 ARCHIVES OF IKTERXAL MEDICIXE

Atwater and Bryant's food tables.' The heat value of 1 gm. fat was
considered 9 calories and of 1 gm. protein or carbohydrate as 4 calories.

The nitrogen determinations on the stools, and on the urines of
about half the cases were made by the Kjeldahl method, and on the
remaining urines by the Folio micro-Kjeldahl method, after the two
methods had been checked against each other. In some cases the
quantitative nitrogen determinations were made daily; in the others
aliquot parts of the acidified daily specimens were used. The stools
were collected for several days, and their nitrogen content divided
equally among the days during which they were passed. In a few
cases, where nitrogen determinations on the stool are not available, the
nitrogen excreted by this route has been estimated as 0.75 gm. daily.
Since the nitrogen of the stool is only in small part derived from the
unresorbed products of protein digestion, and since it is not directly
dependent for its quantity on the food nitrogen, this is thought more
accurate than the estimation of Mosenthal and Harrop- that 10 per
cent, of the ingested nitrogen is excreted in the stool. Rieder ^ found
a fecal excretion of 0.54, 0.87 and 0.78 gm. nitrogen daily from a man
fed on a protein-free diet. The average fecal nitrogen of those cases
in our series who were in balance on small amounts of ingested nitro-
gen, and who were not undergoing active catharsis was about 0.77 gm.
daily.

In most cases, the nitrogen studies were incidental to the treatment,
and in only a few was an effort made to find the lowest level at which
nitrogen balance could be established. It is probable that had such
an effort been made it would have given us lower figures for some of
the patients. The treatment used was the low protein, high fat, low
carbohydrate diets previously described by Newburgh and Marsh.*
These diets were arranged with the intention of satisfying the caloric
needs.

REPORT OF CASES

Case 1 (19-293).— Patient was an American .school teacher, aged 56, who
entered the hospital, June 9, 1919, complaining of weakness. There was no
history of dialietes in the family. The symptoms appeared five years hefore,
with polyuria, polyphagia and polydipsia. There had been gradually increas-
ing weakness, with obstinate constipation and mental depression. Nothing of

1. Atwater and Bryant: The Chemical Composition of American Food
Materials, U. S. Dept. Agriculture, Bull. No. 28, 1906.

2. Mosenthal, H. O., and Harrop, G. A.: The Comparative Food Value of
Protein, Fat and Alcohol in Diabetes Mellitus as Measured bv the Nitrogen
Equilibrium, Tr. Assn. Am. Phys. 33:302, 1918; also Arch. Int. Med. 22:750
(Nov.) 1918.

3. Rieder, H. : Bestimmung der Menge des im Kothe befindlichen, nicht von
der Nahrung hcrruhrcnden Stickstoflfes, Ztschr. f. Biol., N. F. 2:378, 1884.

4. Newburgh. L. H.. and Marsh. P. L. : The Use nf a High Fat Diet in the
Treatment of Diabetes Mellitus, Arch. Int. Med. 26:647 (Dec.l 1920.

MARSH-XEIVBURGH-HOLLY— DIABETES MELLITUS

99

importance was found on physical examination. His blood sugar was 0.55 per
cent., and his urine contained 50 gm. glucose per 1,000 c.c.

On a diet containing 16 gm. protein, 100 gm. fat and 10 gm. carbohydrate,
with a total of about 1,000 calories, his urine became free of sugar on the
fourth day, his ferric chlorid reaction became negative on the fifth day and
his blood sugar had fallen to 0.14 per cent, on the sixth day. On an oatmeal
and butter diet, which was given him a few days later, and which contained
9 gm. protein, 155 gm. fat and 31 gm. carbohydrate, a total of 1,550 calories,
his blood sugar rose to Q2i per cent, and glycosuria returned. .\ day of
starvation sufficed to clear his urine of sugar. The nitrogen determinations

TABLE 1. — P.\RT OF Record of C.vse 1

Food

Urine
, , ' , Blood

Pro-
tein,
Gm.
16.0
16.4

Carbohy-
Fat, drate.
Gm. Gm.
9T.4 9.1
101.5 6.3

Calo-
ries
975
1.0O5

Date

1919

7/1

2

cose, Dia- per Urine,"

Gm. cetle Cent. Gm.

0 -f .... 5.74

0 0 .... 5.74

16.6 96.8

5.74 0.60 2.66 6.24 —3.58
5.74 0..T0 2.45 8.24 — 3.T9
5.74 0.50 2.48 6.24 —3.76

Patient away from hospital for two weeks but did

0.50
0.50
0.50

2.77
2.74
2.74

6.24
6.24
6.24

lOt adhere to diet

6.73
0.73
0.73

2.61
2.61
2.61

6.87
6.87
687

6.14 0.73 2.61
6.14 0.73 2.61

125.4

6.5

1,225

8.2

1,575

157.0

8.0

1,580

157.0

7.5

1,575

157.0

7.5

1,575

163.0

7.5

1,645

168.4

6.9

1,680

lfiR.4

6.9

1,680

1,680

166.4

6.9

1.880

168.4

6.9

1,680

168.4

1,680

1,680

168.4

6.9

1,680

168.4

6.9

1.680

5.,15

0.44

5.35

0.44

0.40

5.49

0.40

5.49

0.40

5.49

0.40

5.49

,1.49

0.40

6.49

0.40

5.49

5.49 5.63

were begun a week later. It will be seen from Table 1 that a diet containing
34 gm. protein and 1,680 calories, with only 7 gm. carbohydrate practically
established nitrogen equilibrium. This represents 0.54 gm. protein and 26 cal-
ories per kilogram of body weight.

C.\SE 2 (19-306).— Patient was a very mild diabetic, an American, 66 years
of age, attendant at a state hospital for the insane. His symptoms were few,
beginning three years before with thirst and polyuria. He had been on no
diet before entrance to the hospital, June 30, 1919, except that he avoided
"sweets." Two days of a low caloric diet sufficed to render his urine sugar
free, and his blood sugar was found to be normal on the tenth day. No dif-
ficulty was met in establishing nitrogen equilibrium with an intake of two-

100 ARCHIVES OF IXTERXAL MEDICIXE

thirds gram of protein and from 33 to 40 calories per kilogram of body weight.
Part of his record is presented in Table 2.

Case 3 '(19-391). ^Patient was an American factory inspector, who entered
the surgical clinic for a herniotomy, and to obtain treatment for an infected
leg. There was no diabetes in the family, and the patient had been well until
six or eight years ago, weighing about 215 pounds. During the past few
years he had noticed that he drank unusually large quantities of water and
that he urinated frequently and copiously. His weight gradually fell to about
135 pounds, but he felt well and worked every day. Three weeks before
entrance he injured his right leg. The wound became infected, was drained,
but healed very slowly, and at the time of entrance was still draining pus.
Except for the emaciation and dry skin, nothing of importance was noted in
the examination. Throughout his stay in the medical ward he was afebrile.

The data from his study are presented in Table 3. The nitrogen studies
were started after his blood sugar had been brought to normal by a diet con-
taining 16 gm. protein, 100 gm. fat, 10 gm. carbohydrate and about 1,000
calories. It will be noted that on this diet his nitrogen output was large,

TABLE 2. — Part of Rfxdrd from Case 2

)ate tein. Fat, drate, Calo- Amount, cose, Dia- N,* per Stool, In, Out, ance, Wt.,

1919 Gm. Gin. Gm. rios. C.c. Gm. cetic Gm. Cent. Gm. Gm. Gm. Gm. l.bs.

723 16.3 97.4 9.7 980 1,610 0 0 4.19 0.14 O.M) 2 61 4.79 —2.18 141

24 16.3 97.4 9.7 980 820 0 0 4.19 .... OM 2.61 4.79 —2.18 ...

25 36.2 219.2 11.4 2,165 600 0 0 4.19 .... 0.60 5.79 4.79 +1.00 ...

26 36.2 219.2 11.4 2,165 825 0 0 4.19 .... O.CO .^.79 4.79 +1.0O ...

27 36.2 219.2 11.4 2,165 710 0 0 4.19 .... 0.60 5.79 4.79 +1.00 ...

28 36.2 219.2 11.4 2,165 725 0 0 5.61 0.12 0.65 5.79 6.16 — 0..37 ...

29 35.8 219.1 10.6 2,155 1,255 0 0 5.61 .... 0.&) 5,79 6.16 —0.37 141

30 36.2 219.2 11.4 2,165 970 0 0 5.61 .... 0..i5 !-..79 6.16 —0.37 ...

31 36.2 219.2 11.4 2,165 926 0 0 5.61 .... 0,55 5.79 6.16 —0.37 ...

8/1 41.5 243.8 14.8 2,420 1,360 0 0 5.61 .... 0.:.5
2 41.5 243.8 -----

64 6.16 +0.48

57 6.16 +1.41

222,6 13.2 2,240 990 0 0 4.80 .... 1.28 7.33 6.08 +1.25

268.8 13.2 2,560 1,100 0 0 4.60 .... 1.28 7.37 6.08 +1.29

23i.7 13,2 2,300 1,000 0 0 4.80 ,,,. 1.2S 5,39 6,08 —0 69

228,8 13,2 2,290 1,190 0 0 4,80 .... 1.28 7.37 6.0S +1,29

228.8 13,2 2,290 1,475 0 0 4.80 0.10 1.28 7,37 6,08 +1.29

amounting to nearly 14 gm. daily. With the addition of moderate amounts
of protein and large quantities of fat to his diet, the nitrogen elimination
gradually decreased until it reached about 8 gm. a day. This very mild
diabetic with a body weight of 60 kilograms did not establish nitrogenous
equilibrium on 40 gm. protein, during the few days allowed. Two explanations
may be offered: First, it may have been that not enough time was allowed;
this factor will be discussed later; or, second, that there was an increased
destruction of protein as a result of his chronic infection. Balance was readily
established by a diet containing 0.90 gm. protein and 42 calories per kilogram
of body weight.

Case 4 (19-444). — Patient was an American farmer, 18 years old, who
entered the hospital Sept. 5, 1919, complaining of weakness and excessive
thirst. On one occasion, at the age of 14, he vomited a large quantity of
blood. He noticed his polyuria and thirst a year and a half before; this was
associated with increasing weakness. A physician found sugar in his urine,
and under treatment the patient remained sugar-free for four months. After
this he returned to an ordinary diet, and his weakness gradually increased.
His best weight had been 135 pounds, two years before, and his weight at
entrance was 95 pounds. For a year lie had had crop after crop of boils.

MARSH-XEIVBURGH-HOLLY— DIABETES MELLITUS 101

Examination showed an emaciated young man with several boils on his
hands and one on the left temporal region. The tendon reflexes were obtained
only on reinforcement. The blood sugar was 0.525 per cent.

On a diet containing 16 gm. protein, 100 gm. fat and 10 gm. carliohydrate,
with a total of 1,000 calories, his blood sugar fell to 0.20; at this time he
left the hospital without permission, and ate a large quantity of carbohydrate.
It was not until September 22 that his urine became free from sugar, and on
September 25 his blood sugar was 0.15 per cent. A diet containing 28 gm.
protein and 1.600 calories, representing 0.67 gm. protein and 38 calories per
kilogram of body weight was found to establish, nitrogen balance. Data from
his case are presented in Table 4.

Case 5 (19-537). — Patient, an American mailcarrier, 21 years of age. entered
the hospital, Oct. 22, 1919, complaining of loss of strength. He had always
been in excellent health until the onset of his diabetes a year before. His

TABLE 3. — P.vRT OF Record from C.\sf, 3

Carbohy- Glu- Sugar.

;, drate, Calo- Amount, cose, Wa- per Urine,* Stool.f In, Out,
I. Gm. ries C.c. Gm. cetic Cent. Gm. Gm. Gm. Gm.

128.0

27 29.6 150.0

28 30.1 128.0

29.8

40.3 219.0
40.6 234.0

228.0

240.0
240.0
240.0

970
970

1.775
2,272

970

1.315
1,450 •
1,500
1,315

2,350
2,400
3,470
4,825

1,400

2,270

1,125
1,660
l.i.V)

2,260

2,430
2,430
2,430

1,830
1.825
1,900

51.1 240.0 13.4 2,430 7.01 1.11 8.66 8.12 +0.54

only symptoms have been extreme weakness, a ravenous appetite, and a mod-
erate polyuria. His weight fell from 165 to 119 pounds. Soon after the
beginning of the diabetes he was placed by his physician on a diet consist-
ing chiefly of milk, eggs, bran flour and fruit. The symptoms had become
increasingly more severe. Examination showed a moderate emaciation, dry
skin and absent knee reflexes.

On a diet containing 22 gm. protein, 110 gm. fat and 10 gm. carbohydrate,
a total of 1,100 calories, he was sugar-free on the ninth day, though it was
nearly a month before his blood sugar was found to be normal. It will be
noted from the table that with a protein intake of 37.8 gm. and a total intake
of 1.380 calories, he .showed a negative balance of 2.11 and 2.83 gm. nitrogen,
respectively, during the two periods in which the determinations were made,
but that this fell to practical equilibrium when the total calories were raised
to 1,650, though the protein and carbohydrate in the diet were unchanged. The
diet contained 0.76 gm. protein and 36 calories per kilogram of body weight.

102

ARCHIVES OF IXTERXAL MEDICIXE

External circumstances made it necessary to discontinue the determinations
before we were entirely finished. The large nitrogen content of the stool is
explained by the fact that the patient took one-half ounce of magnesium
sulphate daily.

TABLE 4. — P.\RT OF Record from Case 4

Diet

Crine

Nitrogen

Pro-

Carboby

Gln-

Sugar.

Bal-

Date

tein,

Fat, drate.

Calo- Amoun

, cose,

Dia

tier Urine.

StooI,t In

Out,

Wt.,

1919

Gm.

Gm. Gm.

ries

C.C.

Gm.

cetic Cent. Gm.

Gm. Gm

. Gm.

Gm.

Lbs.

9m

24.9

10/ 1

24.9

141.4 9.9

1.410

2.500

n

S5

24.9

141.4 9.9

1,410

2,2.W

0

24.9

24.9

141.4 9.9

1,410

1,510

0

85

24.9

141.4 9.9

1,410

2,f««>

0

PB

24.9

2.(1«B

24.9

141.4 9.9

1.410

1,200

0

0.15

24.9

24.9

141.4

9.9

1,410

24.9

141.4

9.9

1,410

.16.8

192.9

M.H

192.9

9.9

1,925

36.8

192.9

9.9

1,925

14 36.8 192.S

17 36.8 192.9

n 0.94 S.98

to 36.S 192.9

192.9

9.9

1,925

1,055

192.9

9.9

1,925

1,115

192.9

9.9

1,925

1,420

192.9

9.9

1,925

166.2

9.9

1,685

166.2

1.685

166.2

9.9

1,685

1.535

166.2

9.9

1.685

1,500

166.2

1.685

166.2

9.9

1,685

1,-10

166.2

9.9

1.685

2,4T0

166.2

9.9

1.685

166.2

9.9

1,685

166.2

9.9

1.685

2.420

166.2

9.9

1.685

l.TOO

166.2

9.9

1,685

162.8

9.9

1,615

i.aw

162.8

9.9

1.615

2,120

162.8

9.9

1.615

1,800

162.8

9.9

1.615

1,920

162.8

9.9

1,615

1628

9.9

1,615

162.8

9.9

1,615

2,235

1(52.8

9.9

1,615

2,365

1S2.8

9.9

1,615

2,425

162.8

9.9

1,615

1.950

1.91 5.89 5.77 +0.12

2.97 0.45 5.S9 3.42

2.97 0.45 4.49

age 28.1 162.8

.15 4.49 2.36 +2.13

Case 6 (19-567). — Patient was a German coal dealer, 49 years of age. who
entered the hospital, Nov. 11, 1919, complaining of weakness and polyuria.
There was no history of diabetes in the family. He had always eaten heartily,
but did not care for sweets and pastries. His symptoms first appeared eight

MARSH-XEIVBL'RGH-HOLL V— DIABETES MELLITUS

103

weeks ago. with increased thirst and polyuria, progressive weakness and loss
of weight, the latter amounting to 30 pounds. A physician placed him on a
diet consisting of eight slices of gluten bread, and half a dozen eggs daily,
with all the meat he wanted. A second physician gave him nothing but milk
and oatmeal. He noticed no improvement from either. There has been some
numbness of the hands and feet, some constipation and no visual disturbances.
Examination showed fair nourishment, with a dry skin. The tonsils were
septic and the teeth in poor condition. There was moderate arteriosclerosis;
the tendon reflexes were active.

T.\BLE 5. — Part of Rfxord from Casf, 5

Pro-

Carbohy

Glu-

— ,

tein,

drate.

(Calo-

Pia-

l!»li»

Gm.

Gm.

Gm.

ries

C.c. Gm.

cetic

0/2S

a. 8

109.2

9.9

1.110

4.000 -H-H-f

-1-

'»

21.6

109.2

9.9

1.110

3,300 -t- + -|-

109.2

9.9

26

21.6

109.2

9.9

1.110

1,800 + +

+

•»7

21.6

87.8

8.5

900

1.630 -1-

+

Sugar.

per Urine,* Stool,t In, Out,
Cent. Gm. Gm. Gm. Gm.

1,380
1,380
1,380

2,860
2.82S
2.170
2,500

12 S7.4 133.7

13 37.4 133.7

14 37.4 133.7

ver-

ige 37.4 133.7

2,750
3,.';90
3.300

1.35 5.99 8.10

19 37.6 147.6

22

37.6

165.6

n

37.8

1H2.0

24

i?7.6

163.6

163,4

?B

37.8

163.6

27

87 Ji

163.6

iver-

age

37.7

ies.7

1,650
1.650
1,650

8.3 1,675 3,070

3.920
3.955
4.380

On the fifth day of a diet containing 16 gm. protein, 100 gm. fat and 10 gm.
carbohydrate, a total of 1,000 calories, his urine was free from sugar, and on
the tenth day the ferric chlorid reaction became negative. He was given a
diet containing 65 gm. protein and 2.065 calories, and was able to add 2.6 gm.
nitrogen to his body daily for several days. .After a few days, his nitrogen
output increased, and he was practically in balance. Part of his record is
presented in Table 6.

104 ARCHIVES OF IXTERXAL MEDICINE

Case 7 (20-18).— Patient, a locomotive fireman, 22 years of age, first entered
the hospital, Jan. 16, 1920, complaining of excessive thirst and polyuria. There
was no history of diabetes in the family. His symptoms were first noticed
about six weeks before, and became gradually more severe, until he consulted
a physician about January 1. The diagnosis of diabetes mellitus was made,
and he was put on a diet of milk. His average weight was 130 pounds, and
at admission it was 113 pounds. The physical examination showed nothing
of importance.

On a diet containing 17 gm. protein, 100 gm. fat and 10 gm. carbohydrate,
with a total of about 1,000 calories, his urine rapidly became sugar-free. His
diet was increased gradually to 2.000 calories, with 40 gm. protein and 25 gm.
carbohydrate, the remainder being fat. Through the summer, he reported to
us at intervals, and his urine was always found to be sugar-free and his blood
sugar within normal limits. He returned to the hospital in October, 1920, at

TABLE 6.— Part of Record from Case 6

Diet Urine Xitrog-'n

, ' , , * , Blood , ' ,

Pro- Carbohy- Glu- Sugar. Bal-

Date tein. Fat, drate, Calo- Amount, cose, Dia- per Urine,* Stool,t In, Out, ance, Wt..

1919 Gm. Gm. Gm. ries C.c. Gm. cetic Cent. Gm. Gra. Gm. Gm. Gm. Lbs.

11/22 15.9 97.2 9.9 9S0 1,800 0 0 0.12 153

23 15.9 97.2 9.9 980 2,630 0 0

24 15.9 97.2 9.9 980 2.250 0 0

25 15.9 97.2 9.9 980 1,450 0 0

26 15.9 97.2 9.9 980 1.5,TO 0 0

27 15.9 97.2 9.9 980 1,735 0 0

age 15.9 97.2 9.9 980 6.02 0.53 2.54 6.55 —4.01

11/28

to 65.0 197.0 9.9 2,065 0 0 n.lO

3 65.0 197.0 9.9 2.065 2,450 0 0 0.07 156

4 6.5.0 197.0 9.9 2,065 1,800 0 0

5 6,5.0 197.0 9.9 2,065 1,830 0 0 ...

6 65.0 197.0 9.9 2,065 1,950 0 0 156

7 65.0 197.0 9.9 2.065 2,440 0 0 0.10 ^^ •.•

age" esio 197^0 9.9 2,065 6.74 1.07 10.40 7.81 4-2.!i9

8-11 65.0 197.0 9.9 2,065 0 0 156

12 65.0 197.0 9.9 2,065 2,195 0 0 156

13 65.0 197.0 9.9 2.0<» 2,535 0 0 •■;

14 65.0 197.0 9.9 2.065 1,940 0 0 loo

15 65.0 197.0 9.9 2,065 1,760 0 0

16 65.0 197.0 9.9 2,065 2,030 0 0 0.09 ...

17 66.0 197.0 9.9 2,065 2,365 0 0 .... ^^ ^^ ^^ ^^ loo

^age' esio IffTo M 2^065 TTT TT r TT 8.48 1.39 10.40 9.87 +0..^3

♦ Daily average of each period obtained from aliquot parts ol urines,
t Daily average ol each period obtained Irom all stools.

our request, so that further adjustment of his diet might be made. It was
during this period that the studies in nitrogen balance were made, the data
for which will be found in Table 7.

This patient was the subject of an interesting experiment that will be
discussed later. Nitrogen balance was established on a diet containing 0.58
gm. protein and 21 calories per kilogram body weight.

Case 8 (20-558). — Patient was a Syrian laborer, 36 years of age, who
entered the hospital, July 25, 1920, complaining of excessive thirst and appetite,
and polyuria. His symptoms were dated back eight months, dtiring which
period he had lost 35 pounds. He was a moderately severe diabetic who
was able to tolerate SO gm. carbohydrate and 30 gm. protein without glycosuria.
His blood sugar was 0.37 per cent, at entrance, and his weight 162 pounds.

He was immediately placed on a diet consisting of 20 gm. protein, 90 gm.
fat and 12 gm. carbohydrate, totaling about 950 calories. After three days.

MARSH-XEIIBURGH-HOLLY— DIABETES MELLITUS 105

his urinary sugar disappeared without the presence of a positive ferric chlorid
test, and his blood sugar fell to 0.124 per cent. He remained on this diet
until July 31. During the week following his carbohydrate intake was grad-
ually increased each day, until August 7, he was eating 51.7 gm. without
glycosuria. From August 8 to 12 he received 28 gm. protein, 115 gm. fat
and 20 gm. carbohydrate with an energy value of about 1.225 calories. In
the period from August 13 to 19 his average daily diet was 34 gm. protein,
165 gm. fat, 26 gm. carbohydrate and 1,725 calories. During this lime his
weight fell to 156 pounds. At this point the nitrogen determinations were
started, the data for which are presented in Table 8. On the basis of a weight
of 69 kg., nitrogen balance was established on 0.81 gm. protein and Zi calories
per kilogram of body weight. There was a large, positive balance at first
which gradually fell, though there was a slight increase in the protein and
caloric intake.

TABLE 7.— Part of Record from Case 7

Diet

Urine

Nitrogen

Pro-

C

arboby

Glu-

Sugar.

Bal-

Date

tein.

Fat,

drate."^

Calo- Amount

, cose.

Dia-

per

Urine,'

Stool,!

\ In, Out,

ance,

Wt.,

1920

Gm.

Gm.

Gm.

ries

C.c.

Gm.

cetic Cent.

Gm.

Gm.

Gm. Gm.

Gm.

Lbs.

11/12

11.1

45.4

9.2

490

0

0.16

13

21.5

80.4

14.6

870

0

i28

14

20.6

87.6

14.6

930

0

6!i6

....

15

195.7

115.9

15.0

1,885

0

16

196.5

96.0

15.0

1,710

1,350

g.Vis

i.42

31.44 li.'oO +20.44

130

17

191.6

107.3

15.0

1,790

2.150

0

24.94

1.42

30.fi6 26.36

+4.30

18

197.9

104.5

14.0

1,790

2,450

14.68

1.42

31.66 16.10 +15.56

19

191.6

107.3

15.0

1,790

6;i2

30.66 ....

20

146.6

80.6

14.0

1,370

1,866

0

18.38

i.«7

23.46 20.05

+3.41

130

21

32.1

121.8

23.9

1,320

2,100

0

29.76

1.67

5.13 31.43

—26.30

22

33.6

169,4

27.5

1,770

1,500

19.38

1.67

5.37 20.85

-15.48

23

34.3

174.7

29.9

950

0

7.7S

167

5.49 9.43

—4 06

127

24

33.8

153.6

28.5

l!630

875

0

0.08

0.75:

5.25 8 05

-3.E0

25

28.7

116.8

20.8

1,250

1,225

0

9!37

0.75

4.59 10.12

-5.53

26

33.5

168.8

26.3

1,670

9.T0

0

7.85

0.75

5.40 8.60

—3.20

27

33.0

89.7

21.9

1,030

1,350

0

2.35

0.76

5.27 3.10

+2.17

127

28

.33 1

159.1

29.2

1,680

1:200

0

o.'ii

2.42

0.75

5.29 3.17

+2.12

29

34.8

167.1

25.0

1,745

1,250

0

3.05

0.76

6.57 3.80

+1.77

30

32.3

149.8

24.2

1.540

950

0

8.IV3

0.75

6.17 9.28

—4.11

i28

31

32.9

169.3

25.9

1,760

1,150

0

8.33

0.75

5.27 9.08

—3.81

1

34.9

166.0

25.9

1,735

960

0

4.30

0.75

5.59 5.05

+0.54

21

82.8

153.7

28.2

1,625

1,900

0

o.oe

15.05

0.76

5.25 15.80

—10.55

3

211.6

37.5

2.270

1,200

0

7.06

0.75

8.56 8.71

-0.15

130

586

202.0

31.3

2,180

1,275

0

0

7.92

0.75

9.37 8.(17

+0.70

5

68.8

213.4

39.4

2,315

1,250

0

9.12

0.76

9.40 9.87

-0,47

Jl

Case 9 (20-677).— Patient was a German pattern maker, 63 years of age,
who entered the hospital, Oct. 19, 1920, complaining of failing vision, polyuria,
polyphagia and polydipsia. There was no family history of diabetes mellitus.
The diabetic symptoms were first noticed about three years before, and though
he was placed on a diet containing little carbohydrate, he continued to lose
strength and weight, until on admission to the hospital he was hardly able to
get about the house. His best weight was 150 pounds twenty years before.
In the past two years he had lost 25 pounds, reaching his entrance weight of
111 pounds. Examination showed a moderate degree of emaciation. There
was a mature cataract in the left eye and a beginning cataract in the right.
The peripheral arteries were markedly sclerotic. There was some edema of
the ankles and the tendon reflexes were normal.

On a diet containing 18 gm. protein, 90 gm. fat and 12 gm. carbohydrate,
his urine became free from sugar on the twenty-fourth day, and four days
later his blood sugar was 0.11 per cent. His diet was increased until he was

106 ARCHIVES OF IKTERXAL MEDICIXE

getting 34 gm. protein, 170 gm. fat and 23 gm. carbohydrate, a total of about
1,760 calories. December 3, he was transferred to the department of ophthal-
mology for operation, and was returned to the medical ward December 23
with a moderate glycosuria and a blood sugar of 0.17 per cent. Three days
of a diet similar to that which he received immediately after admission suf-
ficed to render his urine sugar-free, and his blood sugar fell to 0.14 per cent.
His diet was again increased, and it was found that 55 gm. protein, 210 gm.
fat and 35 gm. carbohydrate, with a total of about 2,250 calories, kept the
sugar content of his blood between 0.19 and 020 per cent, though there was
no glycosuria. His diet was accordingly reduced to 32 gm. protein, 160 gm.
fat and 24 gm. carbohydrate, containing about 1,650 calories. On this diet

-Part of Record from Case 8

Drine

Carbo- GIu-

lydrate, cose, Dia- Urine,* Stool,t

Gm. Calories Gm. cetic Gm. Gm.

2,115
1,850

1R4.6
159.1
134.9

206.9 29.8 2,240
207.6 37.0 2,230
209.0 39.3 2,270

207.8

249.2
234.2

2,290
2,370
2,275
2,230

his blood sugar remained 0.15 per cent., and, as is shown in Table 9, he \va.> in
nitrogen balance. With a body weight of 50 kg., he was receiving on this
diet about 0.64 gm. protein and about 33 calories per kilogram of body weight.
Ca.se 10 (20-615).— Patient, a very mild, elderly diabetic, returned to the
hospital two years after he had been discharged on a moderately restricted
diet. In the interval he had had no symptoms, except those of an associated
chronic bronchitis, and he returned for examination. At entrance, his weight
was 200 pounds, his age, 80 year.s, and his blood sugar 0.15 per cent. At no
time during his stay did he have glycosuria or a positive ferric chlorid test
on his urine. He was placed immediately on a diet containing about 1.500
calorics of which about .34 gm. were protein. This was rapidly increased to
a very liberal diet, as shown in Table 10.

MARSH-NEWBURGH-HOLLY— DIABETES MELLITUS 107

Case 11 (21-276).— Patient was an American farmer, 22 years old, who
entered the hospital complaining of polyuria, weakness and loss of weight. A
brother died of diabetes mellitus at the age of 17. His symptoms started seven
months before and the diagnosis was made immediately. His diet was only
moderately restricted, however, and he constantly lost weight and strength.
His best weight had been 164 pounds just before the onset; his weight at
admission was 127. Except for septic tonsils, his examination was negative.
His blood plasma was creamy, and the total lipoids in the whole blood were

TABLE 9.— P.\RT OF Record from Case 9

Diet

Urine

Nitrogen

Pro-

Carbohy

Glu-

SuRar.

Bal-

Dia-

Urine,'

Stool,t In,

ame.

(im. G.n.

rits

C.c.

Gni.

cetic Cent.

• Gm.

Gm

Gm

1/-25

13.6

104.3 15.9

1,15.5

«V0

0

3.94

0.75

2..'iO

4.69

-219

0.15

4.06

0.75

i.M

4.fc3

-0.24

110

550

3.66

3.98

0.75

.s,r3

157.5 26.2

1.655

765

0

3.73

0.75

b.lV

4.48

+069

109

1.915

0.75

2

32.6

172.8 29.6

1.800

745

0

0.15

3.17

0.75

,5.25

3.92

34.2

174.7 29.9

1.830

770

4.73

0.75

a.4V

0.75

6

165

164.3 19.9

i;705

1.220

0.15

0.75

7

167.3 18.4

1.710

1,020

0

3.40

0.75

Aver-

age

31.8

157.4 23.8

1,640

4.12

0.75

5.09

4.8'/

Daily determinations. t Estimated

T.A.BLE 10.— P.\RT OF Record from Case 10

Pro-

Carbo-

Glll-

'

Dla-

Urine.'

Stool,f

1(W

Gm.

Gm.

6m. Calories Gm.

cetic

Gm.

9/26

24.1

1(».6

28.8

.1.735

0

0

29.2

1,660

0

0

28

34.5

133.0

26.4

1.460

29

33.8

111.1

29.3

1,255

30

34.1

143.0

26.3

1,530

0

0

age

33.9

139.0

27.8

l.liOO

6.06

0.68

10/1

2S.4

173.8

22.4

l.ins

0

g

».(,

223.4

36.5

2.390

0

0

37.0

64.8

236.4

38.7

age

51.4

214.1

33.9

2,260

found to be between 8 and 9 per cent, on several examinations immediately after
admission to the ward. Part of the data of this case are presented in Table 11.
He was established in nitrogen balance on a diet containing 0.74 gm. protein
and about 37 calories per kilogram of body weight.

Case 12 (21-678).— Patient was an American clerk, 18 years old, who entered
the hospital March 24, 1921, complaining of weakness. There was no family
history of diabetes. His symptoms began with polyuria, polydipsia, polyphagia,
and progressive weakness, and his weight fell from 135 to 95 pounds. Treat-
ment had been erratic, though at times his diet was restricted to green
vegetables. Twelve days before entrance starvation was started, lasting

108 ARCHIVES OF IXTERNAL MEDICIXE

seven days : during the next five days he received green vegetables and a few-
eggs. For three months he had numbness and tingling of his hands and toes
and there have been a few Ijoils. The physical examination showed nothing
of importance. Nitrogen balance was established on a diet containing 30 gm.
protein. 180 gm. fat and 15 gm. carbohydrate, allowing him 0.71 gm. of protein
and 43 calories per kilogram of body weight. The data are presented in Table 12.

TABLE 11. — Part of Record from C.\se 11
( Daily Aver.\ges for Each Period)

No. Pro- Oarbohy- Glu- Sugar, Fat, Stool , ■—

ates of tein, Fat, drate, Calo- cose, Dia- N, per per N, Wt., In, Out,

Days Gm. Gm. Gm. ries Gm. eetic Gm. Cent. C0= Cent. Gm. Lbs. Gm. Gm.

8 to 24.7 ++ + +

216 9 20.0 90.0 14.0 950 to 0 to 0 .... 0.26 61.4 8.50 .... 126

'.84 0.13 62.6

•^,•11

20.8

82.8

2/22 to

2 ■24

40.1

1.04.4

20",

1,630

2/2St0

3/1

«>y.

152.6

■>l).b

1.615

3/ 2 to

3/7

3/ 8 to

-3/12

40.0

180.3

20.4

1,865

3/13 to

3/17

40.0

179.4

WKW

1,855

3/18 to

3/20

M).t,

200.3

2.').0

2,065

3/21 to

3/29

40.9

200.9

2.5.0

2,070

3/30 to

4/4

40,6

200.2

«5I)

2,065

4/8

43.1

229.1

24.6

2,340

5.53 0.10

121

3.23

7.00

122

6.42

7.90

122

6.43

6.70

119

6.53

6.25

124

6.40

4.32

126

R.40

3.12

l..?9

125

6.48

6 67

-0.19

2.S9

098

121

6.54

7.10

-0.56

2.71

0.95

121

6.50

6.48

+0.02

2.54

0.95

121

6.90

4.98

+1.94

TABLE 12.— Part of Record from Case 12

Pro-

Carbohy-

'

l>atfi

tein,

Fat,

drate,

Calo-

Amt., (

1921

Gm.

Gm.

ries

C.c. 1

3/24

21.7

99.4

14.8

995

1,085

26

29.1

129.1

21.1

i,;ko

2,000

2,225

?8

21.2

85.9

13.4

910

2,900

29

20.2

92.4

14.6

960

2,925

30

30.0

147.1

14.2

1,600

3. 280

31

31.7

151.9

15.5

4,490

41 1

29.9

149.4

14.8

1,525

2,310

St>.2

149.3

14.9

1,525

3.435

28.K

1.52.4

14.8

1.545

1,760

30.3

152.0

14.9

1,550

2.220

31.1

180.7

14.2

1,800

3,220

30.0

1,785

2,260

8

29.7

180.2

15.3

1,800

3,080

lu- Su?ar, Fat. Stool N N

e.Dia- N.' per ^e" N.t Wt.. In, Out,

i.ceticGm. Cent. CO2 Cent. Gm. Lbs. Gm. Gm.

0 0 6.21 0.75 91.5 4 80 6.9ti —2.16

0 0 6.21 0.75 .... 5.07 6.96 —1.89

0 0 4.62 0.75 .... 4.78 5 37 —0.59

0 0 4.62 0.75 91.0 4.83 5..37 -0..t4

0 0 4.62 0.75 .... 4.61 i">.37 —0.76

0 0 4.62 0.164 63.6 1.44 0.75 .... 4 85 5.37 —0.52

0 0 4.11 0.76 .... 4.98 4.86 +0.12

0 0 4.11 0.75 92.0 4.fO 4 86 —0.06

l.ll 0.164 61.7 1.40 0.75

2. Recently Sherman "' has added to his own careful studies of the
protein requirement for maintenance of nitrogen balance in normal
men a complete collection of data ohtaincd from the literature, 109

S. Sherman. H. C. ; Protein Rc(|uirenient of Mainlcnaiue in Man and the
Nutritive Efficiency of Bread Protein. J. Biol. Chcm. 41:97. 1920.

MARSH-SEWBURGH-HOLLY— DIABETES MELLITVS 109

experiments in all. An average of all minimum daily protein intakes
on which nitrogen balance was established gave 0.635 gm. protein per
kilogram of body weight. If the seventy-six determinations showing
the least variation are averaged, a lower figure, 0.58 gm., is obtained.
He points out that the more recent data are, as a result of more accurate
methods and a better understanding of the problem, somewhat lower
than those obtained by the earlier observers, and he himself was able
to maintain several subjects in nitrogen balance on 0.5 or even 0.45 gm.
protein per kilogram of body weight per day. The experiments of
Hindhede,'^ Chittenden ' and others are too well known to require com-
ment. It is an established fact that nitrogen balance may be main-
tamed on less than 0.66 gm. protein per kilogram of body weight per
day, provided certain other conditions are satisfied.

The conditions necessary for the establishment of nitrogen balance
at this low level are several. Chief among them is the presence of
sufficient total calories in the ingested food ; there must be enough fat
or carbohydrate in the diet to supply all the body needs for heat and
energy, so that the protein may be used only for restoring body tissue.
The protein-sparing qualities of carbohydrate and fat were discovered
by some of the earliest students of metabolism, and it is well known
that carbohydrate is the more efficient of the two in sparing protein,
though in a mixed diet fat may replace carbohydrate in isodynamic
quantities. In spite of this difference in the effectiveness of the two
foodstuffs, the ability of fat to spare protein cannot be doubted. For
example, Thomas, quoted by Lusk,"* "could not maintain nitrogen
equilibrium when twice the amount of the fasting nitrogen elimination
was given to man in the form of meat alone, but was able to accom-
plish this when meat to the extent of that destroyed in the fasting was
administered with fat." It is important, then, to remember that if
nitrogen balance is to be maintained on a low protein intake the total
calories in the food must be sufficient to supply metabolic needs.

While a partial replacement by fat of carbohydrate in a low protein
diet will not affect the protein metabolism, complete withdrawal of
carbohydrate and substitution of fat will not permit the establishment
of nitrogen balance at low levels. Fat alone will not decrease the
amount of nitrogen found in the urine of a fasting animal." It is
generally believed that fat in a low protein diet loses part of its effec-
tiveness when the carbohydrate calories fall below 10 per cent, of the
total calories. Thus Zeller '" found that a man receiving little nitrogen

6. Hindhede. M.: Skand. Arch. Physiol. 30:97, 1913; 31:259. 1914.

7. Chittenden, R. H. : Physiological Economy in Nutrition, 1904; Nutrition
of Man, 1907.

8. Lusk, G.: Science of Nutrition, Philadelphia, 1919, p. 254.

9 Voit: Physiologie des StofTwechsels und der Ernahrung, 1881, p. 128;
Bartman: Ztschr. f. Biol. 58:375. 1912.

10. Zeller: Arch. f. Physiol.. 1914. p. 213.

110 ARCHIVES OF INTERNAL MEDICINE

in his food and varying amounts of carbohydrate and fat showed little
variations in his nitrogen excretion until the carbohydrate gave less
than 10 per cent, of the total calories.

It was stated that fat alone will not .decrease the amount of nitrogen
eliminated during starvation. It is also of great interest in connection
with diabetes to note that in an animal still possessing body fat, fat in
the diet does not bring about any change in the amount of fat metab-
olized. Voit," for example, gave 100 gm. fat to a dog, which in starva-
tion burned 96 gm. fat, and found that it burned 97 gm. In other
words, the same amount of fat was burned whether it was derived
from the body or from the diet. The amount of fat stored in the body
has, however, a direct influence on the amount of protein metabolized
during starvation. Lean animals die of starvation sooner than fat
animals, and as starvation progresses and the relative quantity of fat
and protein in the organism decreases the output of urinary nitrogen

T.\BLE 13. — Effect of Previous Diet on Nitrogen Excretion
(C. VoiT)

.\mount ol meat in previous diet... 2,500 gm. 1,500 gm. Mixed diet poor in N
Amoimt oi Protein Used as Determined by Urine N

First fast day 175 77 40

Second fast day 72 54 33

Third fast day 53 46 30

Fourth fast day 50 53 36

Fifth fast day 36 43 35

.Sixth fast day 39 37 36

increases. Folin and Denis " found that the protein metabolism in
fasting was low in people in whom ample fat was present. On the
other hand, if the fat be supplied in the diet, there is no difference in
the protein metabolism of a lean subject and that of a fat one. The
total metabolism of an obese man is the same in proportion to his body
surface as in a lean man. These facts are of practical value in the
arrangement of diets for diabetics of different degrees of nourishment,
and must not be ignored in the study of their nitrogen requirements.

A very marked effect of the previous diet is noted during the first
few days of the study of nitrogen balance. This is shown very nicely
in the experiment by C. Voit, tabulated by Magnus-Levy,^- the data
from which are shown in Table 13. The diets containing the amounts
of meat indicated had been continued for several days before each
experiment.

It will be noted that it is not until the sixth day that the nitrogen
outputs reach a common level. This may be considered as due to the
fact that excess of protein in the tissues or in the blood stimulates

11. Folin and Denis: .1. Biol. Chem. 21:1&3. 1915.

12. Magnus-Levy, A. : Von Noorrlen's Hantllnich dor Patlmlogie des Stoff-
vechsels, 1906. p. 312.

MARSH-NEWBURGH-HOLLY— DIABETES MELLITUS 111

protein metabolism. The experiment is of great importance in empha-
sizing the fact that studies in nitrogen balance made during the first
few days following a change in diet are of no value; the organism
requires some time for its readjustment to the new dietetic regime.

On the other hand, Rubner, quoted by Lusk,* states that the "greater
the impoverishment of the protein supply in an animal fed with fat, the
more powerful is the protective effect of small quantities of ingested
protein over the loss of body protein. Also the retention of protein
depends on the protein content of the dog as well as on the quantity of
protein ingested." This is illustrated in Table 14.

"It is evident from this that of the same diet of protein more will
be retained when the nitrogen content of the dog is low than when it
is high ; and also that a small protein intake may cause the same reten-
tion of nitrogen as a large protein intake, if in the first instance there
be a relative impoverishment of the protein content of the animal."

T.ABLE 14. — Protective Effect of S.m.\ll Qu.vntities of Protein when

Body Protein is Impoverished

(Rubner)

Total X Content
of Dog

318.8

N in Terms of lOO N in Dog

It has been stated that in order to maintain nitrogen balance on a
diet with a small protein content, the total caloric requirement of the
subject must be satisfied. This requirement is not represented by the
rate of basal metabolism. The caloric needs of the fasting, resting
subject are not the same as those of the same subject after the ingestion
'if food or in the performance of even the smallest amount of work.
A certain amount of energy is used in the metabolism of the food, and
this "specific dynamic action" is different for each of the three major
foodstuffs. For protein the increased heat production is especially
high, amounting to from 25 to 35 per cent, of the ingested calories. In
the case of fat the figure is about 12 per cent. ; for carbohydrate much
less, about 6 per cent. This factor must enter into our calculations of
the total daily requirement, and if the diet contains an excess of protein,
the amount of food necessary to supply the caloric needs of the body,
and hence to maintain nitrogen equilibrium, is considerably greater
than when fat and carbohydrate furnish the major part of the calories.

The amount of work done by the subject obviously influences his
caloric needs. A man whose resting requirements are satisfied by 1,500
calories daily may burn 9,000 calories in a bicycle race. It is evident
that this fact must be considered in the formation of a low protein

112 ARCHIVES OF IXTERXAL MEDICIXE

diet that will maintain nitrogen balance. On the other hand, there is
abundant evidence to show that work does not increase the rate of
protein metabolism when the energy is supplied from other foodstuffs.
The experiment by Shaffer " presented in Table 15 may be cited as an
example ; during the first period he was at rest in bed ; during the
second he was doing laboratory work, and during the third, laboratory
work plus a ten-mile walk daily ; the periods were six, five and four
days respectively.

The other end-products of protein catabolism showed a similar con-
stancy, and he concluded from .this that muscular activity had no effect
on protein metabolism. This fact is important because we may arrange
the protein content of a diet without regard to the amount of work that
the patient will do. If he is in nitrogen balance at rest, the calories
required for energy for work may be added as fat or carbohydrate.
Furthermore, it has been shown that there is no difference in value of
fat and carbohydrate as a fuel for mechanical work.'*

TABLE 15.— Effect of Work on Protein Metabolism
(Shaffer)

In the formation of diets the age of the subject must receive con-
sideration. The total caloric requirement of children is much greater
in proportion to their size than is that of adults, and this requirement
declines as the subject becomes older. Furthermore, beside the "repair
quota" of protein needed to replace that lost in the wear and tear of
tissue, children require a "growth quota." While, as far as we know,
there is no record of a careful determination of the minimum protein
intake that will supply the normal growing child with the needed nitro-
gen, it must be considerably higher than that required by adults.

Sex is of no importance in this connection. Some of the studies
discussed by Sherman '* had for their subjects women. Since women
have a somewhat lower metabolic rate than men, a diet that is sufficient
to supply the needs of a man will certainly supply those of a woman
of the same size. There is no change in the oxidation processes
during menstruation,'"' and the amount of nitrogen lost is evidently

13. Shaffer, P. .-\. : Diminislicd Muscular .\ctivity aud Protciu Mctaliolism,
Am. J. Physiol. 22:445, 1908.

14. Zuntz: Arch. f. d. gcs Phy.siol. 83:557, 1900.

15. Gephart, A. B.. and Du Bois, E. F. : Ba.sal Meta)iolism of Normal Adults.
Arch. Int. Med. 17:907 (June) 1916; Bhmt, K.. and Dye, M.: Basal Metabolism
of Normal Women, J. Biol. Chcm. 47:69, 1921.

MARSH-XEIV BURGH-HOLLY— DIABETES MELLITUS 113

small. The pregnant woman requires an extra quota of nitrogen and
extra calories during the gestation period in order to supply the embryo ;
as the weight of the child at birth is only 5 or 6 per cent, of that of the
mother, the extra ration need not be very great.

The weight of the subject is of importance in giving us a basis on
which we may calculate his nitrogen requirements and, roughly, his
total diet. Gain and loss of weight during the periods of treatment
when the diets are being changed mean little, as a rule, because such
changes are very often due to changes in the water content of the body.
In the normal subject, change from a carbohydrate diet to a fat or
protein diet causes a loss of water ^* amounting to as much as a kilo-
gram a day. Dehydrated subjects, on the other hand, may absorb and
retain enormous quantities of water, with the resulting increase in body
weight. This is seen not only in diabetes, but in other diseases, as
pyloric obstruction. We therefore frequently see the apparent paradox
of a patient who gains weight rapidly on a diet containing only 1,000

TABLE 16.— Effect of Ingested Water on Nitrogen
(Abderhalden and Bloch)

Excretion

N Balance
Normal food +1.36

N in^Urine

Normal food +1.47

18.09

calories daily. More than this, the more severe cases of diabetes show
a marked tendency to the de\elopment of edema. Because of these
facts, one is not entitled to draw conclusions as to the state of the
patient's nutrition from changes in body weight. As pointed out before,
the important consideration is the fat and protein content of the body.

A small increase in nitrogen elimination has usually been noted fol-
lowing the drinking of large quantities of water. Hawk,^' for example,
found that the ingestion of 4.500 c.c. water with an unchanged nitrogen
content of the diet caused the urinary nitrogen to rise from 11.03 to
12.48 on the first day, and 11.82 on the .second, with a fall of 10.91 on
the succeeding day when no water was given. Alderhalden and Block '*
gave a subject on a fixed diet 5 liters of water with the results shown
in Table 16.

It is probable that two factors explain this rise in nitrogoii elimina-
tion; first, the water removes from the body any accumulation of the
end-products of nitrogenous metabolism, and second, causes a true
increase of the protein metabolism. This is of interest in the case of

16. Bischoflf and Voit : Die Gesetze der Ernahrung des Fleisch fresscrs,
Leipsic. 1860; Benedict. Milner: U. S. Dept. Agric. Office, Exper. Stat. Bull.
175. 1907.

17. Hawk: Univ. Penn. Med. Bull.. March, 1905.

18. Abderhalden and Bloch: Ztschr. f. Physiol. Chem. 53:464, 1907.

114 ARCHIVES OF IXTERXAL MEDICIXE

a diabetic who drinks large quantities of water, eithei as a symptom of
his untreated disease or during periods of very restricted diet when he
attempts to reheve some of the pangs of hunger by filling his stomach.

There is also an irregular but important day to day variation in the
nitrogen output that is independent of variations in fluid intake. These
fluctuations are the result of lag in excretion, and in the long run they
are compensatory. A nitrogen determination for a single day is an
unsafe indication of the usual excretion ; safety is in the average of
several days.

Beside all these factors influencing the protein metabolism, there
are a few manifestations of nondiabetic disease which may cause pro-
found disturbances in the nitrogen balance. The great increase in the
rate of total metabolism and especially of protein metabolism in hyper-
thyroidism is well known. In the diabetic, the most important cause
of such disturbances is fever. The increase in nitrogen elimination
during fever may be enormous. In a study of the metabolism in pneu-
monia, for example, ^^ the daily loss in nitrogen on a diet adequate to
protect a normal individual from loss may be from 20 to 25 gm., repre-
senting from 125 to 150 gm. protein. Of chief importance in the study
of diabetes is tuberculosis, in which the destruction of protein, while
smaller than in pneumonia, is continued over a much longer time.'"'
Even during the afebrile periods there is some increase over normal in
the protein destruction. Futhermore, in these infectious fevers, the
total caloric requirement is greater than normal because of the increased,
heat production. The presence of such a fever must modify our diet
in diabetes.

3. It is seen, then, that the normal subject may be maintained in
nitrogen balance on less than 0.66 gm. protein per kilogram of body
weight, provided the total caloric intake is sufficient to supply heat
and energy. Certain fundamental laws governing protein metabolism
have been discussed, and because of their importance are here restated.

In mixed diets carbohydrate and fat are equally efficient in sparing
protein and of equal value as fuel for work. Carbohydrate may not be
entirely replaced by fat, however. An animal having a supply of body
fat will burn the same amount of fat, whether it is supplied in his diet
or is used from his store. A good supply of body fat thus saves pro-
tein in the same way as ingested fat. More nitrogen is excreted during
the days following a diet rich in protein than following one poor in
protein. The increased heat elimination due to the metabolism of food

19. Wolf and Lambert: Protein Metabolism in Pneumonia, Arch. Int. Med.
5:406 (March) 1910.

20. May: Ou's Chemiscbc Pathologie der Tiiberculose, Berlin. 1903, p. 335.
McCann. W. S.. and Barr. D. P.: The Metabolism in Tuberculosis, Arch. Int.
Med. 26:663 (Nov.) 1920.

MARSH-SEWBURGH-HOLLY— DIABETES MELLITUS 115

has been pointed out, and the excessive "specific dynamic action" of
protein has been emphasized. Work causes an increase in the total
metabolism, but no increase in the protein metabohsm provided the
calories needed are supplied in other form. Children require more
protein than adults, and the total requirement decreases as age
advances. Sex is of no importance in studies of protein metabolism.
There is a marked day to day fluctuation in nitrogen excretion, partially
dependent on increased water intake. Changes in weight must be inter-
preted with caution, because they are often due to changes in the water
content of the body. The infectious fevers cause a great increase in
the amount of protein destruction.

The laws governing protein metabolism of normal subjects have
been stated. The question then arises as to whether or not these facts
apply equally to the diabetic. Is there any peculiar disturbance of his

T.-\BLE 17. — Protein Sp.aring Qualities of Fat in Diabetes Mellitus,
(Weintbaud)

18.19 0.9 16.49

18.19 0.9 16.34

18.19 • 0.9 16.88

18.19 0.9 16.01

18.19 0.9 14.00

18.19
18.19
18.01
18.01
18.19
18.19
18.19
18.19

protein metabolism that makes his nitrogen requirement higher than
that of his normal brother? Is his protein metabolism affected to the
same degree and by the same factors as is that of a healthy subject?

That the urine in diabetes may contain more nitrogen than that of
any other disease has been known for a long time and the French even
gave this finding a name, azoturia. \'on Noorden.^' however, pointed
out that this is due to the large nitrogen content of the diabetic's diet,
either because of his own tendency to replace with protein the carbo-
hydrate calories lost in the urine, or because of the very high protein
content of the contemporary diabetic diet.

Weintraud ^^ demonstrated conclusively as long ago as 1893 that
fat fed to the diabetic subject was very powerful in saving his protein.
An example of his experiments is shown in Table 17.

21. Von Xoorden. C: Die Zuckerkrankheit, Berlin, 1507, p. 97.

22. Weintraud. E.: Stoffwechsel in Diabetes Mellitus. Cassell, 1893.

116 ARCHIVES OF IXTERXAL MEDICI XE

It will be noted ( 1 ) that a decrease in fat calories without change
in the nitrogen in the food resulted in the immediate production of a
negative nitrogen balance; (2) the return to the diet rich in fat left the
patient in negative balance for the first four days of the regime; but
(3) during the second four days he was able to add nitrogen to his
body. Had the experiment begun on the fourteenth and ended on the
twenty-first, we would have been led to the conclusion that the fat did
not spare nitrogen ; a similar conclusion would result from the discon-
tinuance of the experiment on the twenty-second, with the average
nitrogen output of the first six days following the change. The fallacy
of drawing conclusions from observations on the days immediately fol-
lowing changes in diet is apparent. The author also made quantitative
determinations of the fat in the stools, in an effort to show whether or
not there was any difficulty in the absorbtion of large quantities of fat
from the intestinal tract of diabetics. Three normal subjects excreted
an average of 3.2 per cent, of 150 gm. of fat. Table 18 presents his
results in cases of three grades of diabetes.

TABLE 18. — Fat Absorbed from High Fat Diets in Diabetes Mellitus
(Weintraud)

Moderately severe..

Fat

Absorbed

ood rat
Gm.

Stoo.^.at.

Gm.

Per Cent.

45.94

7.41

38.53

83.87

146.4

272.04

3.31

268.73

9S.78

39.42

2.16

37.24

94.47

2.52

88.28

206 02

3.10

202.92

98,61

56.41

3.15

52.26

94.31

236.11

4.35

231.76

98.15

This general relationship held in all his determinations, namely
that the percentage of fat in the stools fell with increase in the fat
intake, and all of his diabetics showed an average absorbtion of about
98 per cent, of the ingested fat when the latter was over 125 gm. daily.

Dunlop '•'' doubted the value of fat in sparing protein in diabetes,
as demonstrated by Weintraud '^ and quoted in support of his doubt
the experiment of Karsener ^* who brought himself into nitrogenous
equilibrium on a diet of protein and carbohydrate, and found that he
was no longer able to maintain this equilibrium when he replaced a
large part of the carbohydrate with the caloric equivalent of fat.
Dunlop fed a diabetic patient 4,660 c.c. of skimmed milk for two
periods of three days each, and then added to the milk diet a daily
ration of 170 c.c. of olive oil. The results are shown in Tabic 19.

2.3. Dunlop. J. C: Dietetic Values of Fat in Diabetes, Edinlnirgh W. T.
42:,W. 1896.

24. Karsener: Von Noorden's Beitragc z. Lehrc v. StoffwcchscI, Heft K

MARSH-SEW BURGH-HOLLY— DIABETES MELLITUS 117

Because the oil did not decrease the nitrogen in the urine, the
author argues that it did not act as a nitrogen sparer. This is, of course,
only in accord with the estabhshed fact that a subject cannot be forced
to store nitrogen, and that nitrogen ingested in excess of the require-
ment will be excreted rather than stored ; the experiments show nothings
in regard to the protein sparing qualities of fat. The experiment of
Karsener merely confirms the greater efficiency of carbohydrate than
fat in sparing protein.

The excessive nitrogen excretion of totally depancreatizcd dogs is
well recognized, and may reach five times the normal amount.-"' Some-
times these dogs develop a rapid and fatal cachexia without glycosuria,
indicating, it would seem to us, that the excessive nitrogenous metab-
olism in such depancreatizcd animals was not part of the diabetic state.
Furthermore, numerous pancreatic operations on human beings have
been reported in which there was greatly increased nitrogen excretion

T.\BLE 19.— The Effect of Adding Fat to the Diet of a Diabetic Already
IS Nitrogen Balance

(DUNLOP)

Period

Food

Nitrogen Output

x.

Added Oil

Urine

Stool N-Balance

1
2
3

4

21.8
24.8
24.8
34.8

iro

19.41
18.02
24.1fi
25.06

0.39
0.44
0.38

+5.0O
+6.34
+0.26
-0.26

Av. 1, 3
Av. 2, 4

24^8

24.8

iro

21.8
21.54

0'.44

+2.62
+2.82

without glycosuria. On the other hand, occasional partially depan-
creatizcd animals have been described in which the D : N ratio was
that of "total diabetes" and yet the cachexia (increased protein metab-
olism) was absent. Allen ^" has been able by excessive feeding of
partially depancreatizcd dogs to develop a condition which he considers
a very satisfactory imitation of human diabetes ; in these dogs there is
no increase in the protein catabolism. Although he is one of the
staunchest supporters of the parallelism between such dogs and human
diabetics, and of the dictum that "without disease of the pancreas there
is no diabetes," he believes that the azoturia in the totally depancrea-
tizcd dogs was due to other factors than those which caused the dis-
turbance in carbohydrate metabolism. That is, in his opinion, the
pancreas possesses other internal secretions in addition to that con-
trolling carbohydrate metabolism, and in human diabetes these other
functions are not disturbed. Even the most ardent adherents to the

25. Falta, \V. ; Grole, F., and Staehelin, R. : Versuche iiber Stoffwechscl
iin(] Energieverbrauch an pancrealosen Hunden, Hofmeistcr's Beitrage 10:199..
1907.

26. Allen, F. M. : Glycosuria and Diabetes, Cambridge, 1913, pp. 427, SOS.

118 ARCHIfES OF IXTERXAL MEDICIXE

hypothesis of the pancreatic origin of diabetes would not contend that
there is as complete destruction of the pancreas in the human diabetic
as there is in the depancreatized dog. Such animals by being deprived
of the external secretion of the pancreas can no longer properly digest
food and must accordingly be in a state of starvation. It is obvious
that evidence of increased protein metabolism obtained from the study
of depancreatized dogs must not be considered proof of a similar
increase in human diabetes.

Lusk and others have shown -' that in dogs with phlorhizin glyco-
suria the nitrogen excretion may be four or five times that of normal
dogs. Allen ^* states that this is due ( 1 ) to the fever caused by the
subcutaneous injection and is not always present if the drug is given by
mouth, and (2) to the secondary breakdown of protein to replace the
urinary loss of sugar in fasting or insufficiently fed animals. Sugar in
doses above a certain quantity is burned and spares protein in the

TABLE 20.— Studies of

THE

M

ET.^EOLISM OF .\ SpOXT.\NEOUSLY

(Maicnon)

Diabetic Dog

Regime

Days

Daily weight loss, gm

Urea,gm

Sugar, urine, gm

Carbohydrate Meat, Starv

as Desired 500 Gm. tion

...4 2 1

30O 250 300

12.24 34.60 16.38

125.47 51.71 19.17

0.662 1.248 0.122

- ii^t''

10

0

5.99

3.78

0.6V0

phlorhizinized as in the normal animal. Important as is the informa-
tion concerning many of the processes of metabolism obtained by the
use of phlorhizin, it may be applied only in a limited way to the problems
of diabetes. Phlorhizin glycosuria and diabetes mellitus are two
fundamentally different conditions.-^ The normal sugar content of the
blood of the phlorhizinized subject, his ability to burn glucose in excess
of a given quantity determined by the dose of the drug and the causa-
tion of his glycosuria by local renal disturbances offer sharp contrasts
to the hyperglycemia of diabetes mellitus, the invariable appearance
in the patient's urine of excess of ingested carbohydrate and his
glycocuria because of a disturbance of the carbohydrate metabolism.
The studies by Maignon ^^ of a dog who became diabetic spon-
taneously offer some interesting contrasts to the obsei-vations made on
animals rendered "diabetic" by either phlorhizin or pancreatectomy.
In Table 20 are found a summary of his experiments.

27. Reilly. F. H.; Nolan, F. W.. and Lusk, G.: Phlorhizin Diabetes in Dogs,
Am. J. Physiol. 1:395, 1898. Lusk, G.: Ueber Phlorhizin-Diabetes, Ztschr. f.
Biol. 42:,?1. 1901. Mandel, A. R., and Lusk, G. : Respiration Experiments in
Phlorhizin Diabetes, Am. J. Physiol. 10:47, 1903.

28. Allen: Loc. cit.. p. 617.

29. Maignon, F. : Du Role des Graisses dans la Glycogenic, J. Physiol, et
Pathol. Gen. 10:866. 1908.

MARSH-SEMBURGH-HOLLY-DIABETES MELLITUS 119

The experiments of Benedict and Joslin ^° on the rate of nitrogen
excretion of diabetics without food receive frequent mention in con-
nection with the nitrogenous metabolism of diabetics. Allen ^' refers
to this work as evidence of increase of nitrogen destruction, and
Foster ^- similarly quotes these authors. If the evidence actually does
show an increased protein destruction in the diabetic, it demands very
careful consideration ; it matters not whether this increase be held to
be a specific action of some factor of the morbid state upon protein, or
as one of the compensatory measures for the carbohydrate deficit, such
an increase if real must be reckoned with in the preparation of a low-
protein maintenance diet.

The urines of thirteen fasting diabetics were collected in the morn-
ing and the hourly rate of excretion of nitrogen per kilogram of body
weight determined. The average of all determinations was 8.4 mg.
nitrogen per kilogram of body weight per hour, with variations among
the individual cases of from 4.2 to 12.1 mg., and among different
experiments in the same cases of as much as 5.3 mg. (from 4.2 to 9.5).
In a series of fourteen normal subjects the average excretion per
kilogram of body weight per hour was 6.8 mg., with variations between
4.8 and 8.9. Five of thirteen of the diabetic patients excreted less than
the average of the normals, and three of fourteen of the normals
excreted more than the average of the diabetics. The fact remains,
however, that the average of the diabetics was higher than the average
of the normals, and four of the diabetics eliminated more nitrogen than
the highest of the normals.

If these cases be studied in relation to their degrees of emacia-
tion, it will be noted that, in general, the more poorly nourished have
the highest rate of nitrogen excretion. In Table 21 the subjects are
arranged in the order of their rates of nitrogen excretion ; added are
columns showing the average weight for the age and height of each
case as found in the actuarial tables, the number of pounds above or
below this and the {x.'rcentage of digression from this average.

Cases X and W were clinically only moderately severe, M was
light and the rest were severe. It will be noted that if the difference
between the weight of the patient and the average normal weight may
be taken as a rough measure of the degree of emaciation the better
nourished patients excreted less nitrogen than the more lean patients.
The average nitrogen excretion of the seven patients showing the least
loss of weight was 6.6 mg. per hour (the controls averaged 6.8 mg.)
jvhile the average of the other 6 was 10.3. The i)roper ex])lanation of

,TO. Benedict, F. C, and Joslin. E. P.: Mcta1)c.li.sm in Severe Diabetes,
Washinetnn, 1912. p. 112.

31. .Mien, F. M.: loc. cit., p. 419.

32. Foster, N. B.: Diabetes Mellitus, Philadelphia. 1915. p. 173.

120 ARCHIVES OF ISTERXAL MEDICIXE

these figures then seems to be that, whereas none of these fasting
patients could supply fuel for energ}' from the store of body carbo-
hydrate, some had a sufficient store of fat to serve the purpose and to
render it unnecessary for them to bum protein for fuel. These
determinations offer further evidence of the ability of fat to save
protein in the diabetic. The results are what one would predict from
the laws of normal metabolism ; they give no evidence of abnormal
protein metabolism in the diabetic.

Very occasionally a case of unusually severe diabetes is seen in
which even the power of burning the glucose derived from small
amounts of protein is lost, as is indicated by the D : N ration of 3.65 : 1
on a carbohydrate free diet.'^ Mandel and Lusk,^* for example studied
carefully the metabolism of such a case . The data they obtained con-
cerning his nitrogen metabolism are presented in Table 22.

TABLE 21. — Hourly R.\tf. of Nitrogen Excretion Per Kilogr.\m in

Diabetes Mellitus

(Benedict and Joslin)

—76 —50

The fact that this patient was not in nitrogenous equilibrium when
his food contained 19 gm. nitrogen, and that he lost 14 gm. body nitro-
gen when he was fed boullion only with a nitrogen content of 7.7 gm.
is frequently cited as evidence of increased protein metabolism in
diabetes. Even after the fuel value of the food which was useless to
the patient because of his inability to utilize carbohydrate is subtracted,
his total caloric intake was sufficient to supply his metabolic needs.
A moment's thought, however, will show that the increase in protein
metabolism can be entirely explained by his total inability to burn
carbohydrate, rather than any disturbance in the protein metabolism
inherent in the disease. Not only is carbohydrate itself unavailable
for energy, but as has been ])ointcd out, the fat is also unavailable to.

33. Geylin, H. R., and Du Bois, E. F. : A Case of Diabetes Mellitus of
Maximum Severity. J. A. M. A. 66:1532. 1916.

34. Mandel, A. R.. and Lusk, G. : Stoffweclisellieobachtungen an einem Fall
von Diabetes Mellitus, Deutsch. Arcb. klin. Med. 81:472, 1904.

MARSH-SEWBURCH-nOLLY— DIABETES MELLITUS 121

spare protein because it requires some carbohydrate combustion to
do this.

Mosenthal has recently studied in two series of cases the main-
tenance diet in diabetes meUitus as determined by the nitrogen equili-
brium. In the first study '^ he fed diets in which fat and protein were
about equal gram for gram containing ten to fifteen grams of carbo-
hydrate. He started with diets low in calories and increased them
until nitrogen balance was established, always feeding equal amounts
of protein and fat. With such diets he found that it was possible to
establish nitrogen balance on a caloric intake equal to that required
by a normal person, and in some cases it could be established at an
even lower level. To provide from 1,500 to 2,000 calories it is
necessary on a protein-fat diet containing equal quantities of protein

TABLE 22. — Studies of Metabolism of a Very Severe Diabetic
(Mandel and Lusk)

Pro-
tein,
Gm.

Fat.
Gm.

Car-
Gm.

Calories

Urine

Stool
Gm.

Nitrogen

Date

JS,'.

Out.
Gm.

Balance.
Gm.

1»
23

Mixed diet
Mixed diet
Mixed diet

137 23.3
119 24.0
150 22.2

IS
1.3
1.3

24.6
2.5.3
23.5

25
26

117.5
117.5

170.7
170.7

84.8
84.8

2,580
2:580

148 20.9
144 19.0

1.3
1.3

18.8
18.8

22.2
20.3

—3.4

27

117.5
117.5
117.5

170.7
170.7
170.7

134.8
181.8
184.8

2,768
2.956
2.936

179 18.5
210 17.7
217 17.8

1.3
1.3
1.3

18.8
111

19.8
19.0
19.1

— o!2

1

113.1

164.0

6.6

2.207

83 17.5

1.3

18.1

18.8

-0.7

2

168.1
168.1
191.1

192.1

5.5
5.5
5.5

2.7S9
2.552
2.645

88 23.2
93 24.1
106 27.7

1.3
1.3
1.3

26.9
26.9
30 6

Si
39.0

+i'.'6
+1.6

15

48.1

0

0

85 21.7

7.7

21.7

—14.0

and fat to raise the protein ration to from 100 to 150 gm., and there-
fore Mosenthal found that nitrogen balance was established with
relatively large amounts of protein — amounts that would not be toler-
ated without glycosuria by severe diabetics. Because of the fear of
acidosis no efifort was made to establish balance on low-protein, high-
fat diets. Since he increased the protein content of the diet at the
same rate as he did the fat it was to be expected that he would not
feed enough calories to supply the body needs until the diet con-
tained large amounts of protein. He showed, then, that when one
feeds a diet containing equal weights of protein and fat. one is
compelled to supply large quantities of protein in order to establish
caloric equilibrium. Tic did not show that the diabetic requires more
protein than the normal subject.

,35. Mosenthal. H. O. : The Maintenance Diet in Diabetes Mellitus as
Determined by the Nitrogen Equilibrium, Tr. Assn. Am. Phys. 32:159, 1917.

122 ARCHIVES OF INTERNAL MEDICINE

In the second series of experiments,- an effort was made to
determine the food value of alcohol, fat and protein by their abilities
to spare nitrogen. After a few days on a control diet which usually
contained about 70 gm. each of fat and protein and a small amount of
carbohydrate, and which induced a negative nitrogen balance because
of its low caloric content, he tried the effect of adding to the diets
either alcohol or fat or protein or combinations of these. These addi-
tions brought the total calories up to from 1,500 to 2,000. Each such
new diet was tried four to six days.

Only in occasional instances did the addition of fat establish nitro-
gen balance; since these were not severe diabetics, tolerating as they
did from 150 to 200 gm. protein without glycosuria, the factors con-
tributing to the enormous nitrogen output in such cases as that of
Mandel and Lusk do not have to be considered. Mosenthal reached
the conclusion that "the addition of an equal number of calories of
protein, fat or alcohol to a low caloric, carbohydrate-free diet in cases
of diabetes mellitus results in the assimilation of considerable amounts
of nitrogen when protein is used, a favorable nitrogen balance in only
occasional instances with fat," and no change in the nitrogen equili-
brium when alcohol is given," and he stated that "this would point to
a high protein diet as the most advisable low-caloric, carbohydrate-free
diet by which to conserve the body tissues and furnish a maintenance
ration for the diabetic."

We cannot accept these conclusions, however, liecause the experi-
mental periods employed by Mosenthal were too brief to permit the
subjects to totally rid themselves of the metabolic products of the
previous diet or to readjust themselves to the new diet.

It was shown by C. Voit (Table 13) that it took fasting dogs five
days to reach a common level of nitrogen excretion and that the higher
the nitrogen content of the previous diet the more striking was the
fall of nitrogen output from day to day. Weintraud showed (Table 17)
in diabetics that the addition of fat to a high-protein, low-fat diet
which induced a negative nitrogen balance would induce a positive
balance only after several days had elapsed. A subject to whom he
was feeding 111 gm. protein and 28 gm. fat had a negative balance
of 6.7 gm. nitrogen. It was not until si.x days after the addition of
245 gm. fat to this diet that the first positive balance of only 1.34
gm. was attained. Three days later, on the same diet, the balance was
3 gm. We have had this same experience and have convinced our-
selves that one is not justified in judging of the effect of a diet until
it has been fed at least a week.

In two of Mosenthal's ten cases, the increase in fat followed immedi-
ately after a period of high protein feeding. The results are shown in
Tab'le 23.

MARSH-SEWBURGH-HOLLY— DIABETES MELLITUS 123

In each of these cases the negative balance during a high-fat period
following ioimediately after a high-protein period was greater than the
negative balance during the control period when the total calories were
a third less; this negative balance represented merely the continued
excretion of the previously ingested protein. Case 5 had, in fact,
been in nitrogen balance on the added fat on a previous occasion, when
it did not follow a high protein period. The extract presented in
Table 24 from the record of Case 7 in our series shows a similar lag
in the excretion of nitrogen, with the subsequent sweeping-out during
the following few days, and the eventual establishment of practical
nitrogen balance.

During the period of high-protein feeding there was an apparent
retention of nitrogen in the body, but during the succeeding low-protein

TABLE 23.— Effect of Previous High Protein Diet on Nitrogen Bal.^nce
(Mosenthal)

:s Balance on

Fat,

Carbohydrate,

N Balance,

Control Diet,

Case

Gm.

Gm.

Gm.

Calorics

Gm.

5

150.6

89.4

13.4

1,602

+2.6

—2.7

125.6

12.1

1.506

-3.1

8

89.4

13.5

1,603

+5.7

—1.5

71.4

125.2

12.0

1,510

T.\BLE 24.— Effect of Previous High Protein Diet on Nitrogen Balance

Food, Dally Average

Dally Average

Total for Period

Days

5

6

Pro-
tein,
Gm.

33.5

Carbo-
Pat, hydrate,
Gm. Gm. Calories

iJfi ^:^ 1:^

130.1 25.4 1,3K

NIn,
Gm.

29..'i8

NOut, Balance,
Gm. Gm.

18.38 +11.20
14.75 —9.52
5.41 -0.06

NIn,
Gm.

147.88
31.38
32.16

N

NOut, Balance,
Gm. Gm.
91.89 +.W.95
88.48 -57.10
32.48 —0.32

period this nitrogen was all excreted; the total nitrogen ingested was
179.26 gm. and the total nitrogen excreted was 180.37 gni., practically
a balance.

.Another error in these short period experiments arises from the
fact that the organism apparently requires some time to adjust itself
to new dietetic conditions. Sherman believes that the variations between
the results of numerous investigators of the minimum protein intake
thjt would support nitrogen balance was due, in part, to "the differing
lengths of the investigations and the extent to which the individual had
accustomed himself to a low protein diet." He says further that
"while it is conceivable that a small loss of body nitrogen may represent
a real inadequacy of the intake, perhaps as regards some particular
amino acids, yet it is usually much more probable that a small negative
balance means simply that the body has not yet completed the adjust-

124 ARCHIVES OF INTERNAL MEDICINE

nient of its output to its intake and that a continuation of the experi-
ment would have shown a smaller output."

Several of Mosenthal's patients were returned to the control diet
immediately following the diet to which fat had been added ; the con-
trol period, the fat period and the second control period are shown
in Table 25.

In all these cases the saving of protein by fat showed in the period
following the one during which the fat was given, as shown by the
smaller negative nitrogen balance on the control diet after the fat
feeding than before it. It seems safe to conclude that if the fat feed-
ing had been continued a longer time in each case, nitrogen balance
would have been established.

T.\BLE 25.— Delayed Effect of Addition of F.\t in Est.\blishinc Nitrogen
Balance

As the result of these studies and others of a similar nature, the
belief has become general that patients with diabetes mellitus require
more protein for the establi.shment of nitrogen balance than do normal
subjects, and that there is, as an inherent part of the disease, an
abnormally high rate of protein metabolism with an increased elimina-
tion of nitrogen.

Most of the |)revious investigators have iiecii handicapiied by the
fear of the use of fat in the treatment of diabetes mellitus, and have
hence in none but the mildest cases, been able to increase the non-
protein calories to a level that would satisfy the energy require-
ment without the production of glycosuria. The use of a high-fat,
low-protein, low-carbohydrate diet in this clinic has enabled us to stufiy
in a more satisfactory manner than was hitherto possible, the minimal
protein intake that will maintain nitrogen balance in the diabetic.

In Table 26 are presented the lowest diets on which nitrogen balance
was established in our series. It must be remembered that this does
not represent in every case the lowest possible level for balance, but
the lc\cl at which, in the course of treatment, balance was established.

MARSH-KEWBURGH-HOLLY— DIABETES MELUTUS 125

Some cases, as for example Case 6 in which the positive balance was
at first over 2.5 gm. nitrogen or 16 gm. protein, could undoubtedly have
been established in balance on a lower ration than that allowed.

In spite of the fact that these diets cannot be considered as the
lowest on which nitrogen balance could have been established, they
contained an average of 0.68 gm. protein per kilogram of body weight.
This agrees with the figure found by Hindhede and others on normal
subjects, and demonstrates that the diabetic patient may be maintained

TABLE 26. — Diets on which Nitrogen Balance was Established in
Twelve Patients with Diabetes MellitOs

Per Cent.
Colo- ol Total
Pro- Carbohy- Protein ries Caloriis as

per Carbohy-
Kg. drate

Ase.

Weight,

tein

Fat,

.irate.

K,

Years

Kg.

Gm.

Gm.

Gm.

Calorics

Gm.

Gm.

5fi

63

343

1(«.4

6.9

1.6S0

5.49

0.54

68

66

38.2

231.4

13.8

2,295

6.27

0.58

37

60

54.1

240.0

13.4

2:430

8.66

28.5

162.8

37.8

49

66.0

197.0

9.9

2,065

10.04

0.93

22

59

33.5

150.1

25.4

1.589

5.36

0 53

36

69

SS.5

207.8

38.7

2,^45

8.80

50

31.8

80

90

51.4

239.1

33.9

2,265

8.22

0.57

22

40.6

20O.2

25.0

2,066

6.50

0.74

18

42

30.0

179.6

14.7

1,795

4.80

0.71

in nitrogen balance on as low a protein ration as the normal subject.
The average number of calories per kilogram of body weight given to
this group of patients was 33.5 and of these calories, an average of
only 3.8 per cent, were in the form of carbohydrate.

The average amounts of each of the foodstuffs used to make up
these 33.5 calories are shown in Table 27. The fat in grams is ten
times the carbohydrate in grams, or if the 58 per cent, of the carbo-
hydrate that may be derived from protein be added, the weight of the

T.ABLE 27. — Average Amounts of Foodstuffs Per Kilogram of Body Weight
Used in Establishing Nitrogen Balance

Protdn 0.68 2.72

Fat 3.28 29.51

Carbohydrate 0.S2 1.27

fat is four and one half times that of the total carbohydrate. The
carbohydrate calories are 3.8 per cent, of the total calories, or, if the
protein carbohydrate be added, the total carbohydrate calories are only
8.6 per cent, of the total calories. Nitrogen balance can be established
in the diabetic on diets low in protein whose energy is chiefly con-
tained in fat.

To this law, however, there is an exception which for the sake
of completeness must be mentioned again. A case such as that studied

126 ARCHIVES OF IXTERXAL MEDICISE

by Mandel and Lusk, although very rare, is so severe as to have lost
its ability to burn carbohydrate even in the small amounts necessary
for the metabolism of fat. Such a diabetic who has so lost the ability
to burn both carbohydrate and fat is inevitably thrown back on protein
as a source of energy. The impossibility of satisfying his caloric
requirement in other ways results in the excessive metabolism of
protein. The rarity of such cases, however, makes them of little
practical importance.

Several observations made during this study and already discussed
should be mentioned again in this place. One case was cited which
showed a delay in excretion of nitrogen during the administration of
a high protein diet, with a sweeping-out during the following days
of low protein feeding. As a result, there was an apparent large
positive nitrogen balance during the high protein period, a large nega-
tive balance during the few days immediately following, and the
ultimate establishment of balance after a few days (Table 23). Atten-
tion has also been called to the fact that nitrogen balance may be
established on a given diet only after it has been administered for
some time.

A diabetic patient suffering from advanced chronic pulmonary
tuberculosis was studied. The data obtained were in accord with the
observations of McCann and Barr and others regarding the increased
protein destruction in tuberculosis.

Case 13 (20-461). — Patient was a German-American ward tender, 28 years,
old. who was brouglit into the ward in impending coma after having tried to
treat his diabetes with drugs and without diet. The diabetes was of about
three years standing, and there had been no symptoms of tuberculosis. After
a few weeks of treatment, he was in excellent condition as far as his diabetes
was concerned, but his tuberculosis was advancing rapidly and came to a fatal
termination about a month after the first nitrogen determinations were made.
The diagnosis of tuberculosis was made by physical examination and confirmed
by roentgenogram, acid-fast bacilli in the sputum and necropsy. The day to
day data of the last month of this patient's life are presented in Table 28 and
a summary in Table 29. He died September 25, without glycosuria or a positive
ferric chloride test on his urine, and there was no evidence of acidosis as
measured by the VanSlyke method. During the last six weeks his temperature,
which had previously been normal, reached from 100 to 103 F. daily.

The contrast between this patient who showed so large a negative
nitrogen balance on 0.8 gm. protein and 40 calories per kilogram of
body weight, and the nontuberculous cases summarized in Table 25
is "striking. The increasing nitrogen output in both urine and stool
as death approached is also interesting.

An observation is made concerning the relation of nitrogen output
to the caloric intake which is very important and which could be
predicted from the fundamental laws governing protein metabolism
established by the earlier workers in this field. .'\s calories are added to

MARSH-XEWBURGH-HOLLY— DIABETES MELLITUS \27

the diet in the form of fat the break-down of protein as measured by
the nitrogen excretion is diminished. This ability of fat to save protein
is especially evident on Cases 5 and 11 in which with unchanged
protein intake the nitrogen elimination was markedly decreased by the
addition of fat to the diet. It is equally striking that as the diet
including protein is increased, with the chief increase, however, in fat,
the nitrogen excretion falls. In Table 30 are presented data from six

TABLE 28.— P.\RT of Record of C.vse 13; Diabetes Melutus Complicated
nv Chronic Pulmonary Tuberculosis

Date

tein.
Gm.

Fat.

GDI.

drate.
Gm.

^S-

3/25
26

is

29

28.8
29.7
28.8
30.9
29.9

138.4
141.3
138.4
112.B
140.8

20.3
21.8
20.3
16.6
19.0

1.440
1.480
1.440
1,205
1.465

Aver.
8/30 1<
9/13

29.6
34.0

134.3
160.0

19.6
27.0

1.405
1.685

16
17

34.9
33.5

38.4

179.7
158.8
142.2
165.5

27.3
26J
28.6
31.0

1.865

i.erro

1.340
1,760

Aver.

34.5

161.5

33.3

1,725

18

34.4
34.7

\^\

25.7
24.4

\^

■20

22
23
24

30.7
34.7
33.5
35.0
35.3

156.1
151.1
174.4
209.5
153.7

25.5
26.4
26.2
26.7
26.8

1.630

\^
2,130
1,630

Carbohy- _ Glu- Sugar,

:ose, Dia-
Gm. cetic

per Urine, Stool,
Cent. Gm. Gm.

1.87* 5.60 7.63 — 2.(

• Average of two days'

TABLE 29.— Summary of Case 13; Diabetes Mellitus with Pulmonary
Tuberculosis

Date

Protein,
Gm.

Fat,
Gm.

Carbohy-

Calorleg

''or-

Stool,
Gm.

8/26-8 '29
9/14-9/17
9/20-9'24

29.6
34.5
33.8

134.3
161.5
169.0

19.6
§1

1,405
1,725
l,7fiO

5.71
6.58
7.57

0.72
0.89
1.15

Out, Balance,

Death from pulmonary tuberculosta

of the cases who show this decrease in protein metabolism ; there are
added to the tables the amounts of protein burned as estimated by
multiplying the excreted nitrogen by 6.25 and the amount of glucose
rlerived from this protein calculated as 58 per cent, of the protein. A
diet too low in calories to meet the energy requirement may, because
of its accompanying excessive nitrogen metabolism, produce (from the
protein) large amounts of glucose which are avoided by the addition
of more calories. In one of our cases the glucose so derived was

128 ARCHirES OF IXTERXAL MEDICI XE

over 20 gm. To the diabetic this amount of glucose may be very
important, and may be replaced by an equivalent amount of carbo-
hydrates in the diet. By decreasing the ingested protein one may
increase the carbohydrate in the diet without' increasing the total
carbohydrate metabolism. Furthermore, by decreasing the endogenous
protein metabolism one may also increase the carbohydrate in the diet
without increasing the total carbohydrate metabolism. Both of these
desiderata may be achieved, with maintenance of nitrogenous equili-
brium, through the addition of calories to the diet in the form of fat.

These facts show a fallacy of starvation in the treatment of diabetes.
During the period of starvation, a subject well supplied with body fat
burns this fat, and burns no less than he would if the fat were given
him in the diet. This was demonstrated by Voit's experiment on a
dog which has already been mentioned. In the case of the fasting
lean diabetic, however, who cannot burn glucose, and whose supply of
body fat is low, energy and heat are developed almost entirely by the
combustion of protein. Destruction of body protein produces glucose
exactly as much as does combustion of ingested protein. In the more
severe grades of diabetes this is a factor of prime importance. Such
patients become sugar free sooner if they are allowed a little carbo-
hydrate and a relatively large amount of fat than they do if starved.
One of our patients was starved ten days on several occasions before
coming to us without his urine becoming sugar-free. On a diet con-
taining about 15 gm. protein, 90 gm. fat and 15 gm. carbohydrate, he
became sugar-free in ten days, and was in relatively excellent physical
condition at the end of the period, in contrast to his exhaustion at the
end of his periods of starvation. A diabetic boy, 6 years old, failed to
become sugaf-free after seven days of starvation following a period
of two weeks of a low caloric diet. On a diet containing 12 gm. protein,
85 gm. fat and 15 gm. carbohydrate his urinary sugar gradually
decreased in amount and finally disappeared. A young woman who
had had a constant glycosuria during ten months on a diet containing
50 gm. protein, 20 gm. fat and 30 gm. carbohydrate and who had
failed to become sugar-free on nine days of practical starvation became
sugar-free in five days on a diet containing 15 gm. protein, 85 gm. fat
and 14 gm. carbohydrate. The enormous metabolism of body protein
during starvation with the production of large quantities of endo-
genous glucose explains, in part at least, the more favorable results of
a diet relatively rich in fat.

The same undesirable production of glucose from body protein
occurs to a lesser degree when an under nutrition diet is used in the
treatment of diabetes mellitus. If the total calories fed the patient
are not sufficient to supply caloric requirement, body protein is broken

MARSH-XEWBURGH-HOLLY— DIABETES MELLITLS 129

down and glucose is produced. If an effort is made to supply enough
protein in the diet to compensate for this excessive destruction of
body protein, the ingested protein is a source of glucose. Just in so
far as the carbohydrate burning function of the patient must be used
for the combustion of glucose derived from protein, just so much
more must his carbohydrate intake be limited. Fat offers the best
agent in the diabetic for the sparing of protein, either endogenous or
exogenous.

The increase in the metabolic rate, due to the specific dynamic
action of protein, has been discussed.

That this increase in metabolism is not insignificant in such a
condition as diabetes mellitus in which an effort is made to establish
the metabolism at a low level is readily seen from a single example.
.\ssuming that a fasting subject requires 2,000 calories a day, and

TABLE 30.— Decre.ase in X-Rxcrktiox with Incre.vse in Caloric Int.ske

Protein,

Glucose

N In.

N Out.

Metabolism,

trom Protcii

Calorics

Gm.

Gm.

Gm.

Gm.

960

2.64

6.24

39.00

22.62

980

2.61

6.87

42.91

24.91

1,050

2.56

6.21

38.81

22.61

1,590

5.42

5. 72

35.75

20.73

1,680

5.49

5.63

35.19

20.41

9V0

2.61-

13.85

86.59

50.22

1.400

4.77

11.81

73.81

42.81

2,2«)

6.53

8.42

52.62

30.52

2,430

8.66

8.12

50.75

29.43

1.410

3.98

7.65

47.81

27.73

1.925

5.89

6.S1

39.64

22.99

1.925

5.89

5.77

36.08

20.91

475

2.46

9.34

68.38

33 86

i.3a>

5.99

8.82

55.13

31.96

1.380

5.99

8.10

50.63

29.37

1.650

6.05

6.62

40.97

23.76

1.500

.'i.42

6.74

42.13

24.44

2.265

8.22

5.00

3I.2.i

18.12

4.93

6.96

43.50

25.23

liM^

4.77

5.37

33.56

19.46

1.795

4.80

4.88

30.37

17.61

that the heat eliminated as a result of the specific dynamic action of
the foodstuffs is for protein, 30 per cent., fat 12 per cent, and carbo-
hydrate 6 per cent., one may easily calculate the actual increase over
the 2,000 calories that will result from various types of diets. A diet
of the von Noorden type will contain 2,000 calories if they are divided
as follows : protein, 200 gm. ; fat, 120 gm. ; carbohydrate, 30 gm. When
the calories resulting from the specific dynamic action be added, how-
ever, the diet becomes the following: Protein, 286 gm.; fat, 136 gm.;
carbohydrate, 32 gm. ; calories, 2,496. The original calories are, on the
other hand, contained in a diet which still contains but 30 gm. carbo-
hydrate, but in which the protein and fat amount to 35 and 193 gm.,
respectively. The diet resulting from the addition of extra calories is
as follows: Protein, 50 gm. ; fat, 219 gm. ; carbohydrate, 32 gm. ;
calories, 2,299. In the case of the high protein diet, the increase was

130 ARCHirES OF IXTERXAL MEDICI XE

500 calories, or 25 per cent, of the fasting requirement, and in the
case of the high fat diet, the increase was 300 calories, or 15 per cent.
In passing, attention should be called to the well known fact that
nitrogen in the food in excess of the body requirement is excreted
and is not stored. This is true in the diabetic as in the normal. In
Case 6, for example, the patient was given 65 gm. protein in his food
daily ; during the first few days of this diet he had a positive nitrogen
balance of more than 2.5 gm. daily, but after a few days the nitrogen
excretion rose until he was established in practical balance. The high
protein diet which was given Case 7 has already been discussed and it
was pointed out that there was no real addition of nitrogen to the body.
Furthermore, this patient was in balance at one time on 2i2) gm. protein
a day and later on 55 gm. protein a day. Since the excess of nitrogen
above the requirement is excreted there is no apparent advantage in
feeding the diabetic patient large amounts of protein. On the con-
trary, as has been pointed out, this excess of protein is undesirable
in the diet of the diabetic.

CONCLUSIONS

1. Nitrogen balance can be established in the diabetic according
to the laws applicable to the normal subject provided his total caloric
requirement can be satisfied. This implies that he can burn enough
glucose to metabolize fat. Diabetics who cannot burn this small amount
of glucose are extremely rare.

2. Protein metabolism above the minimal is undesirable in the
diabetic because of (1) the great glycogenic property and (2) the
large specific dynamic action of protein. Excessive protein metabolism
results from a diet containing either too much protein or too few
total calories.

MYCOTIC EMBOLIC ANEURYSMS OF
PERIPHERAL ARTERIES *

de WAYNE G. RICHEY, B.Sc.. M.D.

AND

W. W. G. MACLACHLAX. M.D., CM.

PITTSBURGH

The pioneer investigators of the etiologic factors and pathologic
processes in arteries have, in a large measure, centered their attention
on the more imminent problems of the subject, so that today a wealth
of information has been contributed to our knowledge of syphilis and
sclerosis of vascular channels. Again, the results of more acute bac-
terial invasion on the wall of both the aorta and the peripheral arterial
tree, as well as the various manifestations of the same organisms on the
different arterial systems of the tree, have been studied carefully by
others. This is especially true in the case of acute rheumatic fever,
with or without an attending acute or subacute bacterial endocarditis,
and, to a lesser extent, in other acute affections, notably scarlet fever,
septicemia, typhoid fever and pneumonia. Thus our understanding of
the modus operandi of acute bacterial invasion of the peripheral arteries
has broadened. It is only natural that more accurate data on the latter
manifestations of these infections have gone far to clarify our con-
ception of the detailed materies morbi of such a condition as aneurysm.
\\'hile the greatest stress has been placed on aneurysm of the aorta,
particularly on aneurysms of syphilitic origin, nevertheless, a certain
light has been thrown on the aneurysms as encountered in the arteries
of a smaller caliber whether specific or nonspecific.

L. Koch,* in 1851, described a case of ruptured aneurysm of the
superior mesenteric artery associated with a verrucose aortic endo-
carditis. Two years later, Tufnell - called attention to the co-existence
of aortic endocarditis and popliteal aneurysm. In 1864, Chauffard '
reported a similar aneurysm of the superior mesenteric artery while
three years after Waterman * discussed one of the brachial artery
in a case of valvular disease of the heart. While these cases suggested
that the aneurysms were mycotic-embolic in origin, Ponfick,'' in dis-

* From the Magee Pathological Institute, Mercy Hospital, Pittsburgh.
*Read before the Association of American Pathologists and Bacteriologists.
Cleveland. March. 1921.

1. Koch. L. : Inaug. Dissertation. Eriangcn. 1851.

2. Tufnell: Dublin Quart. J. M. Sc. 15:371, 18S3.

3. Chauffard : Union med.. 1865. p. 54.

4. Waterman : Western J. M.. 1867. p. 584.

5. Ponfick: Virchows Arch. f. path. Anat. 58: 1873.

132 ARCHIVES OF IXTERXAL MEDICINE

cussing an aneurysm of the superior mesenteric artery in a patient with
recurrent endocarditis, emphasized the mechanical effect rather than the
infective character of the embolus. Legroux '* recounted the instance
of an infected embolus in the axillary artery, occurring in a girl with
endocarditis. The embolus gave rise to an arteritis and resulted in
aneurysm. Again, Jacobson ~ believed that his case of double aneurysin
of the superior mesenteric artery precluded the possibility of an embolic
origin. In this case there was a concomitant aortic endocarditis in
which streptococci were isolated from the valvular vegetations. On
the other hand, Humphry * described an instance of multiple embolic
aneurysms of the pulmonary artery which had their genesis from
vegetations around the orifice of the pulmonary artery. It was for
Eppinger," in 1887, however, to give the results of the first compre-
hensive study of a group of aneurysms of peripheral arteries which
he called '"mycotic-embolic." Since then analagous cases have been
described by Lazarus,^" Nasse," von Gabriel,'- Libman,'^ Schmey,"
Routier," Weinberger,'" Roger and Gouget," Lewis and Schrager,'*
Wieland," Lindbom ^" and others. Save for the endeavors of the last
three authors, few attempts have been made to collect the reported
cases. At the present time it would appear that over a hundred
well-authenticated instances of mycotic-embolic aneurysms of peri-
pheral arteries are on record. Although no artery or system of
arteries is exempt, this type of aneurysm apparently occurs most fre-
quently in the superior mesenteric artery.

The term "mycotic-embolic" aneurysm, as coined by Eppinger,'"'
implies that two factors are operative — the embolus and the mycosis
or infection. It is imperative that this idea be borne in mind for it
is recognized that an embolus which is not infective may produce an
aneurysm, as in the case of a calcareous plac|ue cutting the intima
(Libman '=') and that all mycotic aneurysms are not necessarily embolic

6. Legrtui.x : Semaine mcd.. 18cS4, p. 425.

7. Jacobson:: Rull. Sec. Anat., Par. 72:S69, 1897.

8. Humphry; I. Path. & Bacteriol. 17:212. 1912.

9. Eppinger: .\rch. f. klin. Chir. 35:1, 1887.

10. Lazarus: Berl. klin. Wchnschr.. 1891, p. 41.

11. Nasse: Deutsch. klin. Wchnschr. 24:259, 1892.

12. Von Gabriel: Zentralbl. f. Chir. 31:2, 1904.

13. Libman: Tr. New York Path. Soc. 88-91. 1505.

14. Schmey: Cited by Roche and Burnaud : Semaine med. 28:145. 1908.

15. Routier: Semaine med., 1905, p. 306.

16. Weinberger: Mitt. d. Gesselsch. f. inn. Med. u. Kindcrh. 5:4, 1906.

17. Roger and Gouget: Nouveau traite de medicine et de therapeutique de
Bronardel et Gilbert, Paris 24:25, 1907.

18. Lewis and Schrager : J. A. M. A. 53:1808 (Nov. 27) 1909.

19. Wieland: Mcd. Cor.-Bl. d. W\irttemb. arztl. I.andesvcr. Stuttg. 82:565.
1912.

20. Lindbom: Mitt. a. d. Gronzgtl). d. Mcd. u. Chir., Jena, Separatabdruck,
1914.

RICHEY-MacLACHLAX— MYCOTIC EMBOLIC ANEURYSM 133

in nature, the infection being capable of entering the arterial wall by
direct extension from the adventitia, by way of the vasa vasorum, as
in the acute mycotic aneurysms of aorta (Osler,^^ McCrae,-- Klotz -^)
or from the lumen at the point of contact of a septic embolus or
thrombus. Not only have there been misconceptions and misunder-
standings in the interpretation of these cases but also considerable
controversy has arisen and opposite views have been offered to explain
the sequence of events after the lodgment of the bacteria laden
embolus. Consequently, we feel that a consideration of at least the
possibilities of the order of pathologic events and an adaptation of the
experiences and observations of other workers to the two such cases
we have encountered at necropsy will not be amiss.

Fig. 1. — Case 1. Ventral surface of mesentery showing branches of superior
mesenteric artery and hemorrhage into mesentery.

REPORT OF CASES

Case 1. — .\ male, aged 38, was admitted to the hospital complaining of
coughing, shortness of breath and intense pain in the epigastrium.

Previous Illness.— "Vhtt history of his ilhiess showed that he had had a num-
ber of attacks of "rheumatism" and that on more than one occasion there had

21. Osier: Allbutt's System of Medicine 6:620. 1909.

22. McCrae: J. Path. & Bacterid. 10:.373, 1905.

23. Klotz: J. Path. & Bactcriol. 18:259, 1913.

134 ARCHIVES OF INTERNAL MEDICINE

been signs of a cardiac decompensation. The intense pain in the epigastrium,
however, had been present only with the last attack.

Physical Examination. — The physical examination revealed an endocardial
lesion of the aortic and the mitral valves associated with a inoderate cardiac
hypertrophy. The lesion was a regurgitant one at both orifices. The pulse
was very rapid, between 120 and 140. weak and collapsing. The temperature
was 102 F. There was some hypostatic congestion at the bases of the lungs.
The abdomen was very painful to touch, and the abdominal muscles were quite
rigid so that accurate palpation was difficult. The leukocyte count was 10,000.

Clinical Course. — The pain in the epigastrium, which commenced about
twelve hours before his admission, persisted until his sudden death — a few
hours after coming into the hospital. Although the pain gradually became
less acute, at the onset it was very intense.

Necropsy. — At necropsy it was found that death had followed the rupture
of an aneurysm of a branch of the superior mesenteric artery with a resultant
bemoperitoneum. The heart showed hypertrophy and dilatation with a vege-
tative aortic, mitral and mural endocarditis, superimposed on a chronic sclerotic
process of these valves. Passive congestion was present in the lungs and in
the liver.

Description of Aneurysm. — The aneurysmal pouch was situated on the
dorsal surface of the mesentery, near the attachment to the coils of the small
intestine. It was irregularly oval in shape and measured 2.8 by 2.5 by 1.3 cm.,
being of such size as readily to accommodate a large marble. On the ventral
surface, the aneurysm was supported by a massive blood clot while on the
dorsum, where a ragged, linear slit indicated the site of rupture and the cause
of the bemoperitoneum, it lay very close to the peritoneum. The aneurysm
was continuous with a terminal branch of the superior mesenteric artery. The
wall of the aneurysm was thin, though reinforced posteriorly by blood which
had escaped into the mesentery, causing a marked thickening of it. The lining
of the sac was roughly corrugated and consisted, for the most part, of a thick,
friable layer of pinkish gray, granular organized blood clot. An intact rim
of intinia remained at the point where the artery entered the aneurysm. At
a slightly higher level, another irregularly oval cavity, which could have easily
admitted a hazelnut and represented a false aneurysm, occurred. It was
encompassed by a thick layer of extravasated blood clot. The lining of the
second cavity consisted only of several lamellae of well organized blood. Both
cavities were connected by a small ragged opening. It was obvious that the
true aneurysm had slowly ruptured at an earlier date with the formation of
a false aneurysm and wholesale extravasation of blood into the mesentery.
The process, as now seen, was relatively late so that the former embolus had
disappeared in the thrombus. There was a stenosis of the artery at its entrance
to the aneurysm. Above this, the arterial wall was thick and the intima was
nodular.

Case 2. — The second aneurysm, that of the right posterior tibial artery,
was recovered at necropsy from a man, aged 39. who for twenty-six weeks
had pursued a typical clinical course of subacute bacterial endocarditis.

History. — He entered the hospital six weeks before his death, complaining
of weakness with a little cardiac distress, particularly on moving in bed. .\t
times he suffered from chilly .sensations and felt feverish in the afternoon.
This had been his condition since the onset of his illness, along with a pro-
gressive weakness. At this time the patient's temperature varied from 101 to
103 F., and the pulse from 120 to 130. There was a marked pallor.

Physical Examinalion. — The heart was somewhat enlarged and presented a
double murmur which could be heard over the whole of the precordium, in
the left axilla and at the bark. The liver was pal|>able immediately below

RICHEV-MacLACHLAX-MVCOTIC EMBOLIC AXEURVSil 135

the costal margin. Lying behind the head of the right fibula and just below
it, there was an expansile round mass about the size of a small tangerine
orange. One could palpate very easily a considerable portion of its surface.
Its expansile pulsation was synchronous with the arterial pulse. This pulsat-
ing mass did not alter, to any great extent, the contour of the leg, although
its pulsation was quite visible. The right leg below this appeared normal in
every respect.

Laboratory Exainimtion.— During the patient's stay in the hospital, which
was a little over six weeks, Streptococcus salivarius was isolated from the
blood stream on three occasions. The leukocyte count was about 12,000.
Albumin, casts and red blood cells were present in the urine at various times.

Fig. 2. — Case 1. Dorsal surface of mesentery showing ruptured mycotic-
embolic aneurysm of superior mesenteric artery.

Clinical Course. — Two weeks before his death he complanied of a sharp
lancinating pain in the region of the pulsating mass below the right knee.
Almost immediately afterward a diffuse indurated swelling, which gradually
extended to the lower portion of the calf of the leg, appeared. The leg
became very painful, quite hard and of a dusky brown color. It was now
impossible to outline, by palpation, the pulsating mass behind the head of
the fibula, although one could still feel the pulsation here, as well as in the
dorsalis pedis artery, and the foot appeared warm. In a few days, however,
the foot became edematous and it was then impossible to palpate this artery.
Even to the end there was no evidence that the circulation in the foot had
been cut off. One week after this accident had occurred, signs of fluctuation

136 ARCHU'ES OF IXTERXAL MEDICINE

were present in the calf of the leg. It was considered wiser to refrain from
an incision. The patient became more septic, dying about two weeks after the
rupture of the aneurysm in the leg.

Necropsy. — At necropsy, the enlarged heart showed an acute and subacute
vegetative mitral endocarditis, which had been planted on the thick, sclerosed
valve cusps. Recent infarcts occurred in the spleen and kidneys.

Description of Aneurysm. — At the inferior margin of the right popliteus
muscle, and lying posterior to the interosseous membrane, was an aneurysm
which had evidently arisen from the tibial artery, a few centimeters distal to the
bifurcation of the popliteal artery. The aneurysm was large, irregular in outline
and presented three definite sacculations. Into the largest of these the artery
opened from above. This compartment was the size of an English walnut, pos-
sessing a thin wall and a smooth, white, corrugated lining. With this main por-
tion, two smaller sacs were associated. One. the size of a hazelnut, was tortuous
and lined by a gray, laminated blood clot so that its wall was thick and stratified.
The other sacculation, which was the smallest, lay posteriorly and inferiorly to
the medium-sized one and its lining was identical with that of the main chamber.
In the dependent tip of the smallest sac, a tiny opening was all that remained
of the distal portion of the artery which had disappeared in the intense sec-
ondary cellulitis of the surrounding soft parts. Posteriorly and above, the
main aneurysm had ruptured, as a result of which a laminated blood clot,
the size of an orange, intervened between the exterior of the true aneurysmal
sac and the soleus muscle. All the sacs contained dark red fluid blood, in
addition to the thrombus which had obscured, no doubt, the original embolus.
The popliteal artery was thick-walled and its intima was nodose. Where the
posterior tibial artery entered the aneurysm, its thick, fibrosed wall was con-
stricted, causing a narrowing of the lumen. It was clear in this case also,
that a rupture of the true aneurysm with slow leakage into the contiguous
soft parts had led to the formation of a false aneurysm. In addition, an
extensive secondary suppurative cellulitis had wrought havoc with the leg from
the knee to the foot. The popliteal vein showed no demonstrable change.

Microscopically, the sections of the walls of these two aneurysms presented
an almost identical picture. The walls were made up chiefly of a relatively
recent fibrous tissue. In the fibrous tissue, one was able to distinguish some
smooth muscle, representing the remains of the muscular coat of the artery.
Attached to the inner lining was a hyaline thrombus, well-infiltrated with
polymorphonuclear leukocytes, lymphocytes, plasma cells and numerous endo-
thelial cells containing blood pigment. This cellular infiltration extended
through the wall to its periphery where there was considerable hemorrhage
and marked invasion of all types of inflammatory cells. On staining for
elastic tissue, remnants of the elastica interna were occasionally seen. In the
deeper portions of the wall a few broken elastic fibrils were found. In the
second case, gram positive cocci in short chains were noted in the cellular
infiltration at the periphery of the wall. The aneurysmal wall, therefore, con-
sisted largely of chronic inflammatory tissue. The finding of muscle and
elastic tissue indicated that the media of the artery had shared in its formation.

Comment. — A comparison of the salient features of these two cases
shows that in both an aneurysn;i of a peripheral artery occurred, one
in the superior mesenteric and the other in the posterior tibial artery.
Both were associated with a definite acute and subacute vegetatve
endocarditis of the mitral or aortic valves. In one case infarcts were
found in the spleen and kidneys. Streptococcus salivariits was isolated
from the blood stream of one case during life. Unfortunately, the

RICHEY-MacLACHLAX— MYCOTIC EMBOLIC AXEURYSM 137

sudden death of the other case prohibited antemortem cuhures, while
the autopsy cuhures, taken 24 hours after death, showed only secondary
invaders. Xo suggestion of syphilis was found in either case at
necropsy. Both aneurysms had ruptured, at first slowly, with the
formation of a false aneurysm. Clinically the rupture of the aneurysms
was characterized by severe, sudden, lancinating pain, which persisted.

Fig. 3

2. Mycotic-embolic aneurysm of posterior tibial artery.

While no embolus was found at the aneurysmal sites in either case,
it must be remembered that the aneurysms were far advanced and
that well organized thrombi were present in them. This, in itself,
could cause a complete disappearance of the original small embolus
a 'id does not preclude the liklihood of the one-time presence of the

138 ARCHirES OF IXTERXAL MEDICI. \E

embolus. From the evidence at hand, it seems clear that both aneurysms
had their beginning in the bacteria-laden embolus which was swept
off the affected heart valve, lodging at the bifurcation of the artery
involved.

As had been stated, the aneurysms, as they came to us, represented
the end results of an acute inflammatory process wherein, at this time,
it was impossible to recognize the sequence of events from the primary
endarteritis produced by the infected embolism. It has been definitely
shown by Mrchow -* and Klotz -^ that a primary inflammation of
the intima can occur and that the cellular exudate found in the intima
appears to arise by a direct immigration of the wandering cells from
the lumen of the artery. Eppinger," in his exhaustive studies of
mycotic-embolic aneurj'sms associated with acute vegetative endo-
carditis, held that the infected embolus produced an inflammation of
the vessel wall at the point of contact. He believed that the infection
spread from this point in the intima to the adventitia and that the
aneurysm was due to the inflammatory process extending inwards from
adventitia to the media, causing a rupture of the internal elastic lamina
with subsequent arterial dilatation. Eppingers' work has been quite
generally accepted, although other conceptions of the mode of forma-
tion of these aneurysms have been advanced. Benda "'' and, later,
Wieland ^^ considered the condition to be secondary to an extension of
the inflammation in the intima through the elastica interna to the media
and then to the adventitia, regarding the primary intimal lesion as an
ulcerative endarteritis. In support of this view, McMeans -' has shown
that when the process is particularly acute and incited under condi-
tions of septic thrombosis or infected embolus, there is destruction of
the intima with extension of the inflammation through the vessel wall.
On the other hand, where the process is not so acute, Klotz -' has
demonstrated that both the intima and the adventitia can be involved
simultaneously and independently, the media escaping — stating that the
infection reaches the intima from the lumen and the adventitia from
the vasa or perivascular lymphatics. Furthermore, McMeans,-" in
studying the vascular changes in the meningeal arteries in septic men-
ingitis, mentioned the primary polymorphonuclear infiltration of the
intima with proliferation of the fixed cells of this coat and the extension
of the process to the internal clastic membrane, with the later involve-
ment of the media from both adventitia and intima. Moreover, inflam-

24. Virchow: Virchows Arch. f. path. .\nat. 77:380. 1879.

25. Klotz: Ref. McMeans, J. M. Research 32:388, 1915.

26. Benda: Ergebn. d. allg. Path. u. Anat. 8:236, 1902.

27. McMeans: J. M. Research 32:388. 1915.

28. Klotz: Brit. M. T. 1:1767, 1906.

29. McMeans: Am. T. M. Sc. 151:249. 1916.

RICHEY-MacLACHLAX— MYCOTIC EMBOLIC AXEURYSM 139

mation of both the intinia and adventitia has been produced experi-
mentally by Klotz -' and others, employing B. typhosus and strepto-
coccus in rabbits, by Saltykow ^" with Staphylococcus aureus and by
Sumikawa ^" with turpentine and silver nitrate, where the inflammation
occurred in all coats or in the intima alone. Benda ^^ further believed
that some of the aneurysms were metastatic in origin, the infection
being carried to the wall through the vasa vasorum. Thus it would
appear that the infection in a given case, as suggested by Chiari "^ in
discussing tuberculosis of arteries, can and does occur either from
within outwards or without inwards, whether carried by \asa or the
perivascular lymphatics, as in periarterts nodosa (Klotz '^) or by direct
extension from without, as in the instance of secondary involvement
from a neighboring tuberculous process — a view concurred in by Hay-
thorn ^* in reporting such a condition. It is our belief that the ideas
of both Eppinger and Benda are applicable to certain cases and, in
addition, it is conceivable that the infection in the case of mycotic-
embolic aneurysms may attack the artery from within outwards with
direct extension from the intima through the elastica interna to the
media and adventitia or, with the accompanying devitilization of the
tissues at the point of lodgment of the embolus, any bacteria which
may be present in the blood stream can enter the wall by way of the
vasa and the mainstay of the arterial coat be approached, therefore,
from both sides.

There are certain clinical phases of this subject which we would
like, in conclusion, to emphasize. Mycotic-embolic aneurysms, though
a very definite clinical entity, should be regarded only as an arterial
manifestaton of the disease — "subacute bacterial endocarditis." This
fact is important to remember, especially in reference to treatment.
Some peripheral aneurysms, not of this type, can be treated radically
by surgical measures but in these cases the local condition is not com-
plicated by an endocardial infection. Radical surgical treatment of
the mycotic-embolic aneurysm, as a rule, is a useless procedure because
the' cardiac condition remains unaltered. From the histories of cases
of surgically treated peripheral aneurysms there is good reason to
believe that a certain number were associated with an active endocardial
lesion. The end result in these cases is practically always bad as the
patient usually dies eventually from his endocarditis, if not from the
immediate effects of the operaton. Finally we would call attention to
the .symptom of sharp stabbing pain at the time of rupture of the

.30. Saltykow: Beitr. z. path. .Anat. u. ?.. allg. Palli. 43:147, 1908.

31. Sumikawa: Beitr. z. path. Anat. u. z. allg. Path. 23:242, 1903.

32. Chiari : Verhandl. Dcutsch. .\crzte in Prag., Dec. 5, 1902.
3i. Klotz: J. M. Research 37:1. 1917.

34. Haythorn: J. .\. M. .\. 60:141.3 CMay 10) 191.3.

140 ARCHirES OF IXTERXAL MEDICINE

aneurysm. The pain is more severe and prolonged than that due to
embolism which is a common occurrence in this form of endocarditis.
Pain of this intense and persistent character in a case of subacute
bacterial endocarditis may therefore be suggestive of a rupture of an
aneurysmal sac. With the knowledge that infective embolism is of
frequency in this disease, one wonders that mycotic-embolic aneurysms
of peripheral arteries are not encountered more commonly. No doubt a
certain number are overlooked so that it behooves the clinician and
pathologist to be alert to the probabilities of this condition.

BOOK REVIEWS

NUTRITIOX AND CLINICAL DIETETICS. By Herbert S. Carter. M.A..
M.D., Assistant Professor of Medicine. Columbia University. Second
edition, thoroughly revised. Pp. 681. Philadelphia and New York: Lea
& Febiger, 1921.
This volume affords an authoritative treatise on the highly important sub-
ject of nutrition and diet in health and disease. The book is divided into four
parts. Part 1 deals practically entirely with food and nutrition in health.
Part 2 deals with food, per se. Part 3 describes feeding in infants and chil-
dren. Part 4 is devoted entirely to the subject of diet in disease.

The authors have brought the subject matter up to date and have added
much new material in the chapters on chemistry, metabolism, and physiology
of digestion, vitamins, etc. Much obsolete matter has been eliminated. In the
last chapter of the book there are numerous tables which are accurate and
verj- valuable for quick reference. The book is very readable and should be
part of the armamentarium of most any medical practitioner.

LA GENESE DE L'ENERGIE PSYCHIQUE. J. D.\nysz. Librarie J. B.

Bailliere & Fils, Paris.
M. Danysz. a scientist of renown in France, has. after devoting a large part
of his life to material investigations in the Pasteur Institute, turned to less
concrete and more speculative fields. Evolution has always attracted him. as
seen in two volumes dedicated to the principles of evolution in infectious dis-
eases and to the origin, evolution and treatment of noncontagious diseases.
This has led him to fields purely metaphysical, and the present volume is a
scientific and scholarly attempt to define the status of the human mind in
relation to its surroundings — geological, biological, and sociological — past,
present and future. Human intelligence, developing very slowly during the
vast prehistoric ages, progresses with a constant acceleration, particularly in
the recent, relatively short space of our authentic knowledge, and leaves us
at our present point aghast at the possibilities, even certainties, of the future,
with a mental power that will surpass physical forces. Present man is merely
a momentary stage, will be rapidly surpassed, and man of the future will
become, through this accelerated intellectual evolution, vastly superior — in fact,
may dominate in the conquest of nature.

These seemingly startling predictions are no mere conclusions reached after
hasty and enthusiastic considerations — but are deductions drawn from evidence
presented by a large, sound and unusually complete summary of our existing
state of knowledge. Inorganic chemistry, selective solubilities, chemical sta-
bility, followed by colloidal states of activity, are thoroughly considered before
the biological realm is entered. Here the ordinary biological principles are
discussed, followed by a minute consideration of the possibilities of evolu-
tionary progression and the forces for and against such progress, including
anaphyla.xis. the role of vitamincs, and the final result, the individual animal,
later the various species of animals.

The final chapter is concerned with conclusions more of an ethical peda-
gogical, and psychological character — the differentiation and classification of
individuals according to their reactions and the motives stimulating them, with
an attempt to define the type most worthy of bearing the torch onward.

The volume is far above the level of the ancient metaphysicians, in spite
of the rather startling conclusions. Fundamentally it is sound. There is no
one point at which one can disagree or take exception — liut though he has made
a brave effort to Ijridge that seemingly limitless abyss lictween human anatomy

142 ARCHIVES OF IXTERXAL MEDICINE

and physiology on the one side, and the cerebral processes on the other,
M. Danysz has only led us a step nearer our own brink, from which we may,
on arriving, possibly see across.

THE BLOOD SUPPLY OF THE HEART (In Its Anatomical and Clin-
ical Aspects). By Louis Gross. M.D., CM., Douglas Fellow in Pathol-
ogy, McGill University and Research Associate Royal Victoria Hospital,
Montreal. With Introduction by Horst Oertel. Strathcona Professor of
Pathology, McGill University, Montreal. Pp. 165. 34 illustrations. New
York, Paul B. Hoeber, 1921. Price, $5.00.

This monograph is a comprehensive work including a critical survey of the
literature. The points of the author are well illustrated. The work is divided
into eight chapters. In the first chapter the author discusses the various
methods that have been employed in the study of the blood supply to the heart,
and in conclusion describes his own technic, which is an improvement over any
heretofore reported. In the second chapter the blood supply to the ventricles
and auricles is described. The author points out that there is a wide varia-
tion in the distribution of the coronary arteries which he reserves for consid-
eration in the third chapter. He overcomes this difficulty by describing the
theoretical heart representing the average construction from the study of one
hundred normal specimens. The fourth chapter details the result of a careful
study of the blood supply to the neuromuscular tissue. The author concludes
from the study of one hundred normal hearts that a specific blood supply e.xists
for both the sino-auricular and auriculoventricular nodes, the main bundle, the
first portion of the left limb and a large portion of the right limb of the neuro-
muscular system.

The fifth chapter deals with the blood supply to the heart valves. Si.x per
cent, of the hearts studied showed valvular injection; of these the aortic cusp
of the mitral valve was involved most frequently. The author believes that
in those instances in which the valves were not injected, vessels did not exist.
These conclusions were substantiated by microscopic examination of serial
sections of the valves. He calls attention to the presence of musculature in
the valves with blood vessels, and points out that the incidence of endocarditis
is strikingly closely related to the existence of these two structures. He sug-
gests that this may explain the frequency of right sided endocarditis in the
fetus, left sided endocarditis in the child and the relative infrequency of this
condition in adult life.

In the sixth chapter the anastomoses between the coronary arteries are
discussed. The conclusions reached are largely confirmatory of those of for-
mer investigators. In the seventh chapter the venous supply of the heart is
described. The eighth and final chapter deals with the changes in blood supply
to the heart incident to different age periods. The author shows that in
infancy and early childhood the vascularity is greatest on the right side. As
the age advances the blood supply to the left side gradually beci^mes greater.
Finally, in the sixth and seventh decades there is a great preponderance on
the left side with a relative anemia of the right side. He attributes a part of
the abundant vascularity of the left side to the development of the arteriae
telea adiposae. or fat vessels. He considers this a compensatory mechanism in
this critical period of life. He suggests that the insufficient blood supply to
the right side may account for some of the sudden deaths in the aged and the
high mortality from pneumonia at this period.

The most notable addition to the present conception of the blood supply of
the heart is supplied by Chapters 4, S and 8. This work will be especially
interesting and helpful to the anatomist and the physiologist in the investiga-
tion of fundamental problems concernmg the heart. It will, perhaps, be equally
interesting and valuable to the pathologist and clinician in a more comprehen-
sive understanding of the pathologic processes and the clinical manifestation
of cardiac disease.

Archives of Internal Medicine

FEBRUARY, 1922

TRACHEAL AND BRONCHIAL STENOSIS AS
CAUSES FOR EMPHYSEMA*

C. F. HOOVER. M.D

CI.KVF.LANIi J

During the past ten years much research has been done on
emphysema and asthma and on the chemistry of blood and respired
air, but few studies on the mechanism of bronchiolar spasm and
emphysema have been published. Some of the most recent work on
this subject has been based on the assumption that the time-honored
teaching of the causal relation between expiratory dyspnea and
emphysema and asthma is unassailable. To me it seems that the pre-
\ailing teaching on these subjects needs revision. To learn the
mechanism of emphysema and asthma by way of the chemistry of the
blood gases and of the respired air seems quite hopeless. When a
clear understanding of the mechanism of emphysema and asthma has
been attained, we shall be in a much stronger position to interpret our
chemical studies of the blood gases and respired air.

biermf.r's theory
The modern interpretation of bronchiolar asthma originated with
Biermer, of Zurich,^ who in 1870 published an exposition of the
subject. Biermer referred to Paul Bert as having proved that vagus
excitation causes bronchiolar contraction, although it had been sus-
pected, claimed, in fact, by Williams in 1840 at a meeting in Glasgow.^
Biermer said that the low position of the diaphragm could not be
reconciled by his contemporaries to bronchiolar spasm as an explana-
tion for pulmonary .emphysema. Bamberger believed phrenic spasm
to be the source of acute pulmonary emphysema, and Lehman took
the same view.

• From the Department of Medicine. Western Reserve University. Lake-
side Hospital.

* Read before the Meeting of the .Association of .\merican Physicians at
.\tlantic City in Xfay, 1921.

1. Biermer: Innere Medicin, Volkmann's Sammhing klin. Vortrage. No. 3.
Leipzig, 1870-1875.

2. Tenth meeting of the Association for the Advancement of Science.

144 ARCHIJ-ES OF INTERNAL MEDICINE

It is with much satisfaction that Biermer quotes the experiments of
Paul Bert.^ Biermer conceived an essential difference to exist between
stenosis in the remote branches of the bronchi and stenosis in the
trachea. He regarded dyspnea of bronchiolar stenosis as essentially
expiratory, because he believed that the active character of the expira-
tion of asthma produced a vicious cycle; an active pressure on the
lung compressed the boundaries of the respiratory units, the air cells
were subjected to compression, and the air passages connecting them
with the larger bronchi were also compressed; that is, the more active
the expiratory effort, the greater grew the resistance to exit of air from
the respiratory units. On this account the expiratory phase was sup-
posedly much prolonged. Biermer says that the difference between
tracheal and bronchiolar stenosis lies in the fact that the former is an
inspiratory and the latter an expiratory dyspnea. As an evidence of
this point of view, he calls attention to the inspiratory retraction in the
supraclavicular, suprasternal and epigastric regions in tracheal stenosis,
and contrasts it with the labored expiration of asthma ; and he calls
attention to the want of evidence of active vigor in the expiratory
phase in cases of tracheal stenosis.

NATURE OF ACUTE EMI'IIVSEMA

By way of introduction to the subject, let it be understood that we
are dealing with two phenomena; viz., bronchiolar hypertonus and
emphysema of the lungs which consists solely of an increase in the
pulmonary residual air. In such cases there is only altered physiologic
function and no anatomic disease of the lung.

Bronchiolar spasm is immediately attended with emphysema,
diminution of vital capacity, inspiratory and expiratory dyspnea and
lessened extensibility and compressibility of the lungs. Thus far, we
have no evidence that disturbances in the pulmonary circulation con-
tribute to rigidity of the lungs in acute pulmonary emphysema. The
loss in extensibility is due entirely to emphysema and narrowing of the
bronchial tree. Although we have no positive evidence for hyperemia
of the pulmonary circulation in bronchiolar spasm, there is good
negative evidence against its responsibility for rigidity of the lung.
That hyperemia can reduce the- vital capacity of the lung quite as much
as does bronchiolar spasm has often been observed. Both conditions
may reduce the vital capacity of the lung to 800 c.c. in persons who
normally may have a vital capacity of 4,000 c.c. It is, however, incom-
prehensible how such an hyperemia could occur without causing great
resistance to the work of the right heart and thus causing ectasis of
the right ventricle and right auricle. I have never seen any evidence

3. Bert, Paul : The Comparative Physiology of Respiration, 1870.

HOOVER— EMPHYSEMA 145

of enlargement of the right side of the heart in acute bronchiolar
spasm. Evidence for it has never been adduced, although in the
literature on the subject this assumption commonly appears. The
mechanism of bronchiolar spasm, with all its attending attributes, may
appear and disappear with such suddenness that it seems inconceivable
that the collective phenomena can have any other origin than a neuro-
muscular performance.

This aspect of acute emphysema is best seen in patients who have
periodic bronchiolar spasm from vagus excitation due to mediastinitis.
Such a patient came under my observation during an attack.

C.^SF. 1. — The patient was a little girl, 8 years old, who for the last two
years had been quite well, according to the grandmother's account, except for
asthma, which occurred every few weeks and lasted only half an hour. The
last attack of asthma this child had suddenly terminated while I was observing
the volume and excursion of her lungs. The lungs filled the pleural sinuses,
and the child was employing her utmost effort to breathe during both inspiration
and expiration. Each respiratory cycle exhibited her vital capacity, when
suddenly in the midst of a single respiration, the lower border of the lung
was seen to ascend 4 cm., and with this rapid disappearance of emphysema
there was a complete return of respiratory comfort.

It can truthfully be said that this child began a respiratory cycle
with emphysema, lessened vital capacity, inspiratory and expiratory
dyspnea, lessened extensibility and compressibility of her lungs, and
ended the cycle with respiratory euphoria. It is inconceivable that
such a rapid disappearance of all these symptoms could be produced
by any other than a neuromu.scular phenomenon. The recovery was
too rapid to include vasomotor phenomena, with consequent changes
of blood distribution, as among the causes of the symptom complex.
There is, therefore, very good reason for excluding all proposed pos-
sibilities except bronchiolar spa.sm as the cause of acute emphysema.
But we find the theory of the mechanism by which bronchiolar spasm
is supposed to produce emphysema and its attendant symptoms does not
withstand clinical and experimental criticism.

TR.\CIIE.\L STENCSIS

Bicrmer's theory teaches that bronchiolar spasm is attended with an
expiratory effort, which produces a vicious cycle by compressing the
bronchiolar exits from the respiratory units. If this be true, a resist-
ance anywhere in the tracheobronchial tree which is sufficient to
demand expiratory compression of the lung should lessen the volume
flow toward the latter part of expiration, as the intrapleural pressure
rises, and result in emphysema and prolongation of the expiratory
phase. If a strong muscular effort is needed to accomplish expiration
against tracheal re.si.stance, the same mechanism of expiratory stenosis
should be operative that Biermer's theory teaches for bronchiolar

146 ARCHIVES OF IXTERXAL MEDICLXE

hypertoinis. If active compression of the distended lung by the
employment of expiratory muscles produces a vicious cycle of
expiratory stenosis when there is bronchiolar hypertonus, then there
is no reason apparent why the same process should not operate when
the resistance to expiration is located in the trachea.

According to the prevailing theory, in the presence of great resist-
ance to expiration located in the trachea there should be pulmonary
emphysema and a prolongation of expiration proportionately greater
than the prolongation of inspiration, but the clinician rarely has an
opportunity to investigate this theory. By the time the patient with
tracheal stenosis comes under observation, there is nearly always a
complicating tracheobronchitis with infection of the bronchial tree.
What the clinician usually sees is tracheobronchitis complicating
tracheal stenosis. Under these conditions the patient may have much
moisture in the air spaces and there may be a complicating bronchiolar
hypertonus, so that it is quite impossible to separate the effects of the
two lesions so far as the clinical manifestations are concerned. I have
seen patients with laryngeal diphtheria who had pulmonary emphysema,
but these same patients had tracheobronchitis as well as the laryngeal
stenosis, so that one would not be justified in saying the patient's
emphysema was the result of laryngeal stenosis. It could just as well
originate from the bronchitis. However, during the influenza epidemic
of 1919 I had an opportunitv to see a case of uncom]ilicated tracheal
stenosis.

C.^SK 2. — A well-grown girl, about 17 years of age, had lieen ill for three
days with epidemie influenza. There had been a very considerable amount of
tracheitis from the start. There was some dulness at the base of the right
lung, but there was no elevation in the pitch of the respiratory sounds and
the patient was not cyanotic. Until the evening of the third day no alarming
.symptoms had developed, but about 10 o'clock on the night of the third day of
her illness, she became quite dyspneic. At 12 o'clock she had severe stenosis of
her trachea, about half way between the glottis and the bifurcation, whicli
was due to edema of the tracheal mucosa. With the laryngoscope the livid
swollen mucosa and the very narrow slit in the lumen of the trachea could
very plainly be seen. This patient was doing her utmost to breathe. She
employed a violent effort of all the inspiratory muscles during inspiration,
and likewise employed violent contraction of all her abdominal muscles and
intercostals during expiration. There was no undue prolongation of the
expiratory phase and there was no evidence of increase in the volume of the
lung. The patient died about two hours later from resi>iratory exhaustion
due to stenosis of the trachea. Init there was not the slightest evidence
of emphysema.

It might be objected that in this jiarlicnlar patient the equality of
resistance to inspiration and expiraliun was (he reason why she did
not have emphysema, but laltT we had the opportunity of seeing a
patieitt who had an expiratory resistance in the tracliea which greatly

HOOVER— EMPHYSEMA 147

exceeded the resistance to inspiration, and tliere was no pulmonary
emphysema.

Case 3. — A man having an aneurysm of the ascending arch of the aorta.
which compressed the trachea at the bifurcation, was brought into the hospital in
one of his attacks. It was quite apparent that the patient accomplished his
inspiratory act with comparatively little effort. There was some resistance
to inspiration, but the inspiratory dyspnea was not excessive, and it was quite
plainly apparent that there was less dyspnea during the latter part of
inspiration than at the beginning of the phase. In other words, as the man
lifted his thoracic cage, the compression of the trachea lessened, but during
expiration it was equally apparent that the resistance to the exit of air
increased as the expiratory phase progressed. The expiratory phase was
greatly prolonged and was attended with violent contractions of the abdominal
and intercostal muscles. The man was employing his utmost effort to e.xpel
air from his lung during the latter part of the expiratory phase. This was due
to the fact that, as the expiratory phase progressed and the thoracic cage
was driven downward on his aneurysm, the stenosis of the trachea increased,
so that the stronger his expiratory effort, the more ineffectual it became.

This was an instance of expiratory dyspnea, of a high degree, located in
the trachea, but the volume of the patient's lung was not increased during the
attacks. We had an opportunity to see him in several of these experiences,
which were very distressing and seemed to threaten the patient's life, but
the volume of the lung during the paroxysms (which lasted several hours) did
not increase. This patient's expiratory dyspnea was just about as severe
as a patient could endure and yet survive.

I could not conceive of a more successfully designed experiment
to prove the fallacy of Biermer's theory of expiratory stenosis than
this patient exhibited in his several paroxysms. It was this experience
that suggested aniinal experiments to see what effect obstruction to
inspiration and expiration, respectively, may have on the volume of an
animal's lung when the seat of the obsti^tction is in the trachea.

EXFERIMEXTAI. TRACHK.\L STENOSIS

A wooden box, 31x10x9 inches, was constructed with a shelving
top, which enabled us to seal the open top of the box with a glass
plate laid in petrolatum. With an animal confined in this box, which
served as a plethysmograph, we could connect the cannula inserted in
the animal's trachea with the exterior through a tube, which was passed
through a rubber cork placed in a hole in the side of the box.
Through a Meltzer cannula inserted in the pleural cavity, we were
able to register the intrapleural pressure by means of a tambour, which
was connected with the tube leading from the Meltzer cannula through
a cork in the side of the box. By the same means we connected a
cannula in the dog's jugular vein with the exterior. The animal could
breathe the room air, or any other atmosphere, and by connecting the
cavity of the box with the spirometer of a Benedict apparatus, we could
record the respiratory excursions and detect any inodifications in the
volume flow of air during inspiration and expiration quite as accurately

148 ARCHIVES OF ISTERNAL MEDICINE

as when the tracheal cannula was connected with the spirometer. In
fact, the plethysmographic excursions of the spirometer were perfect
reproductions of the excursions of the spirometer whea the animal
breathed directly into the reservoir. Each millimeter of elevation of
the reservoir represented a volume of 22 c.c. When the spirometer
was used for recording the plethysmographic excursions, the tubes
leading to the canister were cut off, so that there was simply a to and
fro movement of air between the Benedict reservoir and the air of the
box which contained the dog. We could thus accurately measure not
only the character of the inspiratory and expiratory excursion and
its exact volume, but we could accurately measure any variation in the
minimum volume of the animal's lung.

The tracheal cannula connected through the side of the box with
room air, and on the end of this tube one could fix various sorts of
appliance ; for instance, a screw clamp was used on a piece of rubber
tubing which was attached to the tube connected with the tracheal '
cannula, and in this way we could give the animal any degree of
stenosis to inspiration and expiration alike. By a Y tube fixed to the
tracheal cannula and clapper valves, we could give the animal any
degree of stenosis to expiration only and leave inspiration unobstructed ;
or we could do the reverse and give the animal any degree of resistance
to inspiration and leave expiration unobstructed.

The animals employed varied greatly in size, the largest weighing
seventeen kilos. Many times there was no change in the volume of
the animal, and we never got an increase in volume above 66 c.c. when
tracheal stenosis to inspiration or expiration was employed.

The dog was given chlorbutanol (dissolved in oil) in the peritoneal
cavity, and morphin was given hypodermically. \\'e attempted to use
the minimum dose of chlorbutanol and morphin, but it was necessary
to keep the animal at rest while fixed within the plethysmograph.

One of the greatest difficulties was to procure active expiration in
an anesthetized animal. Obstructing respiration and permitting the
carbon dioxid to accumulate within the body, or having the animal
rebreathe into a bag containing 80 per cent, oxygen and 20 per cent,
carbon dioxid, would not secure active expiration. Resistance to
inspiration very readily activated the muscles to increased effort, but
when tracings of intrapleural pressure were made, we found that
resistance to both inspiration and expiration was not sufficient to induce
active expiratory effort. // zvas only after a high degree of hyperpnca
was indwed by having the dog rebreathe an atmosphere of oxygen and
carbon dioxid that ive could induce an active respiratory effort by
shunting in a resistance to expiration. This is mentioned because, in
the literature on the subject of respiratory excursions, many writers

HOOTER— EMPHYSEMA 149

have assumed that, because respiration had been suspended or the
animal was compelled to breathe an atmosphere with a high concentra-
tion of carbon dioxid, therefore the hyperpnea was accompanied by an
active expiratory effort. /; zms found after vmny trials that, to procure
an active expiratory effort so that during expiration the pressure in the
pleural cavity ■would be raised above barometric pressure, it was neces-
sary first to induce a very active hyperpnea and then during the period
of hyperpnea to shunt in the expiratory resistance. Under these circum-
stances the threshold for active respiration was passed and a positive
pressure in the pleural cavity was obtained above that of barometric
pressure.

Experiment 1. — Figure 1, line A, should be read from right to left. From
1 to 2, we see four normal respiratory excursions recorded by the plethysmo-
graph. Line A is the base line and marks the minimum volume of the dog at
the end of expiration. At 2 the animal was made to rebreathe into a bag
containing 80 per cent, oxygen and 20 per cent, carbon dioxid. Each 5 mm.

. KaaaaaAaaaaaaaAaaaaaaa^

y\y\^y\^... ^ VV VV ^ , V ., . . . ^^

_j ±^ '■'^^^

Fig. 1. — Experiment 1.

in the time marker represents 1 second, so that 1 mm. equals 1/5 second.
As can be seen by the tracing, inspiration and expiration were of about the
same duration ; and while the animal was breathing room air, 1 second was
occupied to complete the respiratory cycle, and the animal breathed about
70 c. c. with each respiration.

When the dog began rebreathing the high concentration of carbon dioxid,
the time occupied in the respiratory cycle was 1 second; the rate was very
much increased, and the volume of each excursion was 286 c. c. From 2 to 4
the animal was rebreathing the carbon dioxid without resistance to either
inspiration or expiration, and then at 4 a valve was inserted in the end of the
tube connecting with the tracheal cannula, which gave the animal a perfectly
free inspiratory movement but very greatly restricted the expiration. The
volume flow through the expiratory valve under a pressure of 80 mm. of
water was 1,200 c. c. per minute. Inspiration occupied 3/5 second, but the dura-
tion of the respiratory cycle was 13'S seconds; so that the proportional duration
of inspiration to expiration was 3 to 10. The volume of excursion as the
animal began breathing at 4 was 132 c. c. As will be seen by the' tracing, in
the first five respiratory cycles there was a gradual rise in the volume of each
respiration, and then it is seen that the stimulus to respiration due to the
carbon dioxid accumulation produced a large excursion of 400 c. c. The duration
of inspiration in this cycle was 5/5 second, but the duration of expiration was

150 ARCHU'ES OF IXTERX.IL MEDICIXE

25/5 seconds, and the volume of the dog at the end of expiration had increased
154 c. c. As this tracing was continued, the respiratory excursion became
more rapid, and the minimum volume of the dog at the end of expiration had
increased about 250 c. c. above the original volume. So that we can say that
this experiment shovi's that, when an expiratory obstruction is shunted into
the trachea in an animal with great hyperpnea, the residual air in the lung
is very greatly increased. When, however, one carried out the experiment in
reverse order, and as shown in line B, the dog breathed room air from 1 to
2 and then at 2 the e.xpiratory valve was placed in the end of the tracheal
cannula, so long as the dog was breathing room air and there was no occasion
for hyperpne* the volume of the dog was unchanged. At 3, however, the
animal continued to breathe with an expiratory resistance, when he was
connected with a bag containing 80 per cent, oxygen and 20 per cent, carbon
dioxid. By the end of the seventh respiratory cycle, there was a large inspiration
of 352 c. c, and the same increase in the residual air of the dog became
apparent as in line A at 4. As seen in one of the larger excursions, the duration
of the respiratory cycle was 30/5 seconds, inspiration occupying only 4 '5
second, and expiration 26/5 seconds.

Although in this experiment the intrapleural pressure was not
measured (because I wished to leave the thorax intact), I think we
are justified in assuming from repeated trials in former experiments
that with this resistance to expiration added to the great hyperpnea
from breathing carbon dioxid, the animal was actually employing
efifort during the expiratory phase. This experiment showed that the
only way in which tracheal resistance will increase the residual air in
the lung is in conjunction with increase in the volume of respiratory
excursion. In this animal the volume of excursion was increased from
70 c.c. up to 300 c.c. by rebreathing carbon dioxid, and the rate of
respiration was increased from 36 to 52 per minute.

The phenomena described in this experiment do not occur in bron-
chiolar spasm. A patient who has a violent spasm can supply his
oxygen needs perfectly well by a 25 per cent, increase in the oxygen
consumption and the minute volume of air. I have repeatedly taken
Benedict tracings on patients during periods of uronchiolar hypertonus,
and during severe spasm we have found both the minute volume of
respired air and the consumption of oxygen just about 25 per cent,
greater than in the interims of respiratory comfort. The increase in
each is imposed on the ])atient by the additional work required to
ventilate his lungs. This inlrodiico another essential consideration of
asthmatic breathing — that is. ih.-il the residual air in llic lung is greatly
increased, although the increase in the minute volume of ventilating
air is slight. In bronchiolar hypertonus without hyperpnea. emphysema
is produced, but in tracheal expiratory dyspnea, emphysema is obtained
only in the presence of liy]ier])neri, which must be superadded to the
expiratory resistance.

The clinic.-il and experimental evidence^ reveal an essential diflfer-
ence between the results of bmnchiolar hv|iertnnus and stenosis of the

HOOVER— EMPHYSEMA 151

trachea. When the tracheal resistance to inspiration and expiration are
equal, there is no increase of residual air in the lung. In the absence
of hyperpnea, resistance to expiration in excess of that to inspiration
will not increase the volume of the lung.

BRONCHIAL HVPERTONUS. REGIONAL AND OF AXV DEGREE

The Biermer doctrine of the mechanism of emphysema in bron-
chiolar hypertonus demands an active expiratory compression of the
lungs. But moderate bronchiolar hypertonus produces emphysema
when there is no active effort attending expiration. Expiration may
remain a passive procedure and still be attended with emphysema.
When the conception of emphysema from bronchiolar hypertonus first
appealed to me as a problem for investigation, I was in the habit of
tliinking of bronchiolar hypertonus as a neuromuscular performance
that belonged to the all-or-nothing law, and also that it was equally
distributed throughout the bronchial tree. Neither of these conceptions
is true. Neuromuscular reaction in the bronchial tree may, like vaso-
motor reaction in the arterial bed, be of any degree and may also be
universal or regional. A patient may have a moderate bronchiolar
hypertonus attended with demonstrable emphysema of the lung but
without sufficient bronchiolar resistance to call out an active expiratory
effort on the part of the patient. A few clinical cases will suffice to
prove these statements.

REPORT OF CASES

C.\SF. 4. — A man, 60 years of age, who had suffered from bronchitis and
occasional asthmatic paroxysms for a period of five years, had a very moderate
pulmonary emphysema, which at the borders of the lungs was sufficient to
till the pleural sinuses, but the total lung volume was not increased so that the
diaphragm was sufficiently flattened to change the normal direction in the
movement of the costal margins during inspiration. .According to this man's
account, his breathing was never completely shunted out of the field of
consciousness. He' was always aware of a certain degree of respiratory
discomfort, but from physical examination and roentgenograms of the lung
one could not determine how much of his discomfort should be ascribed to
atrophic emphysema and how much to bronchiolar hypertonus. When he was
given a subcutaneous dose of 20 minims 1:1000 epinephrin, the vital capacity
of his lung increased from 2,000 to 2,500 c. c, and he then said that for the first
time- in a year he experienced complete respiratory euphoria. Before the
epinephrin was given, this patient did not have an active expiratory phase,
although he had a certain degree of bronchiolar hypertonus with consequent
emphysema, as shown by the respiratory comfort and increase in vital capacity
directly after adrenalin was given. Mis respiratory discomfort had evidently
consisted in an increased efTort during inspiration. Of this effort he was
conscious, but his pulmonary ventilation was still adequate to preserve the
passive character of his expiration. His expiration was prolonged, but the
moderate dyspnea permitted the expiratory prolongation.

Case 5. — .Another patient was a young, vigorous man, 25 years of age, who
complained of respiratory discomfort, which had lasted for twenty-four
hours. It was quite apparent that the man had labored breathing, but examin-

152 ARCHH'ES OF IXTERXAL MEDICINE

ation revealed that his labor was all on the inspiratory side; there was no
effort during the expiratory phase, the abdominal muscles being perfectly
relaxed. The costal margins moved outward throughout their entire extent
during inspiration. There was sufficient emphysema, however, at the borders
of the lungs so that they filled the pleural s.inuses as far as the eighth interspace
in the nipple line and the eleventh rib in the axillary line.

The patient showed very marked relief from the administration of adrenalin.
Before the administration his vital capacity was 3,350 c. c. when seated, and
2,951 c. c. when recumbent. Ten minutes after he had been given 20 minims
1:1000 epinephrin hypodermically, his vital capacity when seated was 3,478
c. c, and 3,415 c. c. when recumbent. His respiratory rate both before and after
epinephrin varied between 22 and 24. Before he had his epinephrin when the
respiration was traced on a rotating drum, which was running at the rate of
5 mm. per second, in a respiratory cycle during which he breathed 528 c. c.
the duration of the entire cycle was 22 5 seconds. The inspiratory phase was
10/5 seconds, and the expiratory phase 12/5 seconds. After he had epinephrin,
in a respiratory cycle in which he breathed 726 c. c, the duration of the cycle
was 19/5 seconds, the inspiratory phase occupying 8/5 seconds and the expira-
tory 11/5 seconds. So it is quite apparent that both before and after epinephrin
there was no active expiratory phase.

Although the test showed only a comparatively slight increase in his vital
capacity after epinephrin, the lung passively expelled 726 c. c. in 11/5 seconds,
whereas prior to the administration of epinephrin the lungs passively expelled
528 c. c. in 12/5 seconds. The rigidity of the lung was considerably less
after epinephrin than before. The inspiratory effort had been quite apparent
to inspection, and the inspiratory and expiratory phases were accompanied
by a number of coarse, moist and sibilant rales, all of which disappeared after
epinephrin took effect. The only direct evidence we procured to show diminu-
tion in volume of the lung after epinephrin was that, while the lower border
of the lung had been at tlie eleventh rib in the axillary line before epinephrin,
it was afterward at the ninth interspace. In the midclavicular line, it was at
tlie eighth interspace both before and after epinephrin was given. However,
it was quite apparent to the observer that the epinephrin relieved the man of a
very considerable degree of respiratory discomfort.

These patients with demonstrable bronchiolar hypertonus had a
very considerable degree of respiratory discomfort (that is, inspira-
tory), but the vital capacities of the lungs were very little modified by
the hypertonus. According to the size and conformafion of the second
man's thorax, he should have had a vital capacity of at least 5,000 c.c.
Without the use of epinephrin it would have been impossible to show
that bronchiolar hypertonus played a part in his respiratory di.scomfort.

The following cases showed regional hypertonus :

Case 6.— This patient had had a moderate general bronchiolar hypertonus,
but on one occasion it was only regional. The resident physician found one
night that his right thorax was comparatively immobilized, with no respiratory
excursion perceptible in the arches of the right ribs. The whole right thorax
was distinctly larger than the left. There was a distinct asymmetry, which
had not been visible on other occasions. During the inspiratory phase, the
entire right costal border was drawn strongly toward the median line, but
the left moved vigorously in a lateral direction, and the respiratory excursion
of the entire left thorax plainly exhibited an exaggerated excursion. The
arches of the left ribs moved in -their normal bucket handle manner much
more actively than during interims of comfort. It was plainly apparent that

HOOyER— EMPHYSEMA 153

the right thorax was enlarged, that the right diaphragm was flattened, and
that there was very slight respiratory excursion on that side. It was also
apparent that there was a compensatory increase in the respiratory excursion
of the left side, the minute volume of air in the left lung being equal to that
of both lungs in the interims between attacks. In other words, the man had
an acute unilateral bronchiolar spasm. This was also confirmed by auscultation.
Over the entire right side there was an abundance of squeaking and moist rales
during inspiration and expiration, but the left lung was free. The respiratory
sounds were only faintly audible over the right side, whereas they were distinctly
heard over the left, where the excursion was magnified. This patient was
probably breathing a little larger minute volume of air than he employed
during his interims of comfort, but the greater part of the pulmonary ventilation
was accomplished with the left lung.

Twenty minims of 1:1000 epinephrin were given this patient hypodermically,
and within a very few minutes he recovered his normal thoracic excursion.
The two sides were again symmetrical, with symmetrical outward movement
of both costal margins during inspiration. Most of the rales of the right side
disappeared. The patient Nvas .perfectly comfortable, and so far as physical
examination enabled us to determine, he was again ventilating both lungs equally.

C.\SE 7. — -This patient had severe bronchitis and atrophic emphysema, and
he had been studied during very many asthmatic paroxysms. On one occasion,
however, he complained of discomfort when the auscultatory evidence of
asthma was limited to the lower lobe of the right lung. The right costal
border was drawn toward the median line during inspiration, whereas the left
costal margin moved in an outward direction. Aiter the hypodermic injection
of epinephrin, the patient's respiratory comfort was restored within a very
few minutes. The auscultatory signs of asthma disappeared from the lower
lobe of the right lung, and the right costal border resumed its outward inspira-
tory excursion, which was quite symmetrical with that of the left side. So
far as physical examination enabled us to determine, this patient had a regional
bronchiolar spasm attended with emphysema restricted to the lower lobe of
the right lung.

Case 8. — On another occasion we had a patient suffering from chronic
bronchitis and emphysema who had an attack of asthma in which all the
evidence of asthma was limited to the two upper lobes. There was no evidence
of flattening of the diaphragm, but the auscultatory signs of asthma, though
very pronounced in the upper lobes, were entirely wanting over the two
lower lobes.

It does not seem possible that a patient could produce the expiratory
vicious cycle which supposedly occurs in asthma and confine the
exhibition of compression of the bronchioles in one instance (Case 6)
to the entire right lung, in another instance (Case 8) to the two upper
lobes, and still in a third instance to the lower right lobe only (Case 7).
Such cases alone offer sufficient evidence to disprove the need of an
active expiratory phase to produce emphysema during an asthmatic
attack.

DURATION OF EXPIRATION

To return to our cases of general bronchiolar hypertonus (Cases 4
and 5), attended with moderate emphysema and prolongation of the
expiratory phase. The expiratory phase in all the i)atients with mod-
erate hypertonus was longer than the inspiratory phase simply because
the resistance to expiration was not sufficient to demand an active

154 ARCHU'ES OF IXTERXAL MEDICINE

compression of the lungs. To cause a transition from passive to active
expiration, the obstruction to the passage of air must be so great that
the time required for passive expiration exceeds the tolerance of the
patient's respiratory needs. So long as bronchiolar hypertonus will
allow a passive expiratory phase, the patient will instinctively refrain
from active compression of the lungs by simultaneous action of the
abdominal and intercostal muscles. The active expiration demanded by
stenosis of the bronchial tree is very exhausting. Resistance to inspira-
tion requires only an increase in the force of activation of the muscles
normally employed ; but resistance to expiration not only requisitions
muscles which are not normally employed (which in itself is very
exhausting), but produces an unfavorable effect on the hydraulics of
the blood circulation in the thorax. In animal experiments the extrem-
ity of the measures which are required — namely, hyperpnea plus
obstruction — before the animal can be induced to raise the expiratory
pressure above the barometric pressure, plainly illustrates the height of
this threshold for active expiration. These facts clearly explain why a
patient with moderate bronchiolar hypertonus will have the duration
of the expiratory phase shortened after the administration of adrenalin.
Adrenalin given to such a patient will also have the effect of dimin-
ishing the residual air in the lung and will increase the pulmonary
vital capacity.

How is the expiratory phase affected in a patient who has a
bronchiolar hypertonus so severe that an active expiration is demanded?
Figure 2 is a tracing taken with the Benedict apparatus during an
attack of bronchiolar hypertonus. On line B is a tracing after the
hypertonus was relieved by 20 minims of 1 : 1,000 epinephrin given
hypodermically. When the tracing in line A was made, the patient
was in great distress. The volume of each lung was so increased that
the flattened diaphragm drew the costal borders toward the median line
during inspiration, and the expiratory phase was accomplished by
violent contraction of the abdominal muscles. These conditions are
exactly such (according to Biermer's doctrine) as to create a vicious
cycle of dyspnea, and should have caused a prolongation of the
expiratory phase with a lessened volume flow of air toward the latter
part of expiration. But the tracing reveals the opposite of that expected
under the Biermer theory.

Experiment 2. — This tracing was taken on a drnin with tlic speed of 1 mm.
per second. The respiratory rate before adrenalin was given was 16 per
minute. The line of the inspiratory phase on the upstroke is straight, and
the line of the expiratory phase is equally straight. There is no lagging or
horizontal drag as must occur in such a tracing should the volume flow lessen
toward the end of expiration. The transition from inspiration to expiration
and that from expiration to inspiration .show very sharp peaks. The tracing
indicates just what we were able to observe in the patient at the time. He

HOOVER— EMPHYSEMA

155

was employing his utmost muscular strength during all of inspiration and
expiration, and the volume flow of air was uniform throughout the entire
respiratory cycle.

On line B we see the effect of the respiratory cycle after the bronchiolar
spasm had been relieved by the hypodermic injection of 20 minims of 1 :1000
epinephrin. The respiratory rate was then 14 per minute. The inspiration
on the upstroke is not quite so nearly vertical as before the epinephrin was
given, and we can see toward the end of the expiratory phase a distinct slope

/ I

I .11

" ^i:!:|ii:|iMli

1/ II ! ^

,nr,i! >

m^r^'

Fig. 2.

hypertoni

A. > >

-Tracing taken with Benedict apparatus during attack of bronchiol;

to the right, which means the volume flow toward the end of the expiratory
phase was distinctly lessened. During the time the tracing in line B was taken,
the patient had distinct inspiratory widening of his subcostal angle. Expiration
was not attended with contraction of the abdominal muscles. In fact, he
was breathing with perfect comfort. The difference was that before he had
his epinephrin he was able to attain his required pulmonary ventilation at the

156 ARCHIVES OF IXTERXAL MEDICIXE

expense of great muscular effort during both inspiration and expiration. The
expiratory labor was very exhausting, although the actual increase in oxygen
consumption was only 25 per cent, in excess of what he employed during
tranquil breathing at rest. The source of exhaustion is not in this instance
measurable by the increased consumption of oxygen but by the employment
of an unusual kind of expiratory effort.

With the Biermer doctrine of expiratory dyspnea in mind, one would be
disposed to say that line B was traced before epinephrin was given, and that
line A was traced after the bronchiolar hypertonus had been relieved. This
tracing shows a longer expiratory phase after bronchiolar hypertonus was
relieved simply because the patient had a passive expiratory phase and his
respiratory needs did not demand compression of the lungs during expiration.
Tracing A not only shows that the patient employed a positive pressure in his
pleural cavity during the expiratory phase, but it shows that the minute volume
flow of air during the latter part of the expiratory phase was more constant
than when he was breathing with a passive expiration.

These experiments not only prove the fallacy of the doctrine of
expiratory dyspnea in asthma, but they also show the error of employ-
ing the prolongation of the expiratory phase as an index of the severity
of bronchiolar spasm.

EXPIRATORY COMPRESSION STENOSIS OF THE BRONCHI

The question of expiratory dyspnea arising from compression
stenosis in the bronchial tubes remains for experimental consideration.
If bronchiolar spasm is severe enough to require active compression
of the lungs in expiration, there are two forces employed to expel air
from the lungs : one the positive intrapleural pressure due to activation
of the abdominal and intercostal muscles, and the other the retractile
property of the visceral pleura and the elastic tissue within the lungs.
If active compression of the lungs produces stenosis of the small
branches of the bronchi in the presence of bronchiolar .spasm, why
should it not do the same in the absence of bronchiolar hypertonus? It
seems reasonable to suppose that in the absence of hypertonus the
bronchi would be inore compressible than when they are hypertonic
If hyperpnea is induced in an animal by suffocation or by breathing a
high concentration of carbon dioxid, and then tracheal stenosis is added
to the respiratory burden so that compression of the lung is required to
accomplish expiration, then under such experimental conditions we
can measure the volume flow of air from the trachea and also the pres-
sure on the proximal side of the tracheal stenosis, and synchronously
with these air pressure and air volume flow measurements we can also
trace the expiratory pressure in the pleural cavity. In such an experiment
the expiratory forces are pitted against the tracheal resistance, with a
certain volume flow as a result. Should an expiratory resistance to the
outflow of air intervene between the air cells and the tracheal obstruc-
tion, then the volume flow of air should lessen toward the latter part of
expiration, and the sum of the expulsive forces should exceed the air

HOOrER—EMPIirSEMA

157

pressure maintained in the trachea on the proximal side of the stenosis.
The expulsive forces are of course the pressure within the pleural
cavity and the retractile power of the lung.

Experiment 3. — A dog weighing 16 kilos was anesthetized with chlorbutanol
and morphin. Tracheotomy was then done, and two cannulas introduced, one
into the tracliea and one into the right pleural cavity, after which the dog was
placed in the plethysmograph. The cavity of the box was connected with the
reservoir of a Benedict apparatus and the passages to the canister clamped
so that there was a free movement to and fro between the confined air in the
box which contained the dog and the air in the registering reservoir. The
volume flow of air during each respiratory cycle could thus be accurately
traced on a revolving drum, which had a speed of 5 mm. per second (Fig.
3. line Pletli). The time is marked in fifths of seconds. By a tube which
passed through a cork in the side of the box, the Meltzer cannula in the pleural
cavity was connected with a calibrated tambour, which traced the pressure in
the pleural cavity for each respiratory cycle (see line Pleura). A T tube was
inserted in the tube to the pleural tambour, by means of which we could

^

h f\

f\

\

\\ !\

I

u

JVJ

jy

v_„

Vj

VJU

vJ

u

JVJUJvJvJvJUJtJ

Fig. 3. — Experiment 3.

introduce into or remove from the pleural cavity any desired quantity of air.
The tracheal cannula was connected with a tube which passed through a cork
in the side of the box, and to the end of this tube was fitted a Y tube, the
afferent limb of which was fitted with a clapper valve, while the efferent limb
had its lumen modified at will by means of a screw clamp.

On the proximal side of this Y tube which terminated the tracheal cannula,
was introduced a T tube, which connected with a calibrated tambour by means
of which could be traced the air pressure on the proximal side of the expira-
tory obstruction. The writing styles for the plethysmograph and tracheal and
pleural tambours were then perfectly aligned, so that a vertical line drawn
anywhere would enable us to make accurate comparison of the three tracings
in any phase of the respiratory cycle. In both the pleural and tracheal pressure
tracings, the horizontal line traces the line of barometric pressure. The time
tracing is in fifths of seconds. In all three tracings the upstroke records
inspiration and the down stroke expiration, and they are read from right to
left. They were made after 1,050 c. c. of air had been injected into the right
pleural cavity. This injection was to prevent any error in getting the exact
intrapleural pressure and also to procure any evidence of expiratory bronchiolar
compression which the partial collapse of the lungs might offer.

158 ARCHirES OF IXTERXAL MEDICIKE

Preceding this part of the tracing the screw clamp on the efferent limb
of the tracheal Y tube was screwed down to procure the maximum resistance
against which the animal would breathe. Respiration was arrested by complete
obstruction of the tracheal cannula for one minute, and the drum was started
as the complete obstruction of the tracheal cannula was released. The afferent
clapper valve offered slight resistance to inspiration, as shown by the rise
of the pressure curve of the tracheal tambour above the line of barometric
pressure (minus pressure). The combination of hyperpnea with resistance in
the efferent limb of the Y tube to the tracheal cannula induced an active
expiratory phase, as shown by the rise above barometric pressure in the pleural
cavity (below the pleural barometric line).

The plethysmographic tracing shows each respiratory cycle to have a duration
of 17/5 seconds, the inspiratory phase occupying 3/5 seconds, and the expiratory
14/5 seconds. The volume of each respiration was 110 c. c. The volume flow
tracing of expiration is straight : there is no drag to the left as there would be
if the flow diminished toward the latter part of the expiratory phase.

When the positive pressure below the barometric lines of pressure for the
trachea and pleura are measured, the pleural pressure of 9 mm. registered
by the tambour style is found to equal 70 mm. of water, and the 12 mm.
registered by the tracheal tambour style at the same time equals 100 mm. of
water. But to the air pressure within the pleural cavity must be added the
elastic contracile force of the lung. In this dog the intrapleural pressure
measurements were taken when the animal was breathing freely without
obstruction, and the pleural cavity was free of all but SO c. c. of air. The
intrapleural tracings then showed a pressure below the barometric pressure
of 70 mm. of water at the height of inspiration, and 30 mm. of water at the
end of expiration. If these 30 mm. are added to the maximum intrapleural
pressure attained during the expiratory phase, we find the sum of the two
factors (70 + 30) exactly equals the positive pressure measured on the proximal
side of the screw clamp on the efferent limb of the tracheal cannula, and
furthermore the minute flow is constant during the entire expiration. IVe are
justified in saying' that under these experimental conditions there can be no
resistance to the exit of air intervening between the air cells and the foint
of obstruction on the efferent limb of the tracheal cannula.

Repeatedly hyperpnea was produced by having the animal rebreathe
into a bag containing 80 per cent, oxygen and 20 per cent, carbon
dioxid, and then resistance was applied in the tracheal path. By these
means we were able to develop a high pressure in the pleural cavity
during expiration, and the same results were always procured. After
eacli experiment the tambours were calibrated for pressures above and
below barometric pressure. In the dog the mediastinal reflections of
the pleura are so mobile and of¥er so little resistance that a positive
pressure obtained on one side of the thorax by prochicing an active
expiration will be the same on both sides.

In measuring the volume of an animal under these experimental
conditions, great care must be observed to guard against leakage of
air into the pleural cavity. If air enters, the vokmictric measurements
of the plethysmograph will record an increase in volume of the thorax
which apparently varies with the resiliency of the animal's thoracic
cage. In one experiment, 100 c.c. air injected into the right pleural
cavity increased the voUmic of the dog 120 c.c. and wlicn .^00 c.c were

HOOVER— EMPHYSEMA 159

injected, the dog increased 360 c.c. in volume. If this source of error
is not considered, the increase in volume may be interpreted as an
increase of the residual air in the lungs.

Thus far the studies of tracheal obstruction show several results
that are of interest in relation to the study of resistance to the respira-
tory passage of air:

1. If the obstruction is the same during inspiration and expiration,
there will be no increase of residual air, although hyperpnea is demanded
for breathing an atmosphere of 80 per cent, oxygen and 20 per cent,
carbon dioxid. When there is no hyperpnea, equal resistance to
inspiration and expiration fails to produce emphysema.

2. When there is no hyperpnea, equal resistance to inspiration
and expiration fails to produce emphysema.

3. In an animal under chlorbutanol anesthesia, the threshold for
active expiration is very high and is passed only when hyperpnea is
added to tracheal resistance.

4. Expiratory resistance with unobstructed inspiration will not
cause emphysema unless the resistance occurs in company with
hyperpnea.

5. When active expiration is induced by the combination of
hyperpnea with tracheal expiratory resistance, there is no intervening
resistance to the exit of air from the lungs that can be located in the
bronchial tree, as Biermer's theory assumes there should be to produce
emphysema.

These experimental results are quite consistent with the previously
described observations on a patient who suffered from severe stenosis
of the trachea (Case 3), which caused equal resistance to inspiration
and expiration ; and they are also consistent with the observations
made on the patient who had severe expiratory with moderate inspiratory
resistance without the development of emphysema (Case 4) — although
the expiratory phase in both patients was accomplished with the utmost
prolonged muscular effort of which the patients were capable.

Acute emphysema was procured by animal ex])eriment only when
the hyperpnea was combined with expiratory dyspnea in the absence
of inspiratory dyspnea. On the other hand, we have .shown that
emphysema with bronchiolar hypertonns occurs in man when there is
no further cause of hyperpnea than a slight effort during inspiration,
and when expiratory dyspnea is so moderate that an active expiration
i-; not required and expiration remains a passive procedure.

F.XPERIMENT.AL BRONCIII.\L IIVPERTONUS FROM III.ST.VMIN

We should, then, expect to find emphysema in dogs accompanying
the bronchiolar hypertonus that follows intravenous injection of his-

160 ARCHIVES OF INTERNAL MEDICINE

tamin. With this in mind, twelve experiments were performed which
were designed to show modifications in the vohime of the lung of a
dog during the period of bronchiolar hypertonus. The histamin was
administered through a cannula, which was previously fixed in the
internal jugular vein ; and a tube from the cannula led through a cork
in the side of the box, so that we could give histamin without disturb-
ing the plethysmographic tracings of respiration. I was never able
to find an increase of lung volume above 60 c. c. during the period in
which the effect of histamin was apparent, and that was in a dog
weighing 16 kilos. This increase in size is practically negligible so
far as a consideration of increase in the residual air of the lung is
concerned, for such a moderate increase might occur under a variety
of disturbances in the cardiorespiratory function.

B.'

rnTTTTTTm^n'

iTrnT»ni|M||^p

Experiiiieiit 4. — In Figure 4 there is a tracing from an experiment which was
designed to show the intrapleural pressure during the period of the histamin
effect, with the idea that, if histamin produced a decided bronchiolar hypertonus,
then there should be a lessened respiratory excursion in the presence of a
lowering minus pressure within the pleural cavity. In this experiment the
plethysmograph was not used. The animal breathed through a tracheal cannula
directly into the Benedict reservoir. The tracing reads from left to right, and
the time-marker is in fifths of seconds. Lines A and A' were taken simul-
taneously. The horizontal line marks the line of barometric pressure in the
pleural tambour. Line A' was a simultaneous tracing in the Benedict apparatus.

.^fter the injection of 5 c.c. of 1: 10,000 histamin into the jugular vein, the
respiratory rate became very rapid and the volume of excursion was very small.
To the right it will be seen that, during the period in which the pleural tracing

HOOVER— EMPHYSEMA 161

shovved a very great descent in pressure, the volume of tidal air in the lung
diminished. In fact, when the excursion of pleural pressure was very consider-
ably extended, from the barometric line to a considerable distance below, the
excursion of the lung was quite small. So we are certainly justified in saying
that in the presence of a low barometric pressure in the pleural cavity, and
also in great excursions of the pressure within the pleural cavity, there was a
marked rigidity of the lung.

Line B and B' are tracings taken from the same animal after a second
injection of the same dose of histamin. The beginning of the histamin effect is
apparent at H in the tracings both of pleural pressure and of lung excursion.
The two experiments show essentially the same thing.

Thus far our investigation has shown that equal tracheal stenosis
to inspiration and expiration in both men and dogs failed to produce
emphysema. It has also been shown in both men and dogs that, when
tracheal stenosis to expiration is greatly in excess of that to inspiration,
there is no emphysema. It has further been shown in dogs that, when
tracheal expiratory resistance is much greater than resistance to inspir-
ation, emphysema can be induced only with the addition of hyperpnea,
as follows of course when the animal is made to rebreathe into a bag
containing a high concentration of carbon dioxid.

Furthermore, the experiment in which the tracing of Figure 4
was made proved that, when a general bronchiolar hypertonus was
produced by the use of histamin, there was a great lowering of the
barometric pressure in the pleural cavity during the entire respiratory
cycle, but in spite of it there was a very small respiratory excursion
and the total volume of the lung was not increased.

Experiment 5. — An experiment was made for the purpose of tracing the
character and volume of the respiratory excursion and measuring the volume
of the animal under the effect of a large dose of histamin with an intact thorax.
The only operative procedure was the introduction of cannulae into the trachea
and the jugular vein, into the latter of which the histamin was injected. The
tracing (Fig. 5) should be read from right to left, and is simply the plethysmo-
graphic respiratory tracing of a 9 kilo dog. The first part, from 1 to 2, shows
the animal breathing atmospheric air, and then at 2, the animal was made to
rebreathe into a bag containing 80 per cent, oxygen and 20 per cent, carbon
dioxid. Under the influence of hyperpnea thus induced, with perfectly free
tracheal cannula, the total volume of the dog was increased about SO c.c. Then
during this period of hyperpnea 10 c.c. of 1 : 10,000 histamin was injected into
the jugular vein, and immediately, at 3, there followed a great modification in
the respiratory excursion. Before the effect of the histamin was apparent and
while the animal was rebrcathing a high concentration of carbon dioxid, the
volume of respirations varied between 500 and 264 c.c. Then before the histamin

162

ARC HI FES OF IXTERXAL MEDICI. \E

took effect, the duration of the respiratory cycle was 6/5 seconds. After the
histamin took effect, the volume of the respiration was 110 c.c, and the time
of a respiratory cycle was 13/5 seconds, the expiratory and inspiratory phases
Iiaving exactly the same duration. It is quite apparent from the prolongation of
both inspiration and expiration, with a great diminution in the volume of the
tidal air, that the animal was breathing against a great resistance in both
inspiration and expiration, and although in this experiment there is no tracing
of the intrapleural pressure, I believe we are justified in assuming from our
former experiments that the animal was employing a vigorous respiratory
effort. The tracing shows that the volume of the dog ivas not increased,
although the bronchiolar spasm zvas induced at a time when great hyperpnea
ivas demanded by rcbreathing 20 per cent, carbon dioxid; and just as we have
seen in our patients with severe bronchiolar spasm when active expiration zvas
demanded, inspiration and expiration are of equal length.

This experiment, in the light of all the preceding observations,
offers very conclusive evidence that a uniform hypertonus throughout
the bronchial tree will not prolong the expiratory more than the

Fig. 6. — Case 9. Bronchiolar spasm not preceded Iiy lironchitis.

inspiratory phase and will not produce emphysema. Furthermore, in
the experiment from which the tracing in Figure 3 was made, there
is very satisfactory evidence that active compression of the lung does
not create a vicious expiratory cycle by compressing the outlet of the
^ccspiratory units, as the Biermer theory demands.

The following case offers clinicnl confirmation of the foregoing
statements :

Cask 9.— The tracing shown in Figure 6 was made from a young, vigorous
man, who had an attack of bronchiolar spasm which was not preceded by
bronchitis, and so far as we are able to determine, the bronchiolar hypertonus
was a symptom of anaphylaxis from hay fever and not a liypertonus of the
bronchial musculature which could be traceable to any nervous excitation within
the bronchial tree or within the mediastinum. There was a considerable

HOOFER— EMPHYSEMA 163

amount of nasopharyngeal irritation. The results of the tracing, which was
made during his period of bronchiolar spasm and continued after epinephrin
was given hypoderniically until he had perfect respiratory relief, present very
good evidence that in clinical experience we may have bronchiolar hypertonus
which will greatly reduce the vital capacity of the lung and not increase its
residual air. In fact, we have in this patient with a uniform bronchiolar
hypertonus quite the same results so far as the vital capacity and lung volume
are concerned as has been shown in dogs when bronchiolar hypertonus was
induced by intravenous injections of histamin.

This patient went to bed with perfect respiratory comfort, and was awak-
ened in the middle of the night by a feeling of great respiratory discomfort.
Inasmuch as the patient was one of the interns on the medical service in
Lakeside Hospital, he was very conversant with the methods of studying
respiratory phenomena, and was able to give ideal cooperation. The tracing
should be read from right to left. The time marker is in fifths of a second.
The two lower lines were traced before the effect of adrenalin was apparent,
and the two upper lines were traced after the bronchiolar hypertonus had been
relieved by hypodermic injection of 1 : 1,000 epinephrin.

Before epinephrin was given, the vital capacity of his lung had been 2,699
c.c. seated, and 1,932 c.c. recumbent. After the effect of epinephrin was produced,
the vital capacity was 4.264 c.c. seated, and 3,893 c.c. recumbent. The respiratory
rate before epinephrin was 18 per minute, and afterward 14. Before epinephrin
there was a great abundance of fine and coarse sibilant rales over each portion
of both lungs, but after epinephrin all traces of adventitious sounds disappeared.
His sense of discomfort was purely on the inspiratory side.

Expiration w^as a passive procedure, involving no contraction of the abdom-
inal muscles. The patient said that on former occasions he had had attacks
much more severe and e.xhausting, and during them he had been compelled to
employ the abdominal muscles during expiration.

In the lower line at A, the duration of the respiratory cycle is 18/5 seconds.
The duration of inspiration is 6/5 seconds, and as will be noted in the tracing
of the inspiratory phase, the up stroke shows a drag toward the left after the
top of the excursion is approached. In other words, there is evidence of a
slowing of the volume flow toward the end of the inspiratory excursion. The
expiratory phase lasted 12/5 seconds, just double the duration of the inspiratory
phase, and the amount of air expelled during the 12/5 seconds was 374 c.c. At B.
after the bronchiolar hypertonus had subsided, we find a respiratory cycle which
lasted 24/5 seconds, the duration of inspiration being 11/5 seconds and the
expiratory phase 13 5 seconds, and the amount of air expelled was 2,210 c.c.
So, before epinephrin was given 324 c.c. were passively expelled in 12/5
seconds, and after epinephrin was given 2,210 c.c. were passively expelled in
13/5 seconds. In spite of the fact that the vital capacity during the attack of
bronchial spasm was lowered over 1,500 c.c, there was evidence of only a very
slight increase in the volume of the lung. Before epinephrin the lower border
of the lung was in the ninth interspace in the axillary line, and after epinephrin
it was in the eighth interspace; and during the attack, although his diaphragm
was coordinately activated with his intercostal muscles, all of both costal
margins moved in an outward direction during inspiration. So we are justified
in saying that the reduction in vital capacity was not traceable to an increase in
residual air in the lung, but was due purely to bronchial stenosis.

Dr. L., from whom the tracing was taken, has repeatedly tested
his vital capacity during the past year and has always found it to be
about 4,000 c.c. In fact, the character of Dr. L.'s tracing shows his
respiratory excursion during the attack to be quite like the respiratory
excursions found in dogs when an equal tracheal stenosis to inspiration

164 ARCHIVES OF IXTERXAL MEDICINE

and expiration is employed. After the attack was over, there was not
the slightest evidence of any kind for bronchitis or any modification
in the extensibility of the lung.

REGIONAL EXPIRATORY DYSPNEA

At a Stage of the investigation when the relation between bronchiolar
hypertonus and pulmonary emphysema seemed very elusive, a patient
came into Lakeside Hospital who presented a group of symptoms that
I had never before seen.

C.-^SE 10. — This man had an aneurysm of the aorta involving the latter
portion of the ascending and the first portion of the transverse arch. The first
part of his aorta showed no evidence of dilatation. All the physical signs and
the roentgenograms confirmed the diagnosis of an aneurysm which lay in front
and above the left bronchus and showed no evidences of contact with the right
bronchus. In all cases of obstruction of the bronchus from aneurysm which
had come under my observation, there was an equal obstruction to inspiration
and expiration, and there never had been in former cases any difference in
degree of obstruction when the patient was seated or recumbent. This man,
however, showed a very curious phenomenon. When he was in the upright
position, there was no demonstrable stenosis to the left bronchus, but when he
was on his back, there were decided evidences of stenosis to this bronchus, and
during expiration it was much greater than during inspiration. The man suf-
fered no pain and no great discomfort w'hen he was going about. His com-
plaint was that when he lay down he had an uncomfortable, "stuffy" feeling
over the left side of his chest.

Physical examination revealed very characteristic signs of aneurysm of the
arch of the aorta. When he was in the upright position, the respiratory
excursions of the two sides of the thorax were nearly equal, but when he was
lying down, the left side became distinctly larger than the right, and the
respiratory excursion of the entire left side was very nuich less than on the
right. Palpation revealed comparative immobilization of the ribs of the left side
during inspiration in this position, and over the left hypochondrium there was
an absence of the piston thrust from the left leaf of the diaphragm, whereas
the right side of the diaphragm had a perfectly normal and very active excur-
sion. The costal margin of the left side moved very slightly in an outward
direction during inspiration, and percussion revealed a decidedly lower position
of the left lung in the axillary line.

When the patient was upright, the respiratory sounds on the two sides of
the thorax were equally loud, but when he lay down, those on the left side were
only faintly audible. There was also much less tactile fremitus on the left side.

When he was seated, his vital capacity was 3,711 e.c. ; lying down, 2,545 c.c. —
a difference of 1,166 c.c, which is a far greater disparity than one will find
in any other condition, for few persons will on lying down show a lessening of
more than 250 c.c. in their vital capacity.

The pulmonary emphysema of the left side when the patient was recumbent
was due to the relation between the aneurysm and the left bronchus. In this
position the weight of the aneurysm compressed the left bronchus, but during
inspiration the lifting of the thoracic cage permitted a freer flow of air into
the right lung than that allowed out of the left lung during expiration. lie was
able to compensate for the impairment in ventilation of his left lung by a very
moderate increase in ventilation of the right, and therefore had no air hunger.
When the man was upright, there was during expiration no demonstrable sign
of compression of the left bronchus.

HOOVER— EMPHYSEMA 165

On comparing this case with Case 4, in which there was an expira-
tory stenosis of the trachea from an aneurysm of the aortic arch, it
will be apparent why this latter patient acquired an emphysema of his
left lung when he was lying down. One can readily see that the
essential difference between the two cases lies in the fact that the
patient last referred to had a very decided expiratory dyspnea involving
the left lung but none involving the right, whereas in the former patient,
although he suffered from severe expiratory dyspnea, both sides of
the lung were equally involved because the expiratory compression
was applied to the trachea. This suggested the idea that pulmonary
emphysema may be traceable to an excess of expiratory obstruction
over inspiratory obstruction ; and that clinically we find a marked
pulmonary emphysema attending the process because the expiratory
dyspnea is unequally distributed throughout the bronchial tree, whereas
under the influence of histamin our experiments failed to show any
emphysema simply because the hypertonus was equally distributed.

The patient with an expiratory dyspnea confined to the left
bronchus had a great increase in the residual air of his left lung,
which reduced the total capacity of his lungs by 1,166 c. c. This was
accomplished without air hunger or hyperpnea or an active expiration,
which is the very essential criticism above presented and must be met
by a satisfactory theory for emphysema caused by bronchiolar hyper-
tonus. If we conceive of these gross relations of expiratory dyspnea
between the two lungs being applied to the terminal bronchioles
throughout the bronchial tree, we then have a theory of emphysema
from clinical bronchiolar hypertonus which conforms to the facts of
clinical experience.

The physical signs produced by an aneurysm of the transverse
arch compressing the left bronchus were very readily reproduced
experimentally in the dog by procuring expiratory dyspnea confined
to one lung with no obstruction to either bronchus during inspiration.

Experiment 6.— A rubber cork was used, through the middle of which was
passed a metallic tube, with a diameter three-fourths that of the cork itself.
This cork was just large enough to fill a bronchus of the dog employed. Over
one end of the cork was spanned a small strip of rubber dam, and the
cork was so inserted into the right bronchus that this end lay m the distal
position During inspiration there was no obstruction to the entrance of air
into the right lung, but during expiration there was very pronounced obstruction.
Under a pressure equal to 80 mm. of water, the minute volume flow of air
through this valve measured UOO c.c, which would just about balance the
elastic retraction of the dog's lung. Figure 7 shows the animal s thorax when it
was comfortablv breathing atmospheric air. The disparity in the elevation of
the two leaves' of the diaphragm was quite like that in our patient with
aneuo-sm when he occupied the horizontal position. When the animal was
made to rebreathe a high concentration of carbon d.ox.d. this disparity was
very considerably increased.

166 ARCHirES OP I XTERXAL MLDICIXE

If we consider the factors attending pulmonary emphysema as a
result of bronchiolar hypertonus, we shall see that the following must
be considered: (1) There is always dyspnea, but it is not necessarily
so pronounced as to cause great discomfort. (2) The process is not
accompanied by hyperpnea. Even in the very severe cases, the minute

Fig. 7. — Experiinent (i
verse arch was iirodiic

ch ail aneurysm of the trans-

volume of air is not increased more than 2.^ jjcr cent. (3) An active
expiratory phase is not essential for the production of emphysema.

An unequally distributed expiratory dyspnea would .satisfy all these
criticisms, as shown in our ]iatient with exjiiratory dyspnea restricted

HOOVER— EMPHYSEMA 167

to the left side, and the dog with expiratory dyspnea restricted to the
right side. From our experimental and clinical observations, there-
fore, it seems reasonable to suppose that in the acute cases an unequal
distribution of the expiratory dyspnea is essential for the production
of emphysema. In the clironic atrophic cases also it is known to be
unequally distributed.

This study does not pretend to oflfer a satisfactory solution of the
problem of pulmonary emphysema. As in massive collapse of the
lung, which is also dependent upon some effect on lung motility, our
present knowledge of physiologj' is not sufficient for a satisfactory
explanation. This study does, however, reveal the inadequacy of the
commonly accepted theory of emphysema, and ju.stifies the suspicion
that an unequal distribution of expiratory stenosis throughout the
bronchial tree is the essential factor.

SUXIM.\RV

1. Though active expiratory compression of the lung is rarely
employed, the vigor with which the expiratory muscles can compress
the lungs is greater than that with which the inspiratory muscles
can distend them. Therefore, the air inspired within a given time can
be expired within the same time, provided the resistance in the trachea
or the branches of the bronchial tree is the same in inspiration as in
expiration.

2. When the tracheal or uniform bronchial resistance to expiration
exceeds that to inspiration, the residual air in the lung is increased
only when hyperpnea attains such a degree that the respiratory need
will not allow adequate time for the volume of the expired air to
equal that of the inspired air.

3. Compression of the lungs in expiration does not jiroduce a
vicious cycle of increasing resistance to expiration.

4. Neither hyperpnea nor an active expiration is essential for the
production of emphysema.

5. Prolongation of expiration in emphysema does not measure the
degree of expiratory resistance, but indicates the patient's respiratory
tolerance of prolongation of the expiratory phase. It is only in the
extremity of respiratory needs that active expiration is employed to
overcome expiratory resistance.

6. In bronchiolar spasm severe enough to demand an active expira-
tion, the inspiratory and expiratory phases have the same duration
and the volume flow within each phase is constant.

7. That an excess of expiratory over inspiratory resistance should
produce emphysema, the excess must be unequally distril)ute<l in the
bronchial tree.

A STUDY OF MICROLYMPHOIDOCYTIC LEUKEMIA

WITH THE REPORT OF A CASE *
SOLOMON FINEMAX. M.D., M.A.

MINNEAPOLIS
PART 1

In 1915, Citron ^ reported a case of leukemia which from the blood
picture and clinical findings, was diagnosed as "micromyeloblastic
leukemia." The postmortem histologic study carried out by Pappen-
heim and Citron showed that the bone marrow was entirely normal.
However, the cells of the follicles, as well as the cells of the interfol-
licular tissue of the lymph nodes and spleen, contained a slightly
enlarged, eccentrically placed, perfectly round nucleus which resembled
very closely the nucleus of a "myeloblast." The largest forms, however,
were not as large as those found in "myeloblastic leukemia."

While the blood showed a definite myelogenous picture, it was
evident that the "myeloblasts" and "micromyeloblasts" of the blood
were not coming to any extent from the bone marrow because the bone
marrow was normal (Pappenheim and Citron). Ou the other hand,
there was clear evidence that an actual proliferation in situ of the
follicular as well as of the interfollicular tissue was taking place, and,
furthermore, that there was an actual metaplasia of lymphocytic cells
into "micromyeloblastic" cells. Citron's conclusion, therefore, was
that the "micromyeloblastic" cells of the blood were being generated
in the follicles and interfollicular tissue, of the lymph nodes and spleen
and that these "micromyeloblastic" cells were passing from the follicles
into the blood stream. Citron regarded his case as being of utmost
significance in that it was the first case on record which would seem
to prove that the dualistic doctrine of the origin of white blood cells,
namely, the complete independence of the origin of myeloid and
lymphoid blood cells, did not always hold true.

Citron's case was one which anatomists and hematologists had
long been seeking. It offered partial evidence in contradition to the
statements made by Naegeli. Schridde, Meyer, Hyneke and Zieglcr = and
other dualists that "myeloid tissue" has never been observed proliferat-

*A thesis submitted tn the facuUy of the Graduate School of the Univer-
sity of Minnesota in partial fulfihncnt of the requirements for the degree of
Master of Arts.

1. Citron, J.: Ueber zwei licnuTkenswcrte Fiille von (akuter) Leukiimie,
Folia haematol. 20:1, 1915.

2. Naegeli. Schridde, Meyer and Hyneke, and Ziegler : Quoted by Ehrlich, P.,
and Lazarus, A.: (Rewritten by A. Lazarus and O. Naegeli, and translated by
H. Armit). Anemia, New York, Rebman Co., 1910, p. 148.

FIXEMAX—LVMPHOIDOCrTIC LEUKEMIA 169

ing in the germinal centers of spleen or lymph node follicles ; that in
all cases showing "myeloid infiltration" of the spleen or lymph nodes
the follicles are passive or atrophied.

This case shows not only a proliferation of the lymph node follicles
but what is of utmost significance, a proliferation of "micromyeloblasts"
and "myeloblasts" in the germinal centers of these proliferating
follicles. Although a necropsy was not permitted, we were fortunate
in obtaining a lymph node during the patient's life. This lymph node
was fixed in Helly's fluid immediately after excision, so that the
material studied was obtained under ideal conditions.

In brief, we had a case of leukemia in which the blood contained
enormous numbers of cells, which ordinarily would be called "the rare
micromyeloblasts." The blood picture would lead one to make a
diagnosis of myelogenous leukemia. Clinically, however, we had
evidence pointing to lymphopoietic activity, namely, a mediastinal
tumor, as revealed by percussion and roentgenograms, a very marked
enlargement of lymph nodes over the entire body, and a markedly
enlarged spleen.

That the lymphopoietic organs were very active was proven by the
sections of lymph node, which showed that the blood cells referred
to as "micromyeloblasts" were in reality proliferating right in the
lymph follicles and germ centers of the node, and could be traced
entering the circulation from the lymph node.

I believe, therefore, that our case offers strong evidence that the
unitarian theory of the origin of white blood cells, which will be
discussed later, is the correct one.

REPORT OF CASE

//u/orj.— Service of Dr. E, T. F. Richards. Patient, a girl, aged 17, single,
American of Swedish descent, came to University Hospital Jan. 20, 1919.

Present Complaint. — Weakness and rapidly enlarging masses in the neck
and axillae.

Family History. — Negative.

Social and Occupational History. — Negative.

Past History.—She has been well up to four weeks ago (December, 1918).
She had measles at the age of 6; no sequels; scarlet fever at 15; no sequels;
tonsillitis, acute, at 16; lasted four days; could not swallow without pain. No
enlarged glands at that time. Felt fine after attack subsided.

Head: £yc.f.— She has had "eyestrain" for the past four months (October,
1918, to January, 1919) and wore glasses, with relief; vision is good. Ears.—
Negative. A'o.fi'.— Occasional "cold"; otherwise negative. Mouth and Throat.
— Teeth filled during summer of 1918.

Cardiorespiratory : Negative.

Gastro-Intcstinal : Negative.

Genito-Urinary : Negative except occasional nocturia.

Catamenia: Began at 15; negative.

Venereal : Negative on direct and indirect questioning.

Neuromuscular : Negative.

Skin : Negative.

170 ARCHU'ES OF IXTERXAL MEDICINE

Habits: Good. Weight: Best in summer of 1918, 135 pounds; now appar-
ently 120 pounds..

Present Illness.—'Patknt was well until November, 1918; felt "fine and
had red cheeks." Beginning in November, 1918, and especially toward the
latter part of the month, she began to complain of being tired, especially so
after coming home in the evening. Neither the patient nor her mother noticed
any paleness at that time nor enlarged masses. Mother did notice, however,
that the patient became very fretful and irritable.

Dec. 23, 1918, the patient came home from work and complained of feeling
very tired and of having a severe headache and pain in her knees. Mother
thought she had "influenza." Patient did not go back to work on account of
weakness. Also at that time she noticed a beginning pallor of the face. The
pain in the knees subsided after a few days, but the weakness and pallor
became progressive. Patient did not go to bed, however, for two weeks. At
the end of that time, in the early part of January, she noticed "lumps" in
her neck, especially behind and below the ears. Her hearing became impaired,
and once, when she blew her nose, she noticed blood on her handkerchief.
A physician was called and diagnosed the case as "mumps."

Two days after the appearance of the masses behind and below the ears,
the patient noticed a "lump" under the right jaw. This lump enlarged at first
and then decreased in size. Three weeks after onset, about January 14, patient
noticed "kimps" in both axillae, but none elsewhere. From that time on the
swellings gradually enlarged and the mother began to notice a definite pro-
gressive paleness.

Patient denied having fever, hemorrhages and sweats. She "caught cold"
about four days before entering the hospital and developed a nonproductive
cough. On entrance she had no headache; she could read; nasal breathing
was free. She had no pains; her appetite was good; bowels were regular,
and she slept well.

Physical Examination.— Temperature, 100.6 F. ; pulse, 146. Patient in bed;
has dry nonproductive cough; voice hoarse; looks very anemic; development
and nutrition, good; no edema, cyanosis or jaundice. Neck and sides of face
appear swollen with irregular shaped subcutaneous masses.

Head: Sinuses and Mastoids.— Tenderness over frontal sinuses and mastoids
and over left antrum. Eyes. — Eyegrounds show extensive fresh hemorrhages
in both retinae and disks; disk margins indistinct but not choked. Sclerae
have a shghtly yellowish tinge. Conjunctivae are very pale. Visual fields
appear to be normal. £(jr.s.— Moderate deafness; with watch, right ear, 3V2
inches; left ear, one-half inch. Bone conduction greater than air conduction
R. and L. Nose. — Negative. Mouth. — Lips show extreme anemia and capil-
lary pulse is detectable. Gums are also extremely anemic; gingivitis is present.
The gums are tender and bleed on pressure. Teeth are verj- crowded, espe-
cially in the upper jaw where one tooth protrudes in abnormal position. Few
teeth are carious. Mucosa of mouth is very pale. The palate is very high and
arched. The tonsils are very large, glistening, pearly in appearance, .so large
that they almost occlude the entire pharynx. The tongue is coated.

Glands: Chains and matted masses of glands are felt in these regions:
anterior and posterior auricular; submaxillary; anterior and posterior cervical,
left and right subclavicular; over both apices anteriorly and posteriorly: over
the trapezii; in the axillae and in the inguinal regions. These are all bilateral.
They vary in size from that of a small pea to that of a hen's egg. Some are
round; some are oval and many are matted. They are hard and not tender
to pressure; are freely movable and not attached to the skin. The size of
the face is shown in Figure 1. A comparison is made of the size of the glands
from day to day in Table 1 and Figure 2.

Chest and Lungs : Anteriorly.— lr\spect\on. negative, except that chest comes
down rather slowly with expiration. Palpation, negative. Percussion, abnor-
mally wide supracordial dulness (Fig. 3). Auscultation, generalized "cog-wheel"
breathing; exaggerated, tubular over area of supracordial dulness. Vocal

FIXEil.AX-I.yMPHOIDOCyTIC LEUKEMIA

171

fremitus and whispered voice increased over same area. Posteriorly. — Inspec-
tion, small masses over both trapezii. Palpation, negative. Percussion, dimin-
ished resonance for 5 cm. to right and left of vertebrae on levels I to VI D.
Auscultation, cog-wheel breathing, generalized. Tubular breathing from I to
VI D. for 5 cm. to right and left. Vocal fremitus and whispered voice increased
over spine from I to VI D. for distance of 5 cm. to the right.

Breasts : . Negative.

Spine : Increased whispered voice and duhiess to sixth dorsal vertebra.
Costovertebral angles, negative.

Heart : Inspection, negative : palpation, negative ; percussion, cardiohepatic
angle obtuse; absolute dulness increased (Fig. 3).

Fig. 1.— Photograph of patient Fel). 14. 1919. Note large preauricular nodes.
See Figure 2.

Me.vsure.mk.nts

Intercostal Space Right, Left,

Cm. Cm.

2 3 4

3 -' 6

4 i 7

5 4 8

Auscultation Sounds.—SiCowA pulmonic i.; accntuatcd : first sound at the
ape.x is accentuated. »A/«rwi«r.s.— Systolic murmur, short, transmitted but slightly
into a.xilla, heard at apex, tricuspid and aortic areas and clearest immediately
to the left of the sternum. Blood vessels, negative. Pulse, rapid; sharp rise
and fall. Blood pressure: systolic, 130; diastolic. 40.

Abdomen: Inspection, negative; palpation, spleen palpable 7 cm. below
costal margin in midclavicular line. No notch felt; freely movable with res-
piration; no tenderness; liver not enlarged. Percussion, splenic duhiess same
as on palpation. Liver dulness 12 cm. in midclavicular line.

Extremities: Upper, negative, except extreme anemia of nails. Lower, old
traumatic scar on left thigh, otherwise negative.

172 ARCHIVES OF JXTERXAL MEDICINE

Sensation: Normal. Reflexes: Negative, except that right knee jerk was
very sluggish and the left was not obtainable. No cloni ; Babinski, Gordon
or bppenheim signs. Vibration Sense: Present over all bony prominences.
Rectum, negative.

Laboratory Data on Entrance. — Urine: trace albumin; sp. gr., 1.010; acid;
few leukocytes and few hyalin casts ; Bence-Jones body negative. Guaiac test
repeatedly negative. Sputum : negative. Stool : negative repeatedly for blood
with guaiac test.

Blood: hemoglobin, 29 per cent. (Dare); red blood cells, 1,900,000; white
blood cells, 99,000. Differential count (Table 4 and Fig. 4).

Phenolsulphonephthalein excretion, 66 per cent, in two hours. Blood Was-
sermann negative.

Blood Chemistry: Sugar, 0.105 per cent.; creatinin, 1.40 mg. ; urea nitrogen,
10.50 mg.

Blood culture negative.

Mosenthal test negative. For metabolic studies with excreta in urine and
feces see Tables 2 and 3. Electrocardiograph tracing : negative.

PROTOCOL

All white and red cell counts and hemoglobin determinations were
done by Dr. Swan Erickson and- myself. We used the same set of
pipets throughout, and on numerous occasions checked 'each other
and followed as closely as possible the same technic throughout the
patient's stay in the hospital. In referring to spleen and liver measure-
ments, we mean measurement in the midclavicular line below the costal
margin.

Jan. 21, 1919: Hemoglobin, 29 per cent. (Dare). Red blood cells, 1,900,000;
white blood cells, 99,000.

January 22: White blood cells before roentgen-ray treatment, 44,000.
Tioentgen-ray treatment over glands of chest and neck.

January 23 : Glands diminished to from one-fourth to one-sixth of their
previous size. Mediastinal width 8 cm. Tonsils only half as large. White blood
cells, 9,600.

January 24: White blood cells, 4,800.

January 25: Transfusion of 200 c.c. blood; no reaction; clinical improve-
ment marked.

January 27: Glands, smaller than on January 23. Tonsils only about one-
sixth of former size. Spleen, 3 cm. Mediastinal dulness diminished.

January 30: Right and left posterior auricular glands enlarging; hearing
diminishing. White blood cells, 33,000 at 11 a. m.; 50,000 at 6 p. m.

Tanuarv 31 : Glands enlarging again ; spleen 7 cm. White blood cells,
50,000.

February 1: White blood cells, 8000; spleen, 4 cm. ^Mediastinal dulness,
5.5 cm. Transfusion of 300 c.c. blood.

February 3: White blood cells, 9,000. »

February 4 : Distinct swellings in front of right and left ears. Tonsils
enlarging; some of other glands enlarging. White blood cells, 26,800 at 11
a. m.; 49,000 at 6 p. m.

February 5: White blood cells, 70.300. Difficulty in breathing present;
restless: hearing poorer. Tonsils markedly enlarged; lacked 1 cm. of meet-
ing in midline. Spleen 7 cm. All glands definitely enlarged.

February 6: White blood cells, 90,000; no reticulated cells; platelets. 90,000
per cubic millimeter.

February 7: White blood cells, 72,000; axillary lymph gland excised.

FINEMAX-LYMPHOIDOCYTIC LEUKEMIA

173

February 8 : White blood cells, 42,000 at 10 : 30 a. m. ; 68,000 at 7 p. m.
Patient's general condition worse in evening; dyspnea; nasal breathing impos-
sible. Tonsils, 0.5 cm. apart. Facial and axillary glands enlarged. Spleen,
7.5 cm. Supracordial dulness, 8.5 cm. in second intercostal space.

February 10: White blood cells, 60.000; patient's general condition the same.

February 11: White blood cells, 89,000.

February 12 and 13 : Practically no change in patient's general condition.
White blood cells, 108,000.

February 14: White blood cells. 108,000 at 9:30 a. m. ; 68,000 at 2 p. m.
The third roentgen-ray treatment was given at 2 : 30 p. m. over the neck, tibia
and femurs. Patient dyspneic ; dilated veins over temples. Mass of glands on
right side of face, 8 cm. in diameter; left, 7 cm. These masses are palpable
from back of the ears to the angle of the eye. They are very prominent. The
tonsils and uvula practically obstruct nasopharynx. Nasal breathing impos-
.sible. Spleen, 9.S cm.

Fig. 2. — Shows variations in size of preauricular lymph nodes and spleen
in relation to roentgen-ray therapy. Scale in centimeters shown on left.
Splenic tumor measured below costal margin in midclavicular line.

February 15: Masses in front of both ears are barely visible. Patient
breathing freely through nose. White blood cells, 29,300 at 10:25 a. m.
Clotting and bleeding time, 5 minutes each.

1:30 p. m. : transfusion of 400 c.c. Felt stronger immediately after trans-
fusion. Facial glands still smaller; not visible, just barely palpable as flat
masses about 2 cm. in diameter. Tonsils about 1.5 cm. apart. Spleen, 7.5 cm.

February 17: White blood cells, 10,000. Facial mass on right side not
palpable at all ; on the left just barely palpable. Tonsils about 2.5 cm. apart.
Patient "feels fine." Hemoglobin and red blood cells approximately the same
as on entrance. Onset of epistaxis and appearance of numerous petechiae on
both legs, from 2 to 3 mm. in diameter.

174 ARCHIVES OF IXTERXAL MEDICINE

February 18: White blood cells, 20,000.

February 19; Condition worse. Hemoglobin, 26 per cent. (Fleischl) ; red
blood cells, 1,860,000; white blood cells, 140.000 at 3 p. m.: 208,000 at mid-
night. Headache all day; constant dull abdominal pains; slow epistaxis all
day. Complained of dimmed vision. Numerous fresh petechial hemorrhages
on legs with several large bluish areas from 2 to 3 cm. in diameter. Edema
of feet. Pressure over sternum, skull, humeri, ulnae, radii, femurs and tibiae
elicited exquisite pain. For the first time numerous mitotic figures were
observed in the blood in wet and dry preparations.

February 20: Condition worse. Hemoglobin, 21 per cent. (Dare); red
blood cells, 1,344.000; white blood cells. 242,000. Glands seemingly not enlarged.
Transfusion of 200 c.c. Pulse weak and irregular, thready; rate. 160. Fifteen
minims benzol by mouth. Blood chemistry: sugar, 0.117 per cent.; creatinin,
0.75 mg. ; urea nitrogen, 9.188 mg.

February 21 : Eight minims benzol by mouth in the morning. White blood
cells at 9:30 a. m., 62,000; at 1 p. m., 44,000. Blood culture negative. Blood
of patient injected into rabbit's ear vein. Nitrogen intake, 9.6 gm. ; output,
16.8 gm. in urine. Blood chemistry: blood sugar, 0.099 per cent.; creatinin,
0.60 mg. ; urea nitrogen, 9.96 mg.

February 22: Spinal puncture, 20 c.c. clear fluid under pressure; Nonne-
negative; cell count, 3 per cubic millimeter. Colloidal gold test negative;
Wassermann +. General condition good. White blood cells, 6,800. Lymph
nodes in general smaller. Bone tenderness diminished. Several epistaxes
during day.

February 23 : Condition further improved ; patient smiling, cheerful, laughed,
insisted on being allowed to sit up in chair. Vision definitely impaired; could
barely see large newspaper headlines, and small type not at all. Glands and
spleen somewhat smaller.

February 24: Condition worse; headache; anxious expression; lips seemed
paler ; spleen and some of glands enlarged somewhat. White blood cells,
10,000.

February 25 : Spleen still further enlarged ; glands about the same. White
blood cells, 44,000; platelets, 92,000.

February 26: White blood cells, 82,000; 23 minims benzol by mouth.

February 27: White blood cells, 76,000; 23 minims benzol by mouth.

Februarj- 28: White blood cells, 34,200; hemoglobin, 13 per cent. (Dare) ; red
blood cells, 900,000; felt fine; asked for second helpings of her meals. Benzot
discontinued.

March 1 : White blood cells, 22,000.

March 2: W'hite blood cells, 61,000. Condition worse; epistaxis in morn-
ing; glands and spleen about the same.

March 3: White blood cells, 105,000 at 1 p. m. Condition worse; severe
headache. White blood cells, 176,000 at 7 p. m. Profuse epistaxis in the
morning. Lymph glands about the same. Spleen 10 cm. and reached to umbili-
cus; not tender and no rub felt. Liver, 7 cm.; pulse, from 140 to 170; gallop
rhythm. Mitotic figures observed in both wet and dry preparations of blood.
Transfusion of 160 c.c. followed by hypodermic of morphin sulphate, Vo grain,
and atropin sulphate, Viso grain.

March 4: White blood cells, 95,000 at 2 p. m. ; 106,000 at 7 p. m.

March 5 : White blood cells, 75,200 at 10 : 30 a. m. Transfusion at 1 : 45
p. m.. 150 c.c.

March 6: White blood cells, 26,300 at 1:30 p. m. I^oentgen-ray treatment
to spleen at 3 : 30 p. m.

March 7: White blood cells, 4,000 at 9:30 a. m. Transfusion at 10:25
a. m., 200 c.c.

March 8: White blood cells, 2.300. Spleen by noon just barely palpable.
General condition worse. Sight poor; eyegrounds show numerous fresh
hemorrhages.

TABLE 1.— CoMPARATRE Data from Day to Day Regarding White Cell
Count, Changes in White Count Observable on the Same Day;
Roentgen-Ray Treatment; Number of Stem Cells Per C.Mm.; Num-
ber OF Mitotic Figures Per C.Mm., Transfusions, Fluctuations in Size
OF Splenic Tumor, Tonsils, Facial Glands and General Condition

White

Eoent^

1

Gen-

Stem Ml

'<^l^

Later on
in Day

Ray
Treat-
ment

Cells I

Ig-

..ion.

Cm.

^■S.'

Glands,
Cm.

(indl-
tlon

1/21

1/22

99,000
40,000

Chi't-

1,386

...

4.6

0.5

1.5

Fair

1/23
1/24

9.600
4,800

27i

7

i"

1.5

Good

1/25
1/26

5,000

200

...

Good

i;i

5,400
6,900

S

s

1

Good

1/2S

13,800

33,000

50,000

900

1 5

Good

51,000

Spleen

6

Fair

540

300

S

2

i.75

Good

V3

12,300

2/4

26,300

49,700

«

I

3

Good

4.6

0.76

3.5

Bad

15,894

72,000

h

68;6o6

2/9

65,000

7.6

0.6

6.5

Bad

2/11

89,000

2/12

2M3

108,000

2/14

108.000

64,000

^^^

31.605

9.6

0

8

Bad

2/1.5

29,300

400

8

1.6

2.5

Good

2/16

10,300

618

20,000

2/19

140.000

2oe,o66

7

2

Bad

2/21

62.000

44,000

8.060

7

1.25

Fair

6.5

2.5

1

Fair

8,500

0.75

Good

10,400

7

2

1.5

Fair

44,600

1.5

Fair

82,000

45,692

8

1.5

Fair

2/27

76,400

8

Fine

2/28

34,200

6,498

1.5

Fine

22,000

1.5

Fine

3/ 2

61,000

8

1.5

Fair

3/3

105,000

i76,oo6

50.336

160 _

10

1.5

Bad

3/ 4

95,000

1.5

Fair

3/ 5

75.000

150

.

...

1.5

Fine

Spleen

10

1.5

Fine

3/7

4,000

200

4

1.6

Fair

3/ 8

2,30OS

153

2.6

1.5

3/ »

3,400

Fair

3/10

5,600

150

8.6

Fair

3/11

5,000

4.6

1

1.5

Fair

3/12

4.5

1.5

Fine

3/13

8,100

6

1.5

Fine

3/14

Fine

3/15

7,900

6

1.5

Fine

3/16

9,600

6

1.5

Fine

3/17

6

...

1.5

Fine

3/18

6,000

6

1.5

Fine

3/19

6.900

200

6

1.5

Fine

3/20

5,900

...

Fine

3/21

lOO

Fine

3/22

12,700

Fine

3/23

7

2

Fine

3/24

17,300

Fine

3/25

25,500

Fine

3/26

34.000

10.6

1.75

Fine

80,000

106

1.75

Fair

3/28

115,000

105

1.75

Bad

178,666

32,605

300

13

2.5

1.75

Bad

3/30

260,000

295,000

13

1.76

Fair

334,000

16.5

1.75

4/ 1

480,000

Spleen

300

16.6

1.75

Bad

75.800

IS.5

2

1.76

Fair

4/ 3

485.000

1.75

Good

578,000

Spleen

l^l'

4/5

545,000

4/ 7

646,000

Spleen

2,684

4/8

...

....

' Splenic tumor palpable below the costal margin
t Approximate distance between the tonsils.
: Lowest white count recorded.

the midclavicular line.

176 ARCHIVES OF INTERNAL MEDICINE

March 9 to 21 : The next thirteen days the "white cell count and the
patient's general condition simulated a lull before an impending storm. The
white count varied from 5,000 to 9,000. The patient felt fairly well and even
insisted on getting out of bed. We gave her her eighth and ninth transfusion
of 350 c.c. of blood in all. Epistaxis occurred frequently. Blood culture was
again negative. The glands remained stationary. The tonsils enlarged some-
what and the spleen measured 7.5 cm. March 13. From then on to March 21
the spleen remained stationary.

March 11: White blood cells, 5,600. Basal metabolism, +20 per cent.

March 14: White blood cells, 7,000. Basal metabolism, +7 per cent.

March 22 : White blood cells, 12,000. General condition same.

March 23 : White blood cells, 17,000. Complained of poor vision ; facial
glands enlarged slightly. Spleen 8.5 cm. Basal metabolism, +29 per cent.
Patient restless.

March 24: White blood cells, 17,300; felt fine; sat up in chair. Spina! fluid:
normal pressure; Nonne negative; cell count, 1 per cubic millimeter; colloidal
gold test negative ; Wassermann negative.

March 25 : White blood count, 25,000. General condition good.

March 26: White blood cells, 34,000. General condition good; sat up in
chair; facial glands little more enlarged. Mediastinum 7.5 cm. on percussion.
Spleen, 10.5 cm. on palpation.

March 27: White blood cells, 80,000. Condition worse; felt miserable;
sat up very little; complained of dull abdominal pain; pulse very rapid and
heart had gallop rhythm.

March 28: White blood cells, 115,000. In bed all day; anxious expression;
ankles and feet edematous; suppuration set in under nail of left big toe; hear-
ing diminished ; missed third menstrual period.

March 29 : White blood cells, 247,000 at 10 a. m. At 10 : 38 a. m. transfusion
of 300 c.c. White blood cells, 185,000 at 1 : 45 p. m. Spleen, 15 cm. and beyond
umbilicoxyphoid line. White blood cells, 178,000 at 4:10 p. m. Liver, 6 cm.
Abdominal pains severe. Ankles and feet more edematous.

March 30: White blood cells, 260,000 at 11 a. m.; 295,000 at 5 p. m. Spleen
larger (Fig. 5).

March 31 : White blood cells, 334,000. Severe abdominal pain — upper half.
Pressure over spleen and sternum gave exquisite pain. Spleen still larger.
Fluid in flanks. Edema of lower extremities increased. Dulncss at base of
left lung posteriorly, probably due to enlarged spleen. Pneumonia not
demonstrable.

April 1: White blood cells, 480,000 at 10 a. m. Tonsils enlarged, 2 cm.
apart ; nasal breathing free. Transfusion of 300 c.c. Roentgen-ray treatment
of spleen, three minutes only on account of poor condition of patient. Edema
of right hand, wrist and sacrum. Patient very stuporous and complained of
severe pain over the spleen.

April 2: White blood cells, 500,000. Condition slightly improved.

April 3: White blood cells, 485,000. Condition still better; talkative and
bright.

April 4: White blood cells, 578,000 at 2:30 p. m. Roentgen-ray treatment
at 1:30 p. m. over spleen. Liver, 10 cm.; tender. Basal metabolism, +29
per cent.

April 5: White blood cells, 545,000 at 4 p. ni. Transfusion of 200 c.c. at
11 a. m. Patient felt better; sat up one hour in bed.

April 6: White blood cells, 541.000. Condition worse.

April 7: White blood cells, 646,000 at 2 p. m. Roentgen-ray treatment at
1 : 30 p. m., 9 minutes to spleen. Frequent severe epistaxis during day, wilh
several cmescs of clotted blood. At night breathing became stertorous and at
midnight the patient could be awakened only with great difliculty. When
awake, she was rational.

April 8: At 12:15 a. m. patient cried out several times, with inspiratory
gasps. Pulse at that time was very rapid but of good quality. Death occurred
a few seconds later of respiratory failure.

FLXEMAX—LYMPHOIDOCYTIC LEUKEMIA 177

DISCUSSION

1. Hemoglobin and Red Blood Cells. — The hemoglobin and red blood
cell count remained low throughout and both progressively diminished,
in spite of twelve transfusions, a total of 2,760 c. c of blood.

2. IVhite Cell Count. — The white cell count showed some extraor-
dinarily sudden, unaccountable fluctuations. The rise or fall of the
white cell count during twenty-four hour periods would sometimes
be so great that we were obliged to make three or four counts in
twenty- four hours to check our findings.

ON FKlwtion

Fig. 3. — Showing percussable medi
n. 21. 1919.

entrance to hospital

On entrance to the hospital the patient's white cell count was 99,000.
In twenty-four hours, before any treatment was instituted, the count
fell to 44,000, followed by a further fall after roentgen-ray exposure
of the lymph nodes of the neck and mediastinum.

From then on the white cell count kept oscillating between counts
as low as 2,300 and as high as 646,000 on the day of death. By
making blood counts several hours apart on the same day we were
able to demonstrate the following:

January 30: rise of 17.000 white cells in 7 hours.

February 8: rise of 28,000 white cells in 8^/2 hours.

February 14: fall of 44,000 white cells in 41/2 hours.

February 19: rise of 68,000 white cells in 9 hours.

178 ARCHIVES OF IXTERXAL MEDJCIXE

March 3 : rise of 71,000 white cells in 6 hours.

March 29: fall of 62,000 white cells in 3 hours.

The rapidity with which a rise or fall of the white cell count would
occur in this case was astounding. Thus, we see a rise from 10,500
cells February 17, to 242,000 cells February 20, and just as rapid a
fall to 6,800 cells February 22. We see another sudden rise from
22,000 cells March 1, to 176,000 cells March 3, and just as sudden a
fall to 2,300 cells March 8. Up to March 8, some of the white cell
count fluctuations seemed to occur entirely spontaneously. At times,
however, it seemed to us as if the transfusions and roentgen-ray
exposures had an immediate effect in causing a drop in the white count.
Figure 6, showing the daily blood counts and therapy, would seem to
indicate that such was the case.

Alarch 22, the white cell count began to rise with lightning-like
rapidity and rose from 6,000 to 646,000 cells in the next seventeen
days. From the beginning of this last rise in the white cell count,
transfusions, and roentgen-ray exposures had practically no effect on
the count, and the patient died with the highest count demonstrable
during her stay at the hospital.

On looking over the blood chart (Fig. 6) one is struck by the
seemingly periodic exacerbation of the white cell count rise, each
period lasting from five to six days. It is of interest to note here that
with each rise in the white cell count, the patient's general condition
became definitely worse. As a matter of fact, we could usually note
a beginning rise in the white cell count by the change in the patient's
general condition. During the periods of low count she would be happy
and feel so well that she would insist on being permitted to sit up in
a chair. As soon as the white cell count would begin to rise, the
patient would stay in bed and complain usually of headache and
abdominal distress, and she would have an anxious expression.

Krjukow ^ describes a case of "microlymphoidocytic" leukemia in
which there were sudden fluctuations in the white cell count, associated
with rapid fluctuations in the size of the spleen.

3. Mitotic Figures. — A very interesting finding, with counts over
80,000, was the presence of numerous mitotic figures in the blood.
These cells, in all stages of mitosis, were easily demonstrable in both
the 1 per cent, acetic acid solution, and in the dry stained blood prep-
arations. The higher the count the greater would be the number of
mitotic figures. At one time we demonstrated as many as -fifteen
cells in mitosis in a single field under the low power lens of the
microscope and a No. 10 eyepiece.

3. Krjukow, A.: Ueber cincn Fall von akiitcr Microlymphoidozyten Lcii-
kamie. Folia Haematol. 15:328, 1915.

FlXEMAX—WMPHOIDOCrTlC LEUKEMIA _ 179

Gordon Ward * describes a "peculiar case of acute leukemia,"
which from his description might very well have been a case of
microlymphoidocytic leukemia, in which he observed as many as
thirty-seven mitotic figures per cubic millimeter of blood. Krjukow
noted in his case numerous mitotic figures in the blood during the
periods of splenic enlargement.

In an attempt to determine whether these mitotic cells, which in
all probability were being forced out from the rapidly proliferating
lymphatic tissue into the blood stream, would continue the process to
completion in vitro, Dr. Swan E. Erickson and I made a study of
these cells in 1 per cent, acetic acid solution, physiologic solution of
sodium chlorid, I14 per cent, sodium citrate in physiologic sodium
chlorid solution, II/2 per cent, sodium citrate in water, stock blood serum

"-

1

!

\

1 -

—1ST

-^i

/

' ^

1

1 1

1

^^.

,

'A

.^

1

\A

/ \ i

A,

A

\

l/N

A

. i^/^>

1 ,'-

\^

/

f^

v,^^yc<^i^

UJ/\

^^

,x^

^h

^

J

.^

-^

k 0

Fig. 4. — Shows relation between the total daily white cell count, total daily
microlymphoidocytes, and daily percentage of inicrolymphoidocytes. Only data
obtained during low and high count days are here recorded.

of Groups 2 and 3, and also in patient's own serum, which was of Group
4. We used a warm stage and kept the cells under observation in the
above mentioned solutions at body temperature for as long as thirty-six
hours. We did not observe a single instance of continuation of the
process of mitosis in vitro.

4. Differential Counts. — The differential counts gave interesting
findings (Table 4). The percentage of the various cells was calculated
on a basis of from 300 to 700 cells. The percentage of polymorphonu-
clear neutrophils varied from as low as 0.4 per cent, to 33.66zp per cent.
The total number of polymorphonuclear neutrophils was not directly
proportional to the relative percentage. For example, with the lowest

4. Ward, G. : .\ Pecuhar Case of .^cute Leukemi;
1919; abstr. Folia hacmatol. 20:158 (Nov.) 1920.

J. 48:93 (.Aug.;

180 ARCHIVES OF IXTERNAL MEDICINE

count of 2.300, the percentage of polymorphonuclear neutrophils was
19.33'+ per cent., and the total number of polymorphonuclear neu-
trophils per cu. mm. was 445. With the highest count of 646,000 the
percentage of polymorphonuclear neutrophils was only 1.4 per cent.
with an absolute number of 9,044 polymorphonuclear neutrophils per
cu. mm. The highest percentage of polymorphonuclear neutrophils,
namely, 33.66+ per cent., was present with a total of 6,000 white cells
per cu. mm. The number of polymorphonuclear neutrophils per cu.
mm. in the normal blood is about 6,000. In this case even with counts as
high as 242,000, the total number of polymorphonuclear neutrophils
per cu. mm. was only 3,388. Only shortly before death did the absolute
number of polymorphonuclear neutrophils go up above the normal,
as follows:

March 29 : Total white cells, 247,000 ; polymorphonuclears, 14,820.
April 2: Total white cells, 500,000; polymorphonuclears, 9,150. April 7:
Total white cells, 646,000; polymorphonuclears, 9,044.

5. Nucleated Red Blood Cells and Myelocytes. — Nucleated red
blood cells and myelocytes were present in small numbers. These
probably were an irritation phenomenon due to the extreme anemia,
which at its lowest point gave a hemoglobin percentage of 13 on the
Von Fleischl-Miescher instrument, and a red count of only 900,000.

6. The "Micromyelohlast." — The most interesting cell was the
so-called "micromyeloblast" of Naegeli and Schridde, or the "micro-
lymphoidocyte" or "stem cell" of Pappenheim. The total number and
relative per cent, of this cell was practically directly proportional to
the total count (Fig. 4). W'ith the rise in total white cell count our
patient was always clinically worse and the disease could be said to
have assumed a more severe aspect. Coincidentally with the increase in
the severity of the disease, the "micromyeloblast" would increase in
number and percentage, a finding which is similar to the findings of
Panton and Tidy ^ in a series of three cases.

With a white cell count of 2,300, the total "micromyeloblast" count
was 153. ^^'ith a white cell count of 500,000, the "micromyeloblast"
count rose to 75.800. On the day before death, the total white cell count
was 646,000, but the "micromyeloblast" count was only 2.548. probably
an exhaustion phenomenon.

7. Therapy.- — Our therapy consisted of a high carbohydrate diet,
roentgen-ray exposure and transfusions. Benzol was tried on two
occasions, but in amounts so small and for so short a period of tirne that
its effect can safely be discounted. We have already mentioned the
peculiar sudden fluctuations in the white cell count, associated with an

S. Panton. P. H.. and Tidv. H. L. : Some Atypical Cases of Leukemia. Foli
haematol. 17:398, 191.3.

FIXEMAX—LYMPHOIDOCYTIC LEUKEMIA

181

improvement in the patient's general condition during the periods of low
count.

Haughwout and Azuzano " report improvement in forty-eight
hours following the administration of benzyl benzoate. They suggest,
however, that this may be a normal fluctuation of the disease.

The roentgen-ray treatment varied a great deal in its eiTect on the
enlarged spleen and lymph nodes, and on the white cell count. Seven
treatments were given by Dr. Frank S. Bissel. The white cell count
diminished after each treatment, with the exception of the treatments
given a few days before the patient's death, at which time the count
rose from 6,000 to 646,000. It must not be forgotten, however, that

Fig. 5. — Composite tracing of spleen and liver from March 9 to April 6>
Note gradual enlargement of both. Compare with Figure 20 showing gradual
diminution of splenic dulness.

on six occasions there occurred diminutions in the count, varying from
44,000 to 236,000 cells, without roentgen-ray treatment.

The transfusions also had a varied effect on the patient. They
were twelve in number and we gave a total of 2,760 c.c. The hemo-
globin and red cell count gradually diminished. In the beginning
it seemed as if the transfusions might be a factor in lowering the
white cell count and improving the patient's general condition. The

6. Houghwout, F. G., and Azuzano : Notes on the Treatment of a Case of
Lymphatic Leukemia with Benzyl Benzoate, New York M. J. 110:180 (Aug.)
1919.

182 ARCHIVES OF LXTERXAL MEDICINE

last few transfusions seemed to have had no effect whatever on the
patient. With the first few transfusions there were very severe reac-
tions. We were able to eliminate these reactions practically entirely by
the use of small doses of morphin and atropin, given immediately after
the transfusions.

An interesting finding during the last few days of the patient's life
was the peculiarly changed morphology of the red blood cells. Whereas
at the beginning the great majority of cells were very pale and irregular
in size and shape, toward the end of the patient's life, the majority
of cells appeared like perfectly normal red blood cells. Our impression
was that these were functionating transfused cells, and that the patient
was practically living on the transfused blood. Ashby ^ has shown that
transfused red blood cells may functionate in the recipient for thirty
days.

8. Temperature Curve. — The temperature varied between 96.6 and
104 F. A rise in the temperature was usually associated with a rise
in the white count, enlargement of the spleen, lymph nodes and tonsils,
and a marked increase in general malaise. The whole picture would
suggest an exacerbation of an infective process.

9. Blood Culture Studies. — Blood studies on culture mediums and
by injection into rabbits were negative. Baar and Kornitzer * and
several other authors report positive cultures in the blood of leukemic
patients. These, however, were usually found just antemortem and
were probably due to secondary infections. To date no one has suc-
ceeded in transfering human leukemia to laboratory animals. Ellerman
and Bang " could produce anatomic and hematologic leukemic lesions in
healthy hens by injections of cell free Berkefeld filtered organ emulsions
of a leukemic hen. Hirshfeld and Jacoby " observed spontaneous leu-
kemia in hens and were successful in transmitting the disease through
five generations. Schmeisser ^^ also observed sixjntaneous leukemia in
fowls and was successful in transmitting it into other fowls. Ellerman ^^
claims that a myeloid type of leukemia may occur in one generation

7. Ashby. W. ; Some Data^ii Range of Life of Transfused Blood Corpuscles
in Persons Without Blood Diseases, M. Clinics N. America 3:783 (Nov.) 1919.

8. Baar, V., and Kornitzer, E. : Ein positiver Bakterienbefund bei einem
Fall von chronischer myeloischer Leukiiraie (Myeloblasten Leukamie), Wien.
klin. Wchnschr. 32:857, 1919.

9. Ellerman and Bang: E.xperimentcUe Leukamie bei Hiihnern, Zentralbl.
f. Bakteriol. 46:4, 1908.

10. Hirschfeld, H., and Jacoby, M. : Zur Kenntniss der iibertragbaren
Hiihnerleukamie, Berl. klin. Wchnschr. 46:159, 1909.

11. Schmeisser. H. C. : Spontaneous and E.xperimcntal Leukemia of Fowls,
J. Exper. M. 22:820, 1915.

12. Ellerman, V.: Untcrsuchungcn vibor das Virus dcr Hiibnerleukainie,
Ztschr. f. klin. Med. 74:43, 1914.

flXEMAX-LVMPHOIDOCVTIC LEUKEMIA

183

.and a lymphatic type in the next, and that this is highly suggestive that
both forms are due to the same infective agent.

10. Metabolic Studies. — A. The urine and stool chemistry were
studied by Drs. Egerer-Seham and Frances Ford. W'e could not demon-
strate any definite relation between our blood and clinical findings and
the chemical findings. Tables 2 and 3 give the findin s of ^y^\y a f. \v
days on which high fluctuations in white cell count and splenic dimen-
sions occurred.

In Krjukow's case the diminution of splenic enlargement was
associated with an increased excretion of ureates.

TABLE 2. — Chemistrv of Twenty-Four Hour Specimens of Uri

Inor-

gaoic

Creat-

Uric

Date

White Blood

NaCl,

N.

Urea N,

NH3,

Total

Acid.

Cells

Gm.

Gm.

Gm.

Gm.

Phos-
phates,
Gm.

Gm.

Gm.

1/20/19

1 99,000

4.ee

4.89

....

1/24/19

4,800

5.73

6.93

219/19

208,800

0.70

1.86

1.09

0.25

6.17

0.10

2/22/19

4.84

7.21

3.82

5.87

1 0.07

1.90

8/ 3/19

1T6.000

2.31

6.24

2.97

l.Ot

1 0.53

1.12

0.88

S/ 4/19

106.000

3.78

6.09

3.11

0.85

1.84

0.99

8/ 7/19

3.22

9.82

4.77

0.88

1 2.59

2.21

2.87

2.39

3.62

2.20

1.T6

3/ 8/19

5,600

2.41

7.14

3.96

0.87

0.81

1.34

1.88

1.43

2.01

l.OO

0.37

0.50

*l 1/19

4801000

4.08

4.66

3.68

0.48

1 O.08

0.4fi

0.25

4/ 2/19

500,000

1.06

3.70

1.62

0.56

0.52

TABLE 3.— Chemi

lEKTY-FouR Hour Feces

White Blood Cells

2 27/19..
2/28/19. .
3/ 1/19..
8/ 8/19..
8/ 4/19..
8/16/19..
3/24/19..

Ordway " reports that radium applications over the spleen in leu-
kemia increase the protein and phosphate constituents of the urine.

B. Blood chemistry studies did not yield data that could be corre-
lated in any way with the white cell count variations.

C. Basal metabolism studies were as follows :
March 1 1 : +20 per cent. ; white blood cells, 5,600
March 14 : + 8 per cent. ; white blood cells, 7.000

13. Ortlwav. T. : Metabolism in Leukemia and Carcinoma During Radi
Treatment. J.' A. M. A. 73:860 (Sept. 3) 1919.

1^ ARCHIVES OF INTERNAL MEDICINE

March 23: +29 per cent.; white blood cells, 17,000
April 3 : +41 per cent. ; white blood cells, 485,000

11. Blood in Feces and Urine. — Feces and urine showed no chemical
or microscopic blood at any time.

12. Spinal Fluid. — The first specimen gave a positive Wassermann,
but was otherwise normal. Fluid obtained at a subsequent puncture
was normal.

13. Kidney Function. — The phenolsulphonephthalein excretion,
Mosenthal test and blood chemistry (sugar, urea nitrogen, and
creatinin) gave normal values.

14. Spleen. — At first the spleen did not seem to play much of a
role in the leukemic process. January 21 it was palpable about 3 cm.
below the costal margin in the midclavicular line. From January 21 to
March 3 marked fluctuations in the size of the spleen occurred. It
was rather a peculiar effect or coincidence that roentgen-ray exposure
of the mediastinal, facial and neck lymph nodes was followed by
marked diminution in the size of the spleen, followed, however, each
time by an enlargement of the spleen greater than on each previous
occasion (Figs. 2, 5, 7 and 20).

March 6, the spleen was palpable in the midline of the abdomen
and 10 cm. below the costal margin in the midclavicular line. It was
again exposed to the roentgen rays, and again, either as a result of
the roentgen-ray exposure or simply as a pure unexplainable coinci-
dence, it had reduced in size in forty-eight hours to such an extent that
it was just barely palpable (Fig. 20). In this connection it is of interest
to note that the white cell count fell coincidently to 2,300 per cm.

March 8, the spleen began to enlarge again very rapidly (Fig. 5)
and by March 31 it was 5 cm. beyond the midhne and 16 cm. below
the costal margin in the midclavicular line. This rapid enlargement
took place in spite of three exposures to the roentgen ray (Fig. 2). The
white cell count also did not seem to be influenced by the roentgen-ray
exposures and rose from 2,300 to 646,000 on the day of death (Fig. 6).

15. Lymph Nodes. — On entrance to the hospital the lymph nodes
of the face, neck, axillae, groin and mediastinum were markedly
enlarged. Photographs (Figs. 1 and 8) taken before and after roent-
gen-ray exposures show a remarkable difference in the size of facial
lymph nodes. Here, again, it is an open question as to whether these
changes were caused by the roentgen-ray therapy or wliether they
were simply a part of the peculiar unexplainable fluctuations of the
lymphopoietic system. Figure 2 shows the relati\e fluctualions in the
size of the facial lymph nodes.

A comparison of Figures 3 and 11 shows a definite diminution in
the size of the mediastinal shadow. This diminution occurred within

FIXEil.iX—LYMPHOIDOCYTIC LEUKEMIA 185

fortj'-eight hours after roentgen-ray exposure of the chest. Plates
9 and 10 showed the same thing.

16. Toiisils. — The tonsils also seemed to take a very active part
in the leukemic process. With rises in the white count and enlarge-
ment of the lymph nodes or spleen, they too would enlarge, so much
so, that not only would nasal breathing become impossible but even
mouth breathing would be very difficult. Roentgen-ray exposure over
the face and neck seemed to produce a definite diminution in the size
of the tonsils so that the patient could breathe with ease through the
nose and mouth (Table 1).

17. Lh'cr. — At first the liver was not palpable. Toward the end,
it also enlarged and could be palpated 10 cm. below the costal margin
in the midclavicular line.

Fig. 6. — Curve of daily white cell count, red cell count and percentage of
hemoglobin.

CLINICAL SUMMARY

1. In spite of the twelve transfusions of a total of 2,760 c. c. blood,
the patient's hemoglobin percentage and red blood cell count steadily
diminished. With a hemoglobin of 15 per cent., on the Von Fleischel
Miescher instrument, and a red count of 900,000, the majority of red
blood cells were full sized, round, stained well and had every appearance
of normal cells. It seems probable, therefore, that the functioning
transfused red blood cells prolonged an illness, usually acute and
of very short duration.

2. The white cell count showed extraordinary sudden fluctuations.
A fluctuation of 70,000 cells in six hours occurred March 29. The
blood chart (Fig. 6) shows fourteen additional sudden fluctuations.

186 ARCHIVES OF IXTERXAL MEDICINE

3. A study of the blood chart (Fig. 6) shows an apparent
rhythmical occurrence of these fluctuations, a seemingly definite cyclic
change of from five to seven days' duration.

4. Numerous beautiful cells in all stages of mitosis were demon-
strable in the blood, both in the 1 per cent, acetic acid solution and in
the dry stained smears.

5. Injection of the patient's blood into a rabbit was negative.

6. Morphin and atrophin, administered after transfusions, elimi-
nated practically all reaction.

7. We could not demonstrate a definitely beneficial effect from
roentgen-ray therapy over lymph nodes, spleen and long bones. At
first, the spleen diminished in size when roentgen-ray therapy was
applied to the chest and lymph nodes or long bones. Whether this
was purely coincidental or whether it had any relation to the roentgen-
ray therapy, we do not know. One exposure over the spleen was
associated with a very pronounced diminution in the size of the spleen,
Subsequent roentgen-ray exposures had no effect whatever.

Irradiation of the facial, cervical, axillary and mediastinal lymph
nodes by the roentgen ray was followed by marked diminution in their
size. Whether this was a direct result of the therapy, or simply coinci-
dent with it, is an open question. It is of interest to note that cervical
and facial roentgen-ray irradiation was followed by a marked diminu-
tion in the size of the tonsils.

8. Enormous rapid fluctuations in the white cell count occurred,
which could not be accounted for by therapy.

9. A marked rise in the white cell count was usually preceded and
accompanied by a rapid enlargement of the spleen or of some group
of lymph nodes.

10. The clinical picture of more or less rhythmically varying white
blood cell counts, clinical improvement during the periods of low
white cell count, and diminution in size of the lymph nodes or spleen,
suggest the possibility of an infectious etiology of the disease.

1 1. The biopsy of a lymph node showed that at least a great number
of "micromyeloblasts," a cell definitely myeloid according to the dualist
view, were being generated in the germ center of the lymph node
follicles and were passing out from the lymph node into the blood
stream. Such a possibility has always been denied by the dualists.

12. Clinically, our case had all the earmarks of a lymphatic leukemia.
The blood, however, showed in great numbers a cell, the so-called
"micromyeloblast," which is believed by the dualists to originate in
the myeloid tissues. To date, the dualists deny the possibility of such
a cell originating in the germ center of a lymph node follicle.

FIXEMAX-LYMPHOIDOCYTIC LEUKEMIA 187

A lymph node obtained under the very best possible conditions
oflfers very good evidence in flat contradiction to the dualistic view.

This case is reported, because it offers strong evidence in favor of
the unitarian theory of the origin of blood cells.

P.\RT II

MORPHOLOGIC STUDY

Modern heniatologists are divided into two strong groups, the
so-called unitarians or monophyletists, on the one hand, and the dualists
and polyphyletists on the other hand. The bone of contention between
these groups is the so-called ".stem cell."

Fig. 7.— Composite tracing showing increase in size of spleen from January
27 to February 14. Slight diminution in size twenty-four hours after roentgen-
ray treatment over spleen.

The dualists and polyphyletists consider the lymphopoietic and
myelopoietic tissues as two separate tissues, entirely distinct from each
other and never interchangeable. They contend that all the cells
produced in lymphopoietic tissues come from their own specific stem
cells, the "lymphoblasts," and in the same manner all cells produced
in myelopoietic tissues come from their own specific stem cells, the
"myeloblasts." They claim to be able to demonstrate a difference
between "lymphoblasts" and "myeloblasts."

188 ARCHIVES OF IXTERXAL MEDICINE

The unitarians or monophyletists deny the specificity of 'lymph-
opoietic" and "myelopoietic" tissues. They present admirable evidence
that under certain conditions myeloid cells may be produced by "lymph-
opoietic" tissues, and, vice versa, lymphoid cells may be produced in
"myelopoietic" tissues. Pappenheim and his followers ^* claim that
all blood cells spring from a single stem which they call "lymphoidocyle."
While some unitarians do not admit that there is a morphologic
difference between the "myeloblasts" and the "lymphoblasts," other
observers grant such a possibility, but deny the specificity of the mother
tissues from which these cells come.

Among the chief exponents of the unitarian theory are Pappen-
heim,^^ Grawitz,''' Weidenreich,^^ Maximow,'* Downey,^" Ferrata,-"
Du Toit,-^ Arnold,-- Neumann,^^ May,^^ and Danchakoff.^^ Among
the chief exponents of the dualistic and polyphyletic theories are

14. Pappenheim and Hirsclifeld : Ueber akute Myeloide und Lymphadenoide
Makrolymphozytare Leukamie an der Hand von zwei vershiedenen Fallen,
Folia haematol. 5:347, 1908. Pappenheim, A.: Atlas der menschlichen Blut-
zellen, Jena, 1905-1912. Morphologische Hematologie, Leipzig, Werner Klink-
hardt, 1919.

15. Pappenheim, A.: Bemerkungen iiber artliche Unterschiede, usw., der
lymphoiden. Zellformen des Blutes, Folia haematol. 9: 321, 1909. Clinical
Examination of the Blood and Its Technique (translated by R. Donaldson),
New York, Wm. Wood Co., 1914.

16. Grawitz, E. : Klinische Pathologic des Blutes, Leipzig, Georg Thieme,
1911.

17. Weidenreich, F. : Die Morphologic der Blutzellen und ihre Beziehung
2u Einander, Anat. Rec. 4:317, 1910.

18. Maximow, A.: Ueber die Zellformen des lockercn Bindegewebes, Arch,
f. mikr. Anat. 67:680, 1906. Ueber embryonale Entwicklung der Blut und
Bindegewcbezellen bei den Saugetieren, Verhandl. der Anat. Gesellsch, 22,
Vers., Berlin, 1908. Experimentelle Untersuchungen zur post fotalen Histo-
geneses des myeloiden Gewebes, Beitr. z. path. Anat. u. z. Allg. Path. 41:

122, 1907.

19. Downey, Hal, and Weidenreich, F. : Ueber die Bildung der Lymphoc>'ten
in Lymphdriisen und Milz, Arch. f. mikroscop. Anat. 80:367. 1912. Downey:
The Development of the Histogenous Mast Cells of Adult Guinea-Pig and Cat,
and the Structure of the Histogenous Mast Cell of Man, Folia haematol. 16:
1913. On the Development of Lymphocytes in Lymph Nodes and Spleen, Tr.
Minnesota Path. Soc, 1912-1914, p. 91.

20. Ferrata, A.: Le Emopatie. Milano, Societa Editrice Lebraria 1: 1918.
(Reviewed in Folia haematol. 20: 182, 1920, by Downey.)

21. Du Toil. P. J. : Beitrag zur Morphologic des normalen mid des leu-
kamischen Rinderblutes, Folia haematol. 21:1, 1916.

22. Arnold, J.: Zur Morphologic und Biologic der Zellen des Knochenmarks,
Arch. f. path. Anat. u. Physiol. 140:411. 1805.

23. Neumann, E. : Hematologischen Studicn der Lcukozyten und Leukamie,
Arch. f. path. Anat. u. Physiol. 207:379, 1912.

24. May: Quoted by Naegcli.

25. DanchakofT, V.: Concerning the Conception of Potentialities in the
Embryonic Cells, Anat. Rec. 10:415, 1916.

The Origin of Blood Cells, Anat. Rec. 10:397. 1916. The Wandering Cells
in the Loose Connective Tissue of the Bird and Tlicir Origin, Anat. Rec.
10:483, 1916.

FINEMAX—LYMPHOIDOCYTIC LEUKEMIA 189

-\aegeli,= Ehrlich," Stockard," Ziegler,^* Turk,=« Schridde,'" Fischer,"
Butterfield, Stillman Meyer and Heinecke.'-

Ehrlich" (1880) divided all white blood cells into granulated and
nongranulated forms. It is this division which forms the basis of the
modem dualistic teaching. He placed the lymphocytes, large mononu-
clears, and transitional cells among the nongranulated cells. The
eosinophilic, basophilic and neutrophilic polymorphonuclears were the
granulated cells. He believed the transitionals to be an intermediate
form between the mononuclears and the neutrophils. Ehrlich believed
that the lymphocytes came from lymphoid tissue, i. e., lymph node and
spleen follicles, and that the granulocytes came from myeloid tissue,
i. e., principally bone marrow. Today, this teaching, practically
unchanged, is accepted by the dualists and the great majority of
clinicians.

According to Naegeli's scheme the mesenchyme cell gives rise to
the normoblast and this in turn to the normocyte. The mesenchyme
cell also gives rise to the lymphocyte of the "quiet zone" of the follicle,
this in turn to the "lymphoblast" of the germ center, and the "lymph-
oblast" gives rise to the small lymphocyte of the blood. The
"myeloblast," the "stem cell" of the monocytes, megakaryocytes and
polymorphonuclears, also comes originally from the mesenchyme cell.
In postfetal life, therefore, all the blood cells come from their own
specific "stem cells." He denies all transitions between the myeloid
and lymphatic systems. He and other dualists meet the argument
that myeloid metaplasia may occur in the spleen and lymph nodes in the
absence of myeloid elements in the blood by declaring that lymph
nodes and spleen are composed of two types of tissue, myeloid and
lymphoid. The lymph node and spleen follicles are supposed to be

26. Ehrlich, P., and Lazarus. A.: (Rewritten by A. Lazarus and O. Naegeli,
and translated by H. .Armit) Anemia. New York, Rebman Co.. 1910, p. 148.

27. Stockard, C. : The Origin of Blood and Vascular Endothelium in
Embryos Without a Circulation of the Blood and in the Normal Embryo,
Am. J. Anat. 18:227, 1915. A Study of Wandering Mesenchymal Cells on the
Living Yolk Sac and Their Developmental Products ; Chromatophores, Vas-
cular Endothelium and Blood Cells. Am. J. Anat. 18:525. 1915.

28. Ziegler. R. : Experimcntellc u. klinischc Untersuchungen iiber die
Histogenese der myeloiden Leukamie, Jena, G. Fischer, 1906.

29. Tiirk, W. : Ueber Regeneration des Blutes unter normalen und krank-
haften Verhaltnissen, Centralbl. f. Allg. Path. u. Anat. 19:895, 1908.

30. Schridde, H. : Die embryonale Blutbildung. Erwiderung an Herrn
Prof. A. Maximow, Zentralbl. f. Allg. Path. u. Anat. 20:433. 1909.

31. Fischer, H. : Myeloische Mctaplasie und fotale Blutbildung, Berlin,
Julius Springer, 1910.

32. Butterfield, E. E., Heinecke, A., and Meyer, E. : Ueber das Vorkommen
der Altmannschen Granulationen in den weissen BlutzcUen, Folia haematol.
8:. 325, 1909. Butterfield, E. E.. and Stillman. R. G.: The Broader Aspects of
Hematologic Diagnosis, Am. J. M. Sc. 154:781 (Dec.) 1917.

33. Ehrlich, P. : Farbenanalytischc Untersuchungen zur Histologic und
Klinik des Blutes, Gesammeltc mitteilungen, I Teil, 1891, Berlin.

190 ARCHIVES OF INTERNAL MEDICINE

lymphoid, while the inter follicular tissue of the nodes and splenic pulp
are myeloid. They further claim that these two types of tissue are
sharply contrasted and are distinct from each other. The myeloid
function, however, does not manifest itself normally, but comes into
play in anemias, infections, and myelogenous leukemias.

Schridde ™ derives his lymphocytes from the endothelial lining of
lymph vessels, and his myeloid tissue from the endothelial hning
of blood vessels. This view, however, is untenable, because Thiel and
Downey^* have shown that in the spleen, where a large production
of lymphocytes takes place, there are no lymph vessels.

Fig. 8. — Photograph of patient Feb. 17, 1919. seventy-two hours after
roentgen-ray treatment of neck and face. See Figures 1 and 2.

Fischer," a strong supporter of Naegeli's views, admits that myeloid
cells may develop from endothelium of blood vessels and from connec-
tive tissue cells.

Hirschfeld ^' admits that the spleen and lymph nodes can produce
granulocytes. They come, however, not from the specific follicular
tissue but from the pulp and interfoUicular tissue, probably from the
perivascular tissue (Hirschfeld).

34 Thiel G. A., and Downey, H. : The Development of the Mammalian
Spleen, with Special Reference to Its Hematopoietic .Activities, Am. J. Anat.
28:279, 1921. . „

35. Hirschfeld, H.: Die Unitarische und die dualistische .Auffassung uber
die Histopathologic der Leukamien, Folia haematol. 6: 1, 1908.

£ '*• •-;

i ^s

^1^1 =1 -1 =1

^3

•25,^.1
.'i.4

f V m ,

'1

!l s| i| i§ ii ii n p

Siii

IPS HP ii|

s

'1

s|i| i| upiipsi

i| ii

=|irlP=«

5=1

i| !i it i^ i M i 1^ P

i^ N

it iiPPi'i

s|

|S p |§ ; ; S5 ■ : |S |s

S8 ::^

n iiss|sas|

^1

51 ::| ipi §i n n n

Si i j §1

if imv%

'1

S| S| || IE P §8 : : p

^MMi

l| |iilP=S|

sl

^H ii n n 2- M ^^

SS 2S : :

S| « 56. P =*|

g

SS^i si iiS^S:^ ni=^

i^ Ml:i

5= iiSiiiS'l

„2
"1

s| 5S P i i ^^ ; M i ^^

ss ; i §s

31 S6 S| ii 5E|

i

2| S| ^i i ; 2s ; ; : ; Si

s| M=l ir*!

"2

il ai SI n SI 21 ;; iS

Si ;; ;;

ii n i

ilS8|

2

p p II M 2§ n i^ IS

S5 MiiP Ml

-|

"1 2| ii M is sg n

P M N

^ nil H Pi

cS^

3M§ ^ ]\n w \\u

=1 se S| IS SS|

-1 n =1 ^ w ^^ s2 is p

S5 ; : ; ;

ii S5 Si P s'l

2

-5

11 P L'§ i^ "§ i^ s^ "i

P M N

Si ii i

|i ais

■1
t

S| ^1 2| 51 :| SI 21 21

n M H

SI i i SI II :«|

^iP iB n^^iii^i^

is n n

?s ii s'Pii!

"i "? S| : ; 2g 2s ; i Sg

"!

ii'sp'si

II nil tlilillillijill 1

iJiWLiJIifiiiiii

1 1 Id i i|

192 ARCHIVES OF INTERNAL MEDICINE

Stockard ^^ adopts the extreme polyphyletic view. The fixed
mesenchymal cells of the embryonic body, even before they begin their
migration are already specialized to such an extent that differentiation
can take place in one direction only.

Pappenheim and his pupils ^* hold a moderated monophyletic view.
The "histioidocyte" a tissue cell gives rise to the "lymphoidocyte."
The lymphoidocyte is the only stem cell and gives rise to megakary-
ocytes, monocytes, polymorphonuclears, lymphocytes and red blood
cells. The lymphocytes are fully differentiated cells and can not change
into granulocytes.

Ferrata,^" and his pupil Neigreiros-Rinaldi, accept a modified
monophyletic view, similar to the one proposed by Pappenheim. They
believe that all the blood cells come from a single stem cell, which
they call the "hemocytoblast," they do not believe that fully differ-
entiated lymphocytes are capable of differentiating into granulocytes or
red blood cells.

Ferrata considers the connective tissue as a dift'use hemopoietic
organ. This tissue gives rise to the "hemohistioblast" (resting-wander-
ing cells of Maximow ; clasmatocyte of Ranvier^^). In the early
embryo this cell differentiates into the primitive transitory hemocyto-
blast, and this, in turn, to the primitive red blood cell of the embryo
(megalocytes) ; while in the adult, lymphoid and myeloid hemocyto-
blasts (functional differences only) and monocytes are the products of
its differentiation.

Grawitz," Weidenreich,^" Danchakoff,"^ and Maximow,'* hold the
extreme monophyletic view. They believe that all blood cells may
come from fixed tissue cells, such as reticular cells, fixed cells of
omentum and messentery, as well as from the widely distributed
clasmatocytes. These may give rise to free lymphoid cells, and these
lymphoid cells may occur anywheres in the body of the adult.

Fully differentiated lymphocytes derived from these lymphoid
cells may, according to some authors, dedifferentiate into the more
primitive type of lymphoid cell, which, in turn, might differentiate
into other forms of white cells, or these fully differentiated lymphocytes
might metamorphose into other forms of leukocytes without dedifferen-
tiation. Grawitz believes that lymphocytes can change into granulocytes
in the circulating blood.

Weidenreich " and Downey '° do not recognize the term "lyniph-
oblast." They have shown that all the types of lymphocytes are
concerned in regeneration and differentiation from one type to another.
The reticulum of lymphoid tissue, wherever found, serves as a mother
tissue. All types of lymphocytes may be derived from it, and these,

36. Ranvier, L. : Des Clasmatocytes, Arch. d'.Anat. micr. 3:122, 1900.

FISEMAS—LYMPHOIDOCYTIC LEUKEMIA 193

in turn, may become transformed into other types of lymphocytes.
Therefore, no one type of lymphoc}'te can be recognized as being
more highly differentiated than any other type. They claim, therefore,
that one cannot speak of a "stem cell" of the lymphocyte in the same
sense that one can speak of the "stem cell" of the myeloid cells.

Lymphocytes are not in any sense of the word a final product.
Downey,'-' Weill, ^' W'eidenreich,'' and other observers have shown that
all forms of lymphocjtes cSn differentiate specific leukocyte granules,
thereby becoming granular leukocytes.

Such widely differing opinions indicate the unsettled state of many
hematologic questions. These questions regarding the relationship
of the blood cells depend primarily on the various theories of postfetal
regeneration. Is Xaegeli ^^ correct, for instance, in assuming that the
"myeloblast" is a specific myeloid cell which is in no way related to
the lymphatic tissues? He claims that he can differentiate morpho-
logically between the myeloblast and large lymphocyte (lymphoblast).
Pappenheim " claims that Naegeli's myeloblast is not a stem cell but is
the cell which he terms the leukoblast, a cell already partly differen-
tiated along myeloid lines.

The following points are usually considered by the dualists as
differentiating between "myeloblasts" and "lymphoblasts."

1. Character of Chromatin Arrangement. — Naegeli ^'■' (1907)
describes very poor chromatin content as characteristic for lympho-
blasts. Klein*" (1910) finds that the chromatin content is not char-
acteristic of the cell. He shows lymphoblasts which are identical
with du Toit's -^ and Pappenheim's lymphoidocytes and with Naegeli's
myeloblasts.

2. Number of Nucleoli. — Naegeli (1907) claimed that nucleoli of
the '"lymphoblasts" are one or two in number, and of myeloblasts from
two to four in number. Pappenheim,*' Klein,*^ Butterfield *■ and others
have shown that this is not true. Naegeli himself admitted later'
(1912 and 1919) that this does not hold true in pathologic blood.

3. Altmann-Schridde Granules. — The dualists believed that they
occurred in lymphoid cells and were absent in myeloid cells. Tiirk

i7. Weill, P.: Ueber die Bildung von Leukozyten in der Menschlichen und
tierischen Thymus des erwachsenen Organismus, Arch. f. mikro. Anat. 83:305,
1913.

38. Naegeli, O. : Blutkrankheiten und Blutdiagnostik, Leipzig, 1919, Dritte
auflagc.

39. Naegeli: Quoted by Du Toit.

40. Klein : Ueber die groszen einkernigen Leukozyten des Leukamieblutes,
Folia haematol. 10: 1, 1019.

41. Pappenheim and Klein : Quoted by Du Toit.

42. Butterfield : Quoted by Damarus.

194 ARCHIVES OF IXTERXAL MEDICINE

(1912)" has shown that these granules are not specific for lymphoid
cells.

4. Oxydase Reaclion.—Tht dualists claim that the oxydase reaction,
the staining of the oxydase grandules with the indophenol blue dye, is
positive in myeloblasts and negative in lymphoblasts.

Hyneke," Decastello,^= Bliihdorn, Jochman,*" Glaus ^' and Dunn *»
report cases in which the so-called "myeloblasts" did not give the
oxydase reaction. Klein *' showed that ordinary lymphocytes may give
a positive reaction.

Schultze =*° found that the myeloblasts of the normal marrow usually
did not give the reaction while the pathologic myeloblasts in the blood
and in the organs usually do give the reaction.

According to Menten ^' the reaction is not specific for myeloid cells.
She found that lymphocytes may give a well marked reaction. She
also found that all tissues with the exception of bone give this reaction,
so that it is by no means specific for blood cells.

Forman and Hugger ^= found "large mononuclears" both with and
without the indophenol oxydase granules.

Boechat," Belz,^* Steftan ^^ and Krjukow " claim that tlie reaction
may be absent in myeloblasts.

Naegeli himself wavers on the question of the specificity of the
oxydase reaction. In his book^^ (p. 211) he makes the statement "the
oxydase reaction is present in normal myeloblasts and generally in
pathologic myeloblasts, but in some of these the reaction may be not
so strong and less diffuse."

43. Turk, W.: Vorlesungen uber klinische Haeniatologie, Wien. u. Leipzig,
W. BVaumiiller, 1912, Pt. 2, p. 131.

44. Hyneke, K. : Zur Monocytenfrage. Folia hacmatol. 13:345, lyi^J.

45. Decastello: Quoted by Hyneke.

46. Jochman and BKihdorn: Ueber akute Myeloblasten Leukamie, Folia
haematol. 12:1, 1911. ^ ,

47. Glaus : Quoted by Kahle, H. : Ueber ein Hamogonien und Leukozyten
erzeugendes Angiosarkom in zirrhotischer Leber, Arch. f. path. Anat. u. Physiol.
226:44, 1919.

48. Dunn, J. S. : The Use of the Oxydase Reaction in the Differentiation
of Acute Leukemias, Quart. J. M. 6:293, 1913.

49. Klein : Quoted by Neumann.

50. Schultze, W. H.: Zur Differentialdiagnose der Leukamieen, Munchen.
med. Wchnschr. 56:167, 1909.

51. Menten, M. L. : A Study of the Oxydase Reaction with a-naphtha-
paraphcnvldiamcne, J. M. Res. 40:433, 1919.

52. Forman. J., and Hugger, C. C: Nature of Mononuclear Cells Seen in
Exudate of Lobar Pneumonia Accompanying Typhoid Fever, Am. J. M. Sc.
155: 317 (March) 1918.

53. Boechat; Uelier akute Myeloblastenleukamie mit teilweise chloroma-
tosen Charaktcr, Frankf. Ztschr. f. Path. 13:489, 1913.

54. Belz, L. : Ueber Leukamie mit besonderer Beriicksichtigung der Akuten
Form, Deutsch. Arch. f. klin. Med. 113:116, 1914.

55. Steffan, M.: Ueber einen Fall akuten Myeloblasten Leukamie und
iiber die Beziehungen Leukamie-Sepsis, Folia haematol. 21:59, 1916.

FIXEMAX-LYMPHOIDOCYTIC LEUKEMIA

195

Pappenheim and Dolirer's '-^ conclusion regarding the oxydase
reaction is, that it does not indicate an absolute unfailing histogenetic
difference between two heterogenous cell races. It is clear, therefore,
that the opinion of many of the observers is against the specificity
of the oxydase reaction and against any of the other differential
points usually picked on by the dualists for the differentiation of
hinphoblasts and myeloblasts.

Du Toit =' says that morphologically we cannot differentiate between
"lymphoblasts" and "myeloblasts."

L-nogram of chest. Jan. 21, 1919.

Krjukow,' referring to his case of microlymphoidocytic leukemia,
says: "It seems to me that the most skilful dualists could not possibly
differentiate our microlymphoidocytes from true micromyeloblasts,
except, perhaps, by the use of the various methods that the dualists
are wont to use."

Hirschfeld '•■ could not see any morphologic difference between
myeloblasts and germ center cells.

56. Diihrcr and Pappenheim. .\.: Ein weitercr Fall von akuten Mikro-
lymphoidozytcn leukamie. Folia haematol. 16:143, 1913.

57. Hirschfeld, H. : Die Unitarische und die Dualistischc Auffassung iihcr
die Histopathologie der Lcukamien, Foh'a haematol. 6:382, 1908.

196

ARCHIVES OF INTERNAL MEDICINE

Damarus,^* Wolf and Tiirk =^ and Butterfield '^° can see no mor-
phologic difference between the stem cells of lymphocytes and granu-
locytes.

Weber ^^ describes a case of leukemia in which "the conclusion
was unavoidable that the lymphoid cells which permeated the various
tissues of the body were of the same kind as the lymphoid cells which
during the patient's life constituted by far the greatest portion of the
white cells of his circulating blood. I thought at the time that the
cells in question were probably to be regarded as lymphoblasts but

Fig. 10. — Roentgenogram of chest, Jan. 31, 1919, ten days after roentgen-
ray treatment of mediastinum. Compare witli Figure 9.

from the fact that a few of the cells gave a positive oxydase reaction
and from a comparison with Case 4, I now think that they were
probably myeloblasts."

58. Damarus : Der gegenwartige Stand der Leukamiefrage, Folia haematol.
6:, 337, 1908.

59. Wolf and Turk: Quoted by Damarus.

60. Butterfield, K. E. : Ueber die ungranuliertcn Vorstufen der Myelocyten
und ihre Bildung in Milz, Leber und Lymphdriisen, Deutsch. Arch. f. klin.
Med. 92:336. 1903.

61. Weber, F. P.: Acute Leukemia and So-Called Mediastinal Leukosar-
comatosis (Sternberg), with the Account of a Case Accompanied liy Myeloid
Substitution of the Hilus Fat of the Kidneys, Quart. J. M. 12:212 (April) 1919.

FIXEMAX-LYMPHOIDOCYTIC LEUKEMIA

197

Similarly Chosrojeff "'- reports a case of "micromyeloblastic
leukemia" in which he found in the blood all transition forms from
myeloblasts to small lymphocytes and "were it not for the oxydase
reaction, one could not tell what type of leukemia he was dealing with."

Walter Schultze,^" a dualist, admits that the morphologic differences
between myeloblasts and lymphoblasts, described by Schridde, are so
slight that the majority of hematologists cannot differentiate the two
types of cells.

We have shown so far that "lymphoblasts" cannot be differentiated
from "myeloblasts" either by morphology, that is cytoplasm, nuclear
chromatin arrangement, nucleolar content, Altmnnn-Schridde granules.

on Percussion
AMD Palpation

Fig. 11. — Percussion tracing January 27,
mediastinal dulness. Compare with Figure 3.

showing diminution in size of

or by the oxydase reaction. The dualists trace all the white blood cells
from these two stem cells and claim that the lymjihoijoietic and myelo-
poietic tissues derived from these two mother cells are entirely distinct,
and that lymphoid tissue cannot produce myeloid cells and, vice versa,
that myeloid tissue cannot produce lymphoid cells. In the presence of a
vast amount of recently accumulated evidence to the contrary, both
experimental and clinical, it becomes doubtful whether there exists
such a cell as is usually described by the dualists as a "lymphoblast."

The following points are evidence against the dualistic and
polyphyletic theories.

62. Chosrojeff. B. P.
Haematol. 20:33. 1915.

Myelosis aleucc

acuta niic

vcloblastica, Folia

198 ARCHIVES OF ISTERSAL MEDICI XE

1. Downey and Weidenreich have demonstrated that cells such as
are described for the germ centers may be found in the interfollicular
tissue, in the spleen pulp, and in the lymph vessels. They have also
shown that morphologically and genetically the lymphocytes of the
cortex and medulla of lymph nodes are identical, and that follicles
with germ centers may arise anywheres in the node.

2. Weidenreich describes these same cells in the chyle of the
thoracic duct.

3. Dominici has shown that in certain infections, in man and in
the laboratory animals, neutrophilic and eosinophilic myelocytes and
normoblasts may develop in the spleen and lymph nodes from
lymphocytes.

4. Sacerdotti "^ and Frattin. by tying off and cutting the blood
vessels of one kidney in dogs have produced true bone and bone marrow
in the pelvis of the kidney.

5. Maximow "* reproduced these same results and found that these
granulated and red blood cells of the bone marrow came from typical
lymphocytes of the blood.

6. Downey and Weidenreich '■' ha\'e shown that the lymphocytes of
the germ center are extremelv irregular in size and form and that only
a very few of them conform to the descrijjtions given bv Naegeli.
Tiirk, Pappenheim and others.

7. Naegeli himself at one time ( 1900) describes the lymphob'.asts as
being very rich in chromatin and at another time (1909) as being very
poor in cliromatin. They are very young lymphocytes related to the
mature lym])liiicytes by innumerable transition forms. In children they
may be found at times in normal blood, and are especially prone to
occur in infectious leukocytoses (Naegeli''"').

8. \\'eidenreich and Weill ''' claim that lymphocytes may differen-
tiate into neutrophil and eosinophil leukocytes.

9. Downey "^ has shown that lymphocytes may differentiate into
histogenous mast cells. Weidenreich ;ui(l Downey '■' ha\e >hi>wn granu-
lated myelocytes developing frcmi the lym])hncytcs (if the spleen. They
have also shown granul.-ited celK in mitosis in the spleen.

10. Dominici '■" has obscrxed myelocxtcs in lyni]ih follicles.

63. Sacerdotti. C, and Frattin. G. : Ucber die heteroplastisclie Knochen-
bildung, Virchows .^rcli. f. path. .^nat. 168:4.31. 1902.

64. Ma.xiniow, A.: Expcrimcntclle Uiitcrsuchungcii ziir po.stfotalcn Ilisto-
genese des myeloiden Gewebes, Beitr. z. path. .Atiat. u. z. allg. Path. 41:122. 1907.

65. Naegeli: Bhitkrankheiten iind Rhitdiagiiostik. Leipzig, 1919.

66. Downey. H. : The Development of the Histogenons Mast Cells of .\ih\U
Guinea-Pigs and the -StriK-turc nf the Histosjennns Mast fell of Man. Folia
haematol. 16:70, 1913.

67. Doiriinici: Quoted by HirschfeUl.'

FIXEM.LX-LYMPHOinOCyTIC LEVKEMIA 199

11. In a case of eosinophilic polymorphonuclear hyperleukocytosis
recently reported by Giffin,'"* Downey has found eosinophilic myelocytes
in the germ centers of the follicles of the lymph nodes and spleen.^^

12. Hertz,"" using pyrogallol, has produced myeloid metaplasia of
the spleen pulp with a hyperplasia of the splenic follicles. This is
contrary to the usual statement that myeloid metaplasia is accom.panied
by reduction in the size of the follicles, which is used as an argument
in favor of the supposed antagonism lietween pulp and follicles.

Fig. 12. — The cell.s in A and B are drawn with 1.9 mm. objective and Nr
eyepiece: magnified X2 actual appearance. .'\. Microlymphoidocyte (stem
cell) (micromyeloblast) in blood. B. White cell in mitosis in blood stream.
C. Microlymphoidocyte or stem cell in section. Objective 1.5 mm.; ocular
No. 8; magnification X 8. D. Stem cell or microlymphoidocyte in section
Objective. 1.5 mm.; ocular. No. 8; magnification. X 8. E. Reticulum cell in
section. Objective. 1.5 mm.; ocular. No. 8; magnification. X 8.

1,1 Downey has demonstrated cosinoijliilic myelocytes in the
germinal centers of lymph nodes of rabbits, the subjects of experi-
mental Jiemorrhages in which there were no myelocytes in the blood so
that the development of the=e myelocytes must have taken place in loco.

68. Giffin. H. Z. : Persistent Eosinophi
megaly. Am. J. Med. Sc. 158:618. 1919.

69. Unpublished observation.

70. Hertz: Quoted by Krjukow.'

with Hyperleukocytosis and Sple

200 ARCHIVES 01- IXTERXAL MEDICI. \E

14. Roman '^ has observed germ centers containing myelocytes.
I have attempted to bring out the following points.

1. Hematologists are divided into two main camps — the dualists,
or polyphyletists, and the unitarians.

2. The dualists and polyphyletists believe that the blood cells and
blood forming organs are divided into two main distinct divisions, the
myeloid and lymphoid. These two never change from one into the
other either in the tissues or in the blood.

3. The unitarians believe that all blood cells come from the "lympho-
cyte" found in the bone marrow, spleen, lymph nodes and other tissues
of the body, or. according to Pappenheim and his followers, from the
"lymphoidocyte," which, in the adult, is confined to the marrow.

4. The dualists claim tliat they can difterentiate between the
"lymphoblast" and "myeloblast." They regard the o.xydase reaction
as being of prime importance for such difTerentiation.

5. The unitarians claim that such a differentiation is impossible;
that the "lymphoblast" as a cell per se does not e.xist ; that the
"lymphoblast" is given various descriptions by various authors ; that
Naegeli himself makes contradictory statements regarding this cell;
that the "lymphoblast" may be found anywhere in lymphoid tissue and
not only in the germinal centers, as claimed by the dualists, and that
the oxydase reaction, the main stay of the dualists, is not a specific
test for the myeloblast.

6. A vast amount of experimental evidence is accumulating, which
is in favor of the unitarian theory.

We now wish to call attention to histologic and liematologic
evidence in which we consider to be in faxnr of the unitarian theory,
reported to date in the literature on leukemia.

Dr. Downey has been kind enough to permit me to sui(l\- the blood
from a recent case of leukemia in which all the transition stages between
the lymphoidocyte (myeloblast of Naegeli) and lymphocyte were
present. ' =

Fleischniann ■ ■ reports a case of moimcyte leukemia (second case
on record) in which in June, 191.^, he found 6.^ per cent, mononuclear
leukocytes and 19 per cent, small and large lymphocytes. In Novem-
ber, 1913, a bone marrow puncture gave polymorphonuclears, small
lymphocytes and "mononuclears" identical in morphology with those
found in the blood. Shortly before death, the "mononuclears" dropped

71. Roman. B.: Zur Kenntniss der mveldisclicn tlilorolciikaiiiic. Beitr. z
path. Auat. u. z. allg. Path. 55:61. 1913.

72. This case is not yet reported in the literature.

73. Fleischmaini, P.: Der zweile Fall von Monocyten leukaniie. Folia
haematol. 20:17. 191.i.

FIXEMAX—LYMPHQIDOCYTIC LEUKEMIA 201

to 2i per cent, and there appeared, in the blood, myelocytes, 121/2 per
cent., promyelocytes and "myeloblasts." At the necropsy there was
found myeloid change everywhere. Among the myeloid cells were
cells which corresponded to the blood monocytes. The follicles were
shrunken in the spleen and the lymph nodes. Fleischmann concludes
the report by suggesting the possibility that the monocytes changed into
myeloid elements under the influence of some pathologic stimulus.

Pappenheim,'* Walz,'' and Dennig '" have reported cases which
hematologically were cases of lym])hatic leukemia. Yet, in the?e cases

■ < • ' ■'■■:M

t

r '1

- %

■ "h

\. .

1

%;>^

Fig. 13. — A.xillary lymph node X '
germinal center.

follicle with a

the spleen and lymph nodes were not inxolved, but the bone marrow
had become lymphoid.

Krjukow ^ (1915; reiKirts a case of microlymphoidocytic leukemia
in which the blood showed all transition forms from the stem cell
fo the mature granulocyte, yet the pathologico-histologic changes were

74. Pappenheim, A.: Ueher I,yniphamie ohnc Lvrnphdniscnschwelliinc,
Ztschr. f. klin. Med. 39:171. 1900.

75. Walz : Quoted hy Damarus.

76. DenniK, A.: Uchcr akiite L-.iikamic. Miinclun. nicd. Wrhn^chr 47:
1297, 19C0.

202 ARCHU'ES Of IXTERXAL MEDICI XE

those of a lymphatic leukemia. The bone marrow showed only slight
activity. The pulp of the spleen and some lymph nodes showed slight
myeloid metaplasia. The thymus was hyperplastic and the spleen and
lymph nodes showed typical lymphatic leukemia changes. He concludes
that the myeloid elements of the blood were coming from the lymph
nodes and spleen. He also states that the most skilful dualist could not
differentiate morphologically between micromyeloblasts and his micro-
lymphoidocytes. He is loath to conclude from this that the myeloid cells
of the blood were coming from lymphoid cells and so decides that he
has here a case of "mixed leukemia." with a prevailing "lymphatic
component." He also concludes that in acute leukemias even the
presence in the blood of myelocytes and promyelocytes and stem cells
does not necessarily diagnosticate a myelogenous leukemia.

St. Klein '" ( 1910) reports a case in which the blood showed all the
transition stages between the lymphoidocyte (myeloblast) and the
ordinary lymphocyte. The histologic studies showed a lympholeukemic
involvement of all the lymphoid tissues.

Tiirk ^^ reports a case of mixed cell leukemia which changed into
a lymphatic leukemia.

Hirschfeld '■' has found in cases of mixed cell leukemia hyperplasia
of the lymph follicles and a myeloid metaplasia in the rest of the
lymph nodes. He has also found in a true case of lymphatic leukemia
clumps of myelocytes in the spleen and lymph nodes.

\'eszpremi,'" Glinski."" Walter Schultze. Pappenheim.'" Hirschfeld "-
and W'echselmann describe cases of acute lymphatic leukemia in which
there was no proliferation or atrophy of the follicles of the spleen or of
the lymph nodes or of both with marked proliferation of the splenic
pulp and interfollicular tissues in some of the cases. Hematologically
these cases were the same as any ordinary acute lymphatic leukemia,
yet histologically we find the proliferation where it should occur in
myeloid cases (Hirschfeld''').

Hertz *^ (1909) reports a case of lymphatic leukemia in which 15
per cent, of myelocytes appeared before death. Sections of lymph
nodes showed lymphatic proliferation as well as neutrophilic granulated

77. St. Klein : Quoted by Du Toit.

78. Tiirk: Quoted by Hirschfeld."'

79. Veszpremi, D.: Beitrage zur Histologie der sogenanntcii "akutei
kamie," Arch. f. path. Anat. u. Physiol. 184:220. 1906.

80. Glinski, L. K. : Zur pathologischen .Xnalomie dir akutin Lvnip
Arch. f. path. Anat. u. Physiol. 171:101. 1903.

81. Schultze and Pappenhoim : Quoted by Hirschfeld.

82. Wechselmann. W.. and Hirschfeld. H.: Ucl)cr cinen Fall akuter
loider makrolvmphozvtarer Leukamie niit eigentiinilichen Zelleinsch
Ztschr. f. klin.'Mcd. 66:349, 1908.

83. Hertz, A.: Zur Fragc der geniischtcn Leuk.Hmif, Wien. klin. Wcl
22:1030, 1909.

FIXEM.tX—LVMPHOIDOCyTIC LEUKEMIA 203

myelocytes and erythrocytes. In the bone marrow, follic'.e hke aggre-
gations of lymphocytes were present, besides the myelocytes and
myeloblasts. The spleen pulp was myeloid, while the follicles were
atrophied.

Chosrojeflf "- (1915) reports a case of "micromyeloblastic" leukemia
in which the blood showed all transition forms between the "myelo-

% ^Sy

Fig. 14. — Germ center of follicle shown in Figurt 13. Objective: Zeiss apu-
chromatic 1.5 mm.; ocular, Zeiss compensating \o. 8. Note the numerous
stem cells in this germ center.

blast" and the small lymphocyte. He bases his diagnosis on a positive
oxydase reaction and says, "were it not for this reaction one could not
tell what type of leukemia he was deah'ng witli."

204 ARCHU'ES OF IXTERXAL MEDICJXE

Herxheimer *^ (1913) reports a case of mixed leukemia. The blood
contained 69.8 per cent, small lymphocytes and 25.4 per cent, "myelo-
blasts." The bone marrow showed mostly lymphoc}'tes with a few
islands of "myeloblasts." The spleen showed a hyperplasia of follicles,
which consisted of large and small lymphocytes. The pulp showed a
hyperplasia of its lymphoid constituents and a profuse scattering of
"myeloblasts." The lymph nodes showed a hyperplasia of follicles and
occasional "myeloblasts." In the bronchial nodes the "myelob'asts"
predominated. The mediastinum showed compact parts consisting
largely of lymphocytes and loose areas containing many "myeloblasts."
The "myeloblasts" were diagnosed by means ot the oxydase reaction.

The author regards this positive oxydase reaction as the best indi-
cation of the dualistic origin of the myeloid and lymphoid cells.

From these cases the following is evident :

1. The blood may show all transition forms from the "myelob'ast"
(lymphoidocyte) to the ordinary lymphocyte (Downey).

2. A monocytic leukemia blood and bone marrow picture changed to
a myeloid picture (Fleischmann). The necropsy showed myeloid prolif-
eration everywhere. Among the myeloid cells some monocytes were
present.

3. The blood picture was one of lymphatic leukemia (Pappenheim
and others). The bone marrow had undergone a lymphoid change
and nodes and spleen were normal.

4. The blood showed all transition forms from the "micromyelo-
blast" to the granulocyte (Krjukow). These "micromyeloblasts" were
coming from lymphoid tissue, "from the myeloid parts of these organs."

5. The blood showed all transition stages between the "myeloblast"
and the ordinary lymphocyte (Klein). The tissues showed only
lymphoid proliferation.

6. A mixed cell leukemia changed into a lymiihatic leukemia
(Turk).

7. Clumps of myelocytes were present in the spleen and lymph
nodes in a case of lymphatic leukemia (Hir.schfeld). In a case of
mixed cell leukemia, both myeloid and lymphoid proliferation were
present in the lymph nodes.

8. Cases of acute lymphatic leukemia siidwed atropliy of the fol-
licles of the s])leen or of the lymph nodes or of Ixitii, and a proliferation
of the interfollicular tissue and spleen ])ul]) in sonu' of the cases, tissue
which according to the dualists is myeloid in naturt' ( \■c^zl)rc•mi and
others).

84. Herxlieimer. G. : Ueber eincn cnnil)inierttMi l-'all Vdii Lynipliati
Myeloblastenleukamic. Zentralbl. f. uIIk. I'atb. 24:«')7. 191.?.

l-IXliMAX—LrMPHOIDOCVTIC LEUKEMIA

205

9. A case of lymphatic leukemia in which numerous myelocytes
appeared hefore death (Hertz). The bone marrow showed lymphocytes
in follicle arrangement and the spleen pulp showed myeloid prolif-
eration with atrophy of the follicles.

10. In the blood all transition forms from the "myeloblast" to the
lymphocyte were present (Chosrojeff). The diagnosis of "myeloblast"
is made on the strength of the oxydase reaction.

11. In a case of mixed leukemia (Herxheimer) the bone marrow
showed mostly lymphocytes with a small number of "myeloblasts."
The spleen, lymph nodes and mediastinal tumor showed hyperplasia
of lymphoid and myeloid cells. Here, too, the diagnosis of "myeloblast"
was based on the oxydase reaction.

Fig. 15. — Small blood vessel in the lymph node, showing
lymphocyte above and typical stem cell below. Objective Zeis
chromatic ; ocuar, Zeiss Xo. 8 compensating.

the lumen a
1.5 mm., apo-

All of the above evidence is more or less of an indirect nature.

Meyer, Heinecke, Ziegler, Naegeli and other dualists claim a
biologic antagonism between the follicles and interfoUicular tissues.
They claim that no one has ever observed myeloid tissue in the germinal
centers of the spleen or lymph nodes.

In 1912, Pappenheim '■■ made the statement that the unitarian theory
will receive great support, and Lydtin,"*" in 1913, declared that the

85. Pappenheim, A. : Bemerkungen zur Frage der akutcn Myeloblastenleu-
kamie und Leukosarcomatose. Wion. klin. Wchnschr. 25:163, 1912.

86. Lydtin, H. : Kin Fall von .Micromvtioblasten lenkiiniic. Folia haeniatol.
15:.316. 191.1

206 .IRCHirES Of IXTERXAL MEDICI. \E

dualistic theory will receive a serious blow at the moment that a case
would be found in which the so-called "myeloblasts" could be shown to
originate from lymphatic tissue, that is, from the follicles and follicular
cords. In other words, when in the blood the cell morphology will be
that of a "myeloblastic" leukemia, while the tissues will show lympho-
leukemic changes, i. e., follicular hypertrophy and presence of myelo-
blasts within the lymph follicles.

Hirschfeld ( 1908) says that so far researches have shown that
although the spleen and lymph nodes can produce granulocytes, yet they
come not from the specific follicular tissue but from the pulp and inter-
follicular tissue.

Naegeli (1910) makes the statement that under no circumstances
lias it ever been proven that the germinal center of the l_\mph nodes
can act as the site of origin of myeloid cells.

In 1915, Citron published his case of which he says, "It is the
only one to date which shows in a seemingly certain and irreproachable
manner, that the dualistic division does not always hold true ; that not
in all cases in which a myeloid metaplasia takes place is it necessary to
conclude that a substitution of the lymphatic tissue has taken place by
extra parenchymatous myeloid tissue; that in some cases a direct
autocellular change of lymphatic follicular lymphocytes into myeloid
cells may take place. Citron bases this statement on the following
findings.

1. The blood showed a monotonously uniform picture of the
so-called "myeloblasts" of Naegeli or "lymphoidocytes" of Pappenheim.

2. The bone marrow was normal.

3. In the lumina of the blood vessels of the spleen and l\mph
nodes the cells were e.xactly the same as those which jiermeated
these organs.

4. The splenic pulp and iiucrfollicular tissue of the lymjih nodes
consisted of cells similar in all respects to the cells of wiiich the
follicles were composed.

.S. No signs of follicular atrophy were present.

6. No evidence of myeloid metaplasia of the interfollicular tissue
was present.

7. The nuclei ni the eelN of the follicles and of the imerf,.Ilicular
tissue were not those of lymphocytes but of "myeloblasts."

8. These "myelobla.sts" could be seen entering tlu' circulation from
the hyperplastic lymph follicles.

iii:.M.\T<>r.nf;ic .\M) iiiSToi.ocic findincs
.;. ninod.— The difi'erential counts are shown in Table 4. The
various interesting findings of the blood ii.-ue already l)een referrecl to

riXEMAX-LVMPHOIDOCVTIC LECKEMIA 207

in the clinical discussion of the case. I wish again, however, to call
attention to the presence of numerous "stem cells" (lymphoidocyte of
Pappenheim; myeloblast of Naegeli). These cells were present in
the blood during the entire stay of the patient in the hospital. They
were extremely variable in size. Some were as large as a large lymph-
ocyte, others were smaller than a red blood cell. In the differential
counts we included all sizes under the headings "lymphoidocytes,"
"myeloblasts." "micromyeloblasts" or "stem cells." As seen in smears
stained in Wright's stain, the cells (Fig. 12. \) are morphologically
identical with those described by Pappenheim. Their cytoplasm is
basophilic, and scant in amount. The majoritv of cells did not contain

Fig. 16. — Portiiin of lymph node capsule. Stem cells in majority. Mitotic
figure in a small blood vessel. Cell 1, forming part of vessel wall, is an endo-
thelial cell. Cell 2, immediately below, is a connective tissue cell. Objective
1.5 mm., Zeiss apochromatic ; ocu'.ar No. 8, compensating, Zeiss.

azure granules. The structure of llie nuclei forms the diagnostic
feature of these cells. The chromatin forms a very fine, evenly
distributed sievelike meshwork, in contrast to the large clumped
chromatin blocks which characterize the nucleus of the lymphocyte.
Nucleoli, in variable numbers, are usually present. The staining of
the chromatin sievelike network of the stem cell is much lighter than
that of a Iym])hocyte. With the .slightest overstaining the typical
appearance of the stem cell becomes altered and differentiation from
the lympocyte is difficult.

The oxydase reaction was ]X)sitive in the iKilymorplionuclear leu-
kocytes but negative in all other cells. .\s already jjointed out, the

208 ARCHirES OF IXTERXAL MEDIC IS E

failure of the stem cells to show the oxydase granules is of no
significance one way or another. Citron and others refer to cases
unquestionably myeloid in character in which this reaction was negative.

Naegeli ( 1912 and 1919) wavers on the specificity of the oxydase
reaction. He says that the oxydase reaction is positive in normal myelo-
blasts and for the most part in pathologic myeloblasts.

Hyneke, Decastello, Dunn, Jochman and Bluhdorn declare that
the oxydase reaction need not be positive in unripe myeloid cells.

Boechat and Belz claim that the reaction may be absent.

Klein has shown that even lymphocytes may give a positive
oxydase reaction.

Menten has shown that lymphocytes and many tissues give a
positive reaction, so that the reaction is not only nut specific for
myeloid cells, but is not even specific for blood cells.

Morphologically, the cells in our case were identical with those the
dualists call "myeloblasts" and "micromyeloblasts." If the lilood alone
were examined, the diagnosis from the dualists' point of view would be
acute micromyeloblastic leukemia. I w'ill endeavor to show that these
"micromyeloblasts," probably the majority of them, certainly a great
many of them, were coming from the follicles and germinal centers
of the lymph nodes.

In the blood numerous transititm forms between these "micromyelo-
blasts" and lymphocytes were found. I have already pointed out that
the slightest overstaining alters the appearance of these cells. Transi-
tion forms, however, were found in well stained smears, side by side
with the tyjiical "micromyeloblasts." I believe that these transition
forms are another bit of evidence showing the relation of "micromye-
loblasts" to lymphopoietic tissues.

The number of the "micromyeloblasts" varied roughly with the
total white cell count. Table 4 shows these \ariations. Figure 4 shows
three curves expressing the rclatiun between the total white count, the
total "micromyeloblast count" and the relative percentage of these cells.
The rise in total number of these "micromyeloblasts" was invariably
associated with an exaggeration of all clinical symptoms, and an
enlargement of either the s])leen, lymph nodes or tonsils. It would
appear, therefore, lliat tin- increa-c of these cells in the circulation

was closely

associa

ted w

itli 1

lIlC ilK

rea

^ed ac

■tivity of

the 1;

nni

ihoi^oietic

tissues.

The nu

nierous

niitoi

ic f

igures

in

the c

irculatin,

^- 1,1, H

><] 1

t-d lis to

suspect the

possibil

ityof

mill

tiplica

tion

of th

ese cells'

in the

bio

od. Our

experiment;

;d work

, whil<

." no

t cone

lusi

ve, le;

ids me t(

. believe.

however.

that such ])

robably

was ■

not

the ca

se.

It is no

t m\- in

tentioi

1 to

pn.vi

■ or

.lis,.r

ove that

"myel

lobl;

;ists" and

"lymphobla

sts"' an

■ idem

ical

cells.

1

i>h lO sh

ow.' h

owi

'ver. that

IIXEMAS-LYMPHOIDOCYTIC LEUKEMIA 209

in this case, great numbers of cells were circulating in the blood, which
were morphologically identical with "micromyeloblasts," cells which
ordinarily are known to be derived from the bone marrow. We have
good evidence that a great many of our "micromyeloblasts" were being
generated in the lymphopoietic tissues and were passing out from these
tissues into the blood circulation.

Myelocytes were not found on some days, but there were as many
as 3 per cent, on other days. Xo promyelocytes were seen. Nucleated
red blood cells were j)rc-ent in small numbers. The appearance of

^:

a

7^v; H (^ €^.

KJK. 17. — Lymph sinus, showing Imtli I
1.5 mm.; Zeiss apochromatic ; ocular. X<

cytos and stem cells. Olijecti-
ouipensating; Zeiss,

the myelocytes and nucleated red blood cells may be considered as the
result of the irritation of the bone marrow due to the severe anemia.

B. Lymph Node. — The lymph node was obtained by biopsy and
was immediately fixed in Helly's fluid. The Dominici and methyl
green pyronin stains were used. The appearance of the node with
a magnification of eight times is represented in Figure 13. With low
power, the usual markings of a lymph node are almost obliterated. The
structure of the medulla is not as dense as that of the cortex. The
follicles of the cortex arc not clearly (liscernil)k-. with the exception

210 ARCHIVES OF IXTERXJL MEDICJXE

of an occasional one. In examining a hundred sections, only six
slides were found showing a single clear follicle containing a germinal
center. The remainder of the cortex shows a merging of the follicles
with the interfollicular tissue.

The peripheral sinus is filled with cells. The lymph sinuses sur-
rounding the trabeculae are practically all filled with dark staining
cells, and the same holds true for the plexus of sinuses throughout
the entire node. The medullary cords do not stand out prominently
but fuse more or less with the rest of the tissue. The blood vessels
contain red blood cells, which are rather indistinct with the low power

Fig. 18 (Upper). — Objective, 1.5 mm.; ocular, No. 8; maguificat'.on, X 8.
A. Large lymphocyte with seven nucleoli, in section. B. Medium sized lympho-
cyte, in section. C. Small lymphocyte, in section.

Fig. 19 (Lower). — Objective, 1.9 mm.; ocular. Xd. 10; X8 actual appear-
ance. A. Stem cell, very small, in section. B. Cell in niitusis in tlic lymiih
node. C. Small lymphocyte in section.

lens, and white blodd cells which show up very distinctly. The capsule
is \cry lutich thickened and iiihltralcd. In ])laccs it a])])cars to be
iiecnitic. The fatty tissue outside of the cai)sule is also itifiltrated
with many dark staining cells. With tliis low power the Dominici
and metiiyl green pyronin sections -hew a .yreat niaiiy cells which
stain paler than the rest.

With the Zeiss apochroniatic \.r nun. oil ininier--icin lens and the
Xo. S Zeiss compensating ocul;ir, it is seen that b-.-sides the usual

FJXEM.-iX—LVMPHOIDOCYTIC LEUKEMIA 211

normal cellular constituents, the entire lymph node is permeated by
atypical cells, quite variable in size. Some are larger than the largest
lymphocjtes to be seen in the node. Others arc very small, about the
size of a small lymphocyte. The cytoplasm of most of the cells forms a
thin ring around the nucleus (Fig. 12c). In a few cells, especially
those of the germ center, the cjtoplasm projects out in the form of
pseudopodia ( Fig. 14 Cells 3, 8, 12 and 15). These cells are free and in
all probability migratory in nature. The cytoplasm stains basic and is
darker than the cytoplasm of the lymphocytes, and has a homogeneous
appearance (Figs. 12d and 12c). In some cells the area immediately
around the nucleus stains considerably lighter than the peripheral
portion. No granules are demonstrable in the cytoplasm of the cells
in the sections.

The nuclei, here as well as in the blood cells, form the diagnostic
feature of the cell. They are usually eccentrically placed. Some
are round, the majority are ovoid, but many are very irregular in
shape, especially in the cortical regions, where proliferation seems
to be very profuse and the cells are crowded. The nuclear membrane
is very thin and hardly to be made out in some cells. The nucleus,
as a whole, stains much lighter than the cytoplasm, and can \ery
readily be differentiated from the majority of the nuclei of the
lymphocytes which stain considerably darker. Practically all of these
atypical cells contain one or more nucleoli. The chromatin arrange-
ment is entirely difTerent from that of the lymphocytes. \\'hereas the
chromatin of the lymphocytes, both large and small, is arranged in
blocks, taking a dark stain and usually arranged at the periphery of
the nucleus, the chromatin of these atypical cells forms a fine network
of tiny particles, which stain very lightly and seem to be linked by fine
threads running in every direction through the nucleus. In some
nuclei one may observe from one to four very small blocks of chromatin,
which stain very lightly and are irregularly placed (Fig. 12d).

The morphology of these cells, therefore, corresponds exactly with
the morphology of myeloblasts and micromyeloblasts as decribed by
Naegeli, or to the lymphoidocytes of Pappenheim.

Throughout the entire lymph node numerous cells in all stages of
mitosis are seen (Fig. 14, Cell 1). The blood vessels show these
"micromyeloblasts" in large numbers. They are easily distinguishable
from the ordinary lymphocytes. Figure 15 shows a small blood vessel
containing a lymphocyte, a "micromyeloblast," and red cells. Occasional
mitotic figures can be seen in the lumina of the blood ves.sels. Figure
16 shows such a figure in a blood vessel. ^

The lymph sinuses likewise are filled with enormous numbers of
these "micromyeloblasts." Figure 17 .shows several such cells in a
lymph sinus in the cfirte.x of the node. They are here also readily

212 ARCHII'ES OF IXTERXAL MEDICIXE

distinguishable from the large and small lymphocjtes. In this figure,
cells 6, 7, 8, 9 and 10 are "micromyeloblasts." They are identical with
those in the germ center. Cells 1, 2, 3, 4 and 5 are lymphocj'tes.

The cells of the reticulum proper can easily be differentiated from
the lymphocytes and the "micromyeloblasts." The cytoplasm of the
reticulum cells is large in amount and stains very lightly. Cytoplasmic
processes extend in all directions so that the cell assumes a stellate
appearance. The cell membrane is exceedingly fine. In the majority
of reticulum cells only part of the cytoplasm is visible, the remainder
fading out into the adjacent tissue or being concealed by overlying
cells. The nucleus is surrounded by a sharply defined membrane,
which in many cells is invaginated and may form long grooves over the
surface of the nucleus (Fig. 14, Cells 31, 32 and 34). The nucleus
is exceedingly poor in chromatin and appears to be very vesicular. A
nucleolus may or may not be present. Occasional small blocks of
chromatin may be present. In some cells the chromatin surrounds
the nucleolus, in others it is scattered about.

The capsule is thickened and is infiltrated with lymphocytes "'myelo-
blasts" and "micromyeloblasts." Figure 16 shows a small blood vessel
of the capsule containing a cell in mitosis. In this figure cell 1 is an
endothehal cell ; cell 2, a connective tissue cell, and the remaining cells
are "myeloblasts." A study of the various cells in the capsule shows
transition forms between our "myeloblastic" cells and lymphocytes as
well as connective tissue cells. These transition forms are of both large
and small cells.

In some places all the cells are flattened and elongated, while in
other places actual necrosis is present.

Mitotic figures are also present in the capsule. The "myeloblasts"
and "micromyeloblasts" are present in far greater numbers than the
lymphocytes. There are also many transition forms between our
atypical cells and the lymphocytes, which are recognizable by the fact
that in these cells, having all the cytoplasmic and nucleolar character-
istics of an atypical cell, some of the chromatin is arranged in the form
of dark staining blocks which varv in number in the different cells.
The majority of the.se chromatin blocks are usually to be found at the
periphery of the nucleus or abutting the nuclear membrane. Similar
transition forms were found in the germ centers of the follicles and
will be described in detail under that heading. In the capsule a whole
series of cells can be found between our atypical cells and lymphocytes.

We also have evidence indicating relationship between the con-
nective tissue cells of the capsule and our atypical cells. There are
typical elongated connective tissue cells, with the clear, pale, almost
homogeneous cytoj)lasmic processes, and vesicular, pale nuclei. The
next cell in the series shows a shortening of these processes. Then.

FINEMAX—LYMPHOIDOCYTIC LEUKEMIA 213

again, a cell may be completely rounded out or oval in shape. These
cells take on a more basic stain, which is usually a sign of cell activity,
and the nucleus shows more and more the characteristics of our
atypical cells.

In the medulla an occasional cord can clearly be made out. The
majority of cells in these cords are lymphocytes. However, scattered
everywhere among them are our atypical cells. Here, too, mitotic
figures are numerous. The rest of the medulla contains a majority of
our atypical cells. These are also present in large numbers in the
blood vessels and sinuses.

The cortex shows only a very few follicles with germinal centers.
In a hundred sections only six were found. Follicles without germinal

A— -'A

' V> •^<^.''/

Fig. 20. — Composite tracing showing diminution in size of spleen from
March 3 to 8. Roentgcn-ray treatment over spleen was given March 6.

centers are more numerous. No signs of follicular atrophy are present.
All evidence points toward increased activity. Mitotic figures are
numerous. Our atypical cells are present everywhere. Between the
follicles they constitute the majority of cells, but they are very numer-
ous throughout the follicle, and are present in large numbers in the
very centers of those follicles which do not show germ centers and
in the germ centers of the six follicles studied.

Transition forms between our atypical cells and lymphocytes, as
well as between lymphocytes and cells of the reticulum, are present
throughout the parenchyma of the node.

The germ center of a follicle, as represented in Figure 14, is typical
of all other germ centers found.

214 ARCHIVES OF IXTERXAL MEDICL\E

The "micromyeloblasts" are of about the same size as a small or
medium sized lymphoc}'te (Cell 4). All graduations in size up to the
size of a very large lymphocyte are present (Cells 5, 6, 7, 8, 9 and 10).
This variation in size is evident throughout the node. With methyl
green pyronin the cytoplasm stains an intense red. The red color is
more pronounced in these cells than in the cytoplasm of the lymphocytes.
In these India ink drawings the red staining cytoplasm is represented
by various shades of gray, arranged proportionately to the intensity
of the staining reaction. Compare cytoplasm of Cell 11, a lymphocyte,
with that of Cell 7, a medium sized "myeloblast." Light staining areas
in the cytoplasm (Cell 12), or around the nuclear membrane (Cells 5
and 9) are present. The shape of the cell varies considerably. It
may be round (Cell 4) or oval (Cells 2 and 7) or considerably
elongated (Cells 13 and 8). Some of the cells are more or less angular,
like Cells 9 and 14. Pseudopodia-like cytoplasmic processes are to be
seen on some cells, probably indicating their migratory nature (Cells
3, 8, 12 and 15). The c\-toplasm appears to be homogeneous and no
granules are distinguishable with either the methyl green pyronin or
Dominici stains. The cytoplasmic rim varies in size. In the majority
of cells it is rather narrow but well defined (Cells 3, 6 and 7). In
some cells it is considerably larger (Cells 8, 9, 10 and 12).

The nucleus may be placed eccentrically or centrally. In the majority
of cells it is eccentric. The nucleus is large and occupies the greater
portion of the cell. It varies greatly in shape. It may be round, oval,
or angular (Cells 6, 7 and 9). Bottleneck-like constrictions may occur
(Cell 16).

The nuclear membrane is well defined but thin. It may invaginate
the nucleus and form folds which may be mistaken in cross section for
nucleoli. Cell 10 shows three such invaginations, which could easily
be mistaken for nucleoli, since the nucleoli and the cytoplasm both stain
red. Careful focusing, however, shows these to be continuous with the
cytoplasm.

One or more nucleoli are usually present. These take the pyronin
dye and stain red. They may be located anywhere in the nucleus and
vary considerably in size. Compare nucleoli in Cells 2 and 9. Small
masses of chromatin may surround the nucleoli (Cell 12).

The arrangement and quantity of the chromatin is the chief diag-
nostic characteristic of the cell. The chromatin and parachromatin
are approximately equal in amount. The chromatin, taking a dark
greenish violet stain with methyl green pyronin, is in the form of
extremely fine dustlike particles, scattered rather uniformly throughout
the entire nucleus (Cells 2, 6, 9 and 12). In some cells coar.se chro-
matin masses are entirely absent (Cell 6), while in others, one or more
may be present (Cells 7, 10 and 13).

riXEMAX—LYMPHOIDOCYTlC LEUKEMIA 215

The ordinary lymphocytes are easily distinguishable from our
atypical cells. They vary greatly in size (Cells 17 and 18). Their
cytoplasm also takes the red stain, but not so intense a red as the cyto-
plasm of the atypical cells. In the majority of lymphocytes, the
c)loplasm forms only a very narrow rim about the nucleus, visible only
where the nuclear membrane is invaginated (Cell 19), or only on one
side of the cell (Cell 11), or not at all (Cell 18).

The nuclei of these lymphocytes are also characteristic. They stain
considerably darker than the nuclei of the atypical cells (Cell 18).
Numerous coarse chromatin strands run in all directions. Chromatin
blocks, triangular, square, oblong and irregular in shape, are present.
These, in many cells, have a tendency to arrange themselves around
the nuclear membrane (Cell 20). Nucleoli may be, but usually are
not present.

In this germ center (Fig. 14), as well as throughout the parenchyma,
many transition forms between lymphocytes and our atypical cells
may be seen. Beginning with a cell like Cell 6, which represents a
fairly small "'myeloblast," we may find a cell like Cell 22. In this
cell the nucleolus is surrounded by chromatin, and small masses of
chromatin line the nuclear membrane. Two other chromatin blocks
are to be seen below the nucleolus. The cytoplasm does not show
any changes.

The next in the series would be one like Cell 5. In this cell t!ie
chromatin blocks are more numerous and larger. A nucleolus is present.
The nucleus is indented in several places. The C)rtoplasm stains a
slightly lighter red.

The next stage is represented by Cell 23. The cytoplasm is definitely
less red than in Cell 5. The nucleus still retains the basic character
of a "myeloblast." It is lightly stained and fine dustlike particles of
chromatin are in evidence everywhere. Here, however, there are eight
triangular shaped, dark staining, large chromatin blocks arranged in
"radkern" fashion about the nuclear membrane. In addition four more
or less large irregular chromatin blocks, and several smaller blocks
occupy the central portion of the nucleus. At first sight such a cell has
all the earmarks of a lymphocyte. A detailed and careful study and a
comparison with typical lymphocytes and "myeloblasts" shows that it
is neither one nor the other but a cell which must be placed halfway
between the two.

The next cell in the series is Cell 24. In this cell the cytoplasmic
rim is considerably narrowed, and is practically invisible in part. The
nuclear membrane is thick and is lined all the way around with chro-
matin strands and blocks. Numerous coarse chromatin blocks and
strands are to be seen in the finer, chromatin groundwork described

216 ARCHIVES OF IXTERSAL MEDICINE

for the "myeloblast." This cell is past the half way mark in the
transition series.

Cell 25 is practically a normal lymphocyte. In comparing it with
Cell 18, which I consider to be a typical lymphocyte, the only diag-
nostic differential point is the character of the nucleus. The nucleus
of Cell 25 is practically a lymphocyte nucleus. However, it lacks the
coarse irregular chromatin strands so frequently found in lymphocytes,
and it still shows a rather fine distribution of small chromatin particles
all through the nucleus.

Other cells which I consider as transition forms are Cells 26, 27 and
28. And so thoroughout the parenchyma of the lymph node complete
series of transition forms are present. Whether the lymphocyte is the
mother cell of our atypical cell or vice versa, I do not know. Our
evidence is such as to show a relation between the two types of cells.

Another type of cell present (Fig. 14, Cells 31, 32, 33 and 34) is
the reticulum cell. This cell varies considerably in size and can very
easily be differentiated from the various types of lymphocytes, atypical
cells and transition forms. The cytoplasm of this cell takes an extremely
light pinkish stain. In most of the cells, as in Cells 31 and 32, no
definite cytoplasmic rim can be made out. Cell 34 shows the c}i:oplasm
and the cytoplasmic processes very well. The nuclei of these reticular
cells are very vesicular. The nuclear membrane is sharply outlined.
The nucleus may be almost triangular in shape, as in Cell 31, or oval
shaped, as in Cell 32, or elongated, as in Cell 33. The chromatin con-
tent is extremely meager, as in Cells 31, 32 and 33. It usually consists
of a few small blocks and a few strands. A rather large nucleolus,
staining red, and usually surrounded by some chromatin may be present
(Cells 31 and 32). Here, too, there is evidence showing a relation-
ship between our atypical cells and the reticulum cells, as well as a
relation between the reticulum cells and the lymphocytes.

A rather incomplete series may be represented by Cells 23, 28, 35
and 36, respectively. The nucleus of a small reticulum cell is shown
in Cell ZS. Cell 36 shows a beginning formation of chromatin blocks
and a darkening of the karyoplasm. Cell 28 shows a small amount of
well defined cytoplasm, the nucleolus is prominent and chromatin
arrangement and karyoplasm staining are suggestive of a cell between
a "myeloblast" and lymphocyte, as well as between a lymphocyte and
a reticulum cell. Cell 23 might very well fit in as the next in the
series. It has already been described as a transition cell between a
lymphocyte and a myeloblast, so that by including it also in our retic-
ulum lymphocyte series I believe to have demonstrated a relation
between the reticulum cells, lymphocytes and atypical cells. This rela-
tion of reticulum cells to the other two is not as definite as the rela-
tion between lymphocytes and the atypical cells. Not enough transition

FINEMAX—LYMPHOIDOCYTIC LEUKEMIA 217

Stages are shown in this one germ center. A study of other parts of
the parench}ma reveals much more conckisive evidence.

In comparing our atypical cells with the various cells found in
supposedly normal lymph nodes, human and animal, I did not find any
cells at all comparable to our "myeloblasts" and "micromyeloblasts."
Even germ centers cells, the so-called "lymphoblasts" were not com-
parable to our atj'pical cells. In these "lymphoblasts" the cytoplasm
takes on a lighter stain, and the chromatin occurs as fairly coarse dark
staining blocks. The fine dustlike particles of chromatin scattered
fairly uniformly throughout the entire nucleus are lacking.

DISCUSSION

It should be pointed out, that in comparing cells in blood smears
with cells in sections we must keep in mind certain morphologic dif-
ferences which are always present, due to the difference in technic. In
making a blood smear, the drop of blood is quickly spread out over the
slide and dried. The capillary traction flattens out the cells so that they
appear much larger than they actually are. In sections, however, the
various cells are exposed to the action of fixing fluids and consequently
undergo a certain amount of shrinkage, even with the best of fixing
fluids such as Helly's solution. Furthermore, an identical Wright's
stain technic cannot be used satisfactorily in sections and blood, so that
in comparing blood and tissue cells one must bear this also in mind.

In comparing a normal lymphocyte in the blood with a normal
lymphocyte in sections we notice a definite morphological difference.
The chromatin of the blood lymphocyte is arranged in a more or less
definite network, while in the lymphocytes of sections the chromatin is
arranged in coarse blocks and irregular strands. The lymphocytes in
the blood vessels of sections also show these coarse blocks and irregular
strands. The same holds true for myeloblasts in the blood and in
sections. Whereas, myeloblasts in the blood show a beautiful sievelike
nuclear chromatin network, in the tissues there is more of a tendency
to formation of fine particles with occasional chromatin block formation.
This same change is also evident in the myeloblasts in section blood
vessels.

The "micromyeloblasts" and "myeloblast" are very frequently mis-
taken for lymphocytes, because they overstain very easily, thus losing
their typical appearance.

The presence of the stem cell (microlymphoidocytes and lymphoido-
cytes of Pappenheim or micromyeloblasts and myeloblasts of Naegeli)
in large numbers in the blood of leukemics is usually associated with
an acute and rapidly fatal course of the disease.

A blood picture alone will often lead to a wrong diagnosis. In
some cases, for a correct diagnosis, histologic studies must supplement

218 ARCHIl-ES OF IXTERXAL MEDICINE

the blood studies. Krjukoff diagnosed myeloid leukemia from the
blood, yet the bone marrow showed only slight activity and he was
forced to conclude that his myeloid cells were coming from the "mye-
loid parts" of lymph nodes and spleen.

Similarly, Citron diagnosed "micromyeloblastic" leukemia, yet he
found the bone marrow normal and lymph follicles hyperplastic and
proliferating these "myeloid" cells.

The dualists, however, deny the possibility of "myeloid" cells
originating from lymphatic tissue. They argue that in lymphatic leu-
kemia the follicles of the lymph nodes atrophy or are, at least, quies-
cent. They deny absolutely the possibility of "myeloid" cells ever
occuring in follicles and especially in the germ centers of follicles.

Citron's case, however, showed the "myeloblasts" of the dualists in
the follicles of the lymph nodes and spleen.

Naegeli, referring to Citron's case, belittles the findings because
"the patient died as a result of overdosage of benzol." Only one dose
of 6 gm. was given by rectum. Boni *' gave 5 gm. daily for three
weeks. Josef son '^^ gave 96 gm. in six weeks. Krokiewicz **" gave
in one case a total of 206 gm. No ill results were observed in
these cases.

Naegeli also belittles the tindings in cases which had excessive
roentgen ray therapy.

The patient at the time the axillary lymph node, on which this study
is based, was removed had not had any benzol, nor had there been
any direct roentgen-ray irradiation of the axillary lymph nodes.

W'e believe our case to be even more convincing than Citron's for
two reasons : The lymph node was obtained in vivo, so that post-
mortem changes can be ruled out. Whereas, Citron speaks of "myelo-
blasts" and "microniyeloblasts" in the follicles he does not say that he
found them in the germ centers. The germ centers in our cases con-
tained many of these cells.

This case is presented in the belief that it offers e\ idence in favor of
the unitarian theory of Pappenheim and Fcrrata. The cell which has
been described as a "myeloblast" and which the dualists have assumed
to be a specific myeloid cell was found to originate in the germ centers
and follicles, as well as in the medullary portion of the node. This case
and the case reported by Citron prove, therefore, that the cell in ques-
tion may be related to lymphocytes as well as to cells of the myeloid
series.

87. Boni : Siir une cas tie leukemie chronique niveloide traitee par le benzol.
Bull, delle clin. (Aug.) 191.3; a])str. in Folia haematol. 16:167. 1914 (.^ubertin).

88. Joscfson, A.: Bcnzolbeliandlung vid Leukami. Hvgica. 1914 (abstr.
Folia haematol. 16:167, 1914).

89. Krokiewic;;, .'\. : Die Benzol Behandlung dcr I-cukaniic, Pzegl. lekarski
52:. =582. 191.3 (abstr. in Folia haematol. 16: 169. 1914)

FJXEMAX-LVMPHOIDOCVTIC LEUKEMIA 219

SUMMARY

Morphologic Part 2

1. The blood, at all times, showed numerous stem cells (lymphoido-
c>'ted of Pappenheim), (myeloblasts of Naegeli) of all sizes. Our
cells, which we shall refer to as atypical cells, had a basophilic cyto-
plasm and a nucleus in which the chromatin formed a very fine evenly
distributed sievelike network. Morphologically our atypical cells were
indistinguishable from typical myeloblasts.

2. Very careful staining was essential in bringing out the finer
details of these atypical cells.

3. The oxydase reaction was negative in these cells in the blood
smears.

4. The diagnosis from the blood alone would be "micromyelo-
blastic" leukemia.

5. The presence of numerous mitotic figures in the blood stream
suggested the possibility of cell proliferation in the blood stream. We
could not demonstrate such to be the case.

6. Lymphocytes, normal in appearance were always present in the
blood. The contrast between the lymphocytes and the atypical cells
was very marked. Numerous transition forms between the lympho-
cytes and the atypical cells were present in the blood.

7. Some myelocytes and nucleated reds were present. The severe
anemia might easily account for these.

8. The biopsy of a lymph node showed these atypical cells pro-
liferating in great numbers in the capsule, interfoUicular tissue, lymph
cords, lymph follicles and in the germ centers of the lymph follicles.

9. Transition forms between the connective tissue cells of the
capsule and these atypical cells as well as between lymphocytes and
these atypical cells were present in the capsule.

10. In the interfoUicular tissue as well as in the follicles and even
in the germinal centers transition forms between these atypical cells
and reticulum and lymphocytes were also present.

11. The lymph follicles and lymph cords showed no signs of atrophy,
but had all the earmarks of marked activity. Mitotic figures were
numerous. The only signs of atrophy or necrosis were found in the
capsule.

12. The.'ie atypical cells formed the majority of the cells of the
parenchyma.

13. These atypical cells constituted by far the majority of the cells
in the lymph sinuses and were very numerous in the blood vessels of
the node.

220 ARCHIVES OF IXTERXAL MEDICINE

14. From the evidence at hand, the conclusion is justified that in all
probability the majority of the "myeloblasts" and "micromyeloblasts"
of the blood were coming from the lymphoid organs, not only from
the portions which, according to the dualists, may give rise to myeloid
cells, but from the sanctum sanctorum of the lymphoid tissues, namely,
the follicles and germ centers.

I wish to express my sincerest thanks to Professors L. G. Rowntree and
S. Marx White for their kind guidance in the preparation of the clinical report,
and to Professor Hal Downey who was the first to recognize the significance
of the case and under whose direction the morphologic study was carried out.

INTRACUTANEOUS REACTIONS IN LOBAR
PNEUMONIA *

GEORGE H. BIGELOW. M.D.

BOSTON

In 1915 Clough ^ reported intracutaneous reactions in lobar
pneumonia patients. He prepared his antigen as follows: Cultures
used were made from the lungs in cases of fatal pneumonia, with no
statement as to t}T3e. From twenty-four to thirty-six hour old cul-
tures in 5 per cent, glucose broth plus calcium carbonate were decanted
off the carbonate, centrifuged, washed with salt, recentrifuged, the
sediment taken up in a few drops of distilled water, dried, weighed,
ground for three hours with sterile sand, e.>ctracted with 10 c. c. saline
for each gram of dried material, incubated for eighteen hours; centri-
fuged at high speed for several hours until no more sediment came
down ; used in this form or precipitated with absolute alcohol. One
gram of the dried culture residue yielded about 0.15 gm. dried
alcoholic precipitate.

He used two methods of inoculation : first, allowing a drop of
1 per cent, solution of his antigen to dry on a scarification as for a
von Pirquet tuberculin test; second, injecting a 0.25 per cent, solution
intracutaneously, causing a painful inflammatory reaction subsid-
ing after twenty-four hours. There was no appreciable increase
in severity or time of the reaction in the pneumonias over the controls.
The more dilute solutions gave discrete papules from 0.5 to 2 cm.
in diameter with an ill-defined area of hyperemia. The actual measure-
ments of the papules averaged a little larger in the pneumonias, but
this was only slight and inconstant. He found no difference in
reactivity early, at crisis, or later in convalescence. As to dough's
work, then, it may be said that he used a rather elaborately prepared
antigen with practically negative results.

Weil = criticizes Clough's antigen for being too strong so that it
produced reactions independent of any immunologic response. He used
forty-eight hour old cultures on Loeffler's serum medium suspended in
distilled water, from 2 to 3 c. c. to a tube, shaken, incubated two
hours, and then heated to 60 C. for one hour. In a few early experi-

*From the Department of Preventive Medicine and Hygiene, Harvard
Medical School.

♦This article is one of a series of studies carried out under the supervision
of the Influenza Commission of the Metropolitan Life Insurance Company. The
Influenza Commission is investigating the etiology and prevention of the acute
respiratory diseases, and the work is being carried on in Boston, New York,
Chicago and Washington. The members of the Commission are Dr. M. J.
Rosenau. Chairman, Dr. G. W. McCoy, Dr. L. K. Frankel. Dr. A. S. Knight,
Dr. E. O. Jordan, Dr. \V. H. Frost and Dr. W. H. Park. .Secretary.

1. Clough. P. W.: Johns Hopkins Hosp. Bull. 24:295. 1913.

2. Weil. R.: J. Exper. M. 23:11. 1916.

222 ARCHIfES OF IXTERXAL MEDICI. \E

ments a Type III organism was used "with success," otherwise Type I
was used exclusively. For the test, a sufficient amount of antigen was
injected to produce a small wheal, that amount being between 0.1 and
0.2 c. c. All showed an immediate superficial ill-defined blush which
was considered merely traumatic. In the negatives this faded within
a few hours and nothing more developed. In the positives, within
twenty hours a fairly well circumscribed area of erythema had
developed with slight infiltration and elevation of the skin surrounding
the point of puncture. This may be a true papule, persisting for
forty-eight hours or even longer, ^\'eil obtained no reaction during
the course of the disease, but "after the subsidence ... a consid-
erable percentage of cases do present reactions." These exceptionally
appeared twenty-four hours after crises, but more commonly later
even after two, three or more weeks. The controls "may or may not
present a reaction dependent presumably on their previous sensitization
by the pneumococcus or an allied organism." A few cases of pneumonia
gave no reaction at any time. In summary, Weil obtained positive
reaction in a "considerable percentage of cases" after crisis, and in
some controls. The reaction which he considered positive readied
its maximum within twenty hours, though it might persist for two
days or more. He used a relatively simple incubated suspension of
pneumococci in distilled water.

Steinfield and Kolmer ^ jjrepared an antigen in nnich tlie same
way, except that they suspended in saline instead of distilled water
and omitted the incubation. .Separate preparations of Types I, II
and III were made. The final dilution was such that 1 c. c. contained
2 billion cocci, that is 200 million were injected for a test. They
found that at the end of twenty-four hours all tests .showed a zone
of hyperemia. By forty-eight hours a definite jiapule liad appeared
in the positives, with an erythema greater than 1 cm. in diameter and
accompanied by slight edema. This reaction persisted for four to
five days, and gradually cleared. There was no instance of a pustule
formation. In the negatives the erythema had largely cleared by
forty-eight hours. Of nineteen cases of clinical loljar jineumonia. six
gave positive reactions with one or more antigens. Rut in none did
the type or types found in the sputum exactly correspond with those
shown by the skin reaction. For instance, one case showed Type I
in the sputum and reacted to Types I and III; another case showed
Types I and II in the sputum and reacted to Types II and III, etc.
The earliest reaction appeared on the ninth day of the disease, or three
days after crisis, and the latest on the thirty-ninth day after onset.
"A large number of tests with healthy persons and patients suflfering
from chronic ailments not involving the chest" were negative. The

,1 Steinfield. F... and K,.lnu-r. J. A.: J. Infect. Vh. 20:.i.44. W?.

BIGELOir—I.XTRACLTAXEOUS REACTIOXS 223

authors conclude that there is no constant relation between the reaction
in the skin and the type of organism found in the sputum, and think
it probable that the allergic reactions to pneumococcus protein are of
a more general character than the agglutination reactions.

In 1918 Weiss and Kolmer * report an intracutaneous test with a
preparation of "pneumotoxin" as opposed to the "pneumoprotein" of
the former tests. Perhaps the most important difference in their
antigen over those used previously is that at no time is the product
heated to a temperature which would endanger the thermolabile
substances. They centrifuged eighteen hour broth cultures of
Type I pneumococcus, washed once with saline, took up in 5 c. c.
saline and dissolved the cocci in 1 c. c. of a 2 per cent solution of
sodium choleate. The total volume was made up to 30 or 40 c. c,
centrifuged to remove undissolved pneumococci and 0.2 per cent,
tricresol added. The "control fluid" was treated as above, except for
the inoculation with pneumococci. The minimal lethal dose for a 250 or
300 gm. guinea-pig was found, and the antigen diluted so that 0.1 c. c.
contains one twentieth the minimum lethal dose (M. L. D.). This
amount was injected intracutaneously for the lest. The preparation
deteriorated rapidly and was made fresh every day. The authors
found the reaction in "all respects similar to that described for the
Schick test with diphtheria toxin." Cases just prior to the crisis
showed a strongly positive reaction marked by vesiculation. Others
showed a definitely circumscribed area of edema and erythema gradually
fading, leaving a scaling zone of brownish pigmentation. The "control
fluid" and heated "pneumotoxin" remained negative. Pseudoreac-
tions involving the entire arm with diffuse erythema and a definite area
of edema never persisted longer than forty-eight hours. Of thirty-
eight cases of "acute lobar pneumonia" all reacted positively; of
sixteen cases of "lobar pneumonia convalescents" five reacted posi-
tively; of five "doubtful" cases ("alcoholism and tuberculosis with
possible superimposed pneumonia"), one gave a positive reaction; and
of twenty-three controls all were negative.

In summary, they state that the tests with "pneumotoxin" were
elicited as early as the fifth day (the earliest under observation) and
as late as the thirteenth day of the disease. In general, the test was
positive throughout the toxemia.

Further in the paper the authors took up the question of the
type specificity of this reaction. The sputum from ten cases which
reacted positively to their type I "penumotoxin" was examined for type
organism. Six of these showed Type IV and four showed Type I. They
consider that "sensitization to the toxin presumably takes place with its
liberation (by action of normal enzymes on pneumococci normally

4. Weiss. C, and Kolmer. J. .\. : J. Imnuinol. 3:39.^. 1918.

224 ARCHIVES OF INTERNAL MEDICINE

localized in the lung alveoli) at the time. of the prolonged chilling due
to exposure." They feel that the method is not as yet of value in
serologic type diagnosis.

In these two papers Kolmer shows that his heated saline suspension
of pneumococci gives positive reactions after crisis which are not
specific for type, and that his unheated bile salt solution of pneumococci
gives positive reactions before as well as aftei crisis which are also
not specific for type. It is rather a surprise after having compared
this latter reaction to the Schick test in diphtheria, to find that the
presence rather than the absence of a reaction is considered specific.
The authors' theory of early sensitization against their "pneumotoxin"
is certainly open to question when we consider the absence of any
sensitization necessary in the case of the positive Schick reaction, and
also the lack of specificity in the toxic radicle as shown by \'aughan's
work.'^

EXPERIMENTAL WORK

It was hoped that by the study of a considerable series of pneumonia
cases and normal and diseased controls, using antigens made from the
various types of pneumococci (homologous and heterologous) pre-
pared in various ways, additional information might be obtained
concerning the role played by allergy in pneumonia with especial
reference to crisis. Furthermore, it was hoped that by varying the
method of preparation and dose of antigens prepared from the various
types of pneumococci, it might be possible to obtain antigens sufficiently
delicate to indicate the type of infection present. If this reaction
should prove to be positive in the early stages of the disease, it might
furnish an earlier, quicker and simpler means of type determination
than can lie obtained by the present method of mouse inoculation, thus
making it possible to institute specific serum therapy in suitable cases
with less loss of time than is at present possible.

Cultures and Their Sources. — The cultures " used in the preparation
of antigens were:

5. Vaughaii. V. C, and Wheeler. S. M. : J. Infect. Di.<;. 4:476, 1907.

6. The cultures used were obtained as follows: I A from Dr. F. T. Lord.
IB, II A, II B. Ill A and III B from Dr. G. B. White. Slate Biological Labora-
tory, Forest Hills. IV B and IV C from Miss E. A. Beckler, Bacteriological
Laboratory, State House, Boston, isolated Dec. 13. 1920. The remaining cuhures
were isolated in the Pneumonia Lal)oratory of the Department of Preventive
Medicine and Hygiene. The following were obtained from blood cultures
taken on patients on the Special Pneumonia Service at the Boston City
Hospital on the dates as indicated: I C, Jan. IS, 1921; ID, January 24; IF,
Feburary 24; I G, February 18; I H, Feburary 28; I J, March IS; III C, Feb-
ruary IS; HID. February 27, and IV E, January 24. The following cultures
were isolated from the sputum of similar patients : I R, February 8, and 1 1,
March 22. The following cultures were isolated from the sputum of patients
outside Boston as indicated: IV A, Foxboro Slate Hospital, and IV D,
Framingham.

BIGELOU'—LXTRACUTAXEOUS REACTIONS 225

Type 1 : 10 strains, designated I A, I B, etc.

Type II : 2 strains, designated II A and II B.

Type III : 4 strains, designated III A, III B, etc.

Type I\' : 5 strains, designated IV A, IV B, etc.

Methods of Preparing Antigens. — The cultures used were from
eighteen to twenty-four hour old growths in 100 c. c. of a 1 per cent,
glucose beef infusion broth, neutral to phenolthalein, which was found
to correspond to a hydrogen ion concentration of from 8.0 to 8.2. On
this medium there was homogeneous clouding of the broth with an
abundant sediment in the bottom of the flask. The morphology of the
organisms used was typically that of pneumococcus, gram-positive,
bile soluble, and the type was repeatedly verified by agglutination
with type serum from the New York State Board of Health and the
Rockefeller Institute. All growths were examined microscopically
for contamination, and serologically to confirm the previous type
determinations.

For all the antigens, the first steps were identical. The growth
was centrifuged, and the sediment was washed twice with sterile salt
solution, centrifuging after each washing. Throughout the work the
centrifuge was run at about 3,000 revolutions per minute. The lime
interval for centrifuging and shaking electrically was one-half hour.

The finished antigen was sealed in ampules containing sterile glass
beads so that if desired the suspension could be shaken thoroughly
before use to insure uniformity of material. W^henever the antigens
were heated to 56 C. special care was taken that the entire ampule
was completely submerged in the water bath. Before using the antigens
were plated on whole rabbits' blood agar to test sterility, and in the
simpler preparations bacterial counts were made according to Wright's
method ' as commonly used in standardizing bacterial vaccines. The
ampules were kept on ice until used. No antigen was used more than
a week after being put on ice. Both the opalescent supernatant fluid
and also the entire opaque suspension of organisms were used. Before
using the antigens were diluted with sterile saline solution and doses
varying from 10 to 1,000 million bacteria or their equivalent were
injected in 0.1 c. c. volume.

With the earlier antigens the strains (except Type IV) were kept
separate. Thus a patient with a known Type I pneumonia might be
inoculated with four strains of Type I, one strain each of Types II
and III, and a polyvalent Type IV, the whole constituting one test. No
especial merit was found in this method, the few homologous strains
used giving no better results than the heterologous strains, and since
it subjected the patients to multiple inoculations, the strains of each

7. Wright. A. E.: Lancet 2:1556, 1900; 2:11, 1902.

226 ARCHIVES OF IXTERXAL MEDICIXE

type were pooled. The Type IV antigen was later omitted since in
specific reactions the two nonspecific types furnished ample controls.

Antigen 1.— Cultures I A and I B, II A and II B, III A and
IV A, IV B and IV C were used. The sediment after the second
washing was taken up in from 20 to 30 c. c. of saline solution and
heated to 56 C. for one hour. The ampules were then vigorously
shaken, and put on ice.

Antigen 2. — Prepared as Antigen 1, except that it was incubated
one week before being used. At the end of that time the sediment
formed on standing showed on smear much amorphous gram-negative
material including faint staining cocci with individual gram-positive
cocci and rare diplococci.

Antigens 1 C and 1 D. — These were prepared as Antigens 1 and 2
from cultures I C and I D. They were used autogenously and on a
few other patients.

Antigen 3. — Prepared as Antigen 1 with the following cultures:
I A and I B ; II A and II B ; III A and III B ; and IV A and lA' C.

Antigen 4. — Same as Antigen 3, except that it was incubated one
week.

Antigen 5. — Same as Antigen 4, except thai it was incubated two
weeks before using.

Antigen 6. — After the second washing with saline solution, the
sediment was taken up with distilled water instead of saline solution.
Cultures used: I A, I B and I C; II A; III B; IV A, lY C. IV D
and R" E.

Antigen 7. — Same as Antigen 6, except that it was incubated one
week.

Antigen 8. — After the second washing with saline solution, the
sediment was taken up in 10 c. c. of a 10 per cent, bile solution,* and
incubated for one hour. Attempts were made to obtain suitable animal
membrane for separation of the bile .salts from the dissolved pneumo-
coccus substances by dialysis, but it was found more feasible to obtain
the separation by precipitation of the protein with four volumes of
absolute alcohol. This was thoroughly .shaken, centrifuged and the
sediment again mixed with absolute alcohol. After recentrifuging,
the sediment was dried overnight in partial vacuum over sulphuric
acid. The residue was weighed, ground with sufficient sodium
chlorid to make the final solution physiologic and taken up by adding
distilled water drop by drop while grinding until 0.1 c. c. contained
1/1.^0 mg dried .sediment." ( Later injections up to 1 10 mg. were used.)

8. Bacto-oxgall was used.

9. Gay. F. P. and Minakcr, .^. J.: J. A. M. .'\. 70:21.S (Ja

BIGELOir—fXTRACUTAXEOUS REACTIOXS 227

This was heated to 56 C. for one hour. For this antigen all the strains
of each type were pooled. The cultures used were : I A, I B, I C, I D
and I E; II A, and II B; III A, III B and III C.

Antigen 9. — After the second washing with saline solution, the
'sediment was taken up in a small amount of saline solution and pre-
cipitated with four volumes of absolute alcohol. This was centrifuged
and taken up a second time in absolute alcohol. After again centrifug-
ing the sediment was extracted by thoroughly shaking with ether.
The sediment was dried for one hour in the incubator, weighed, and
ground with sufficient salt to make the final solution physiological,
and taken up in distilled water, added slowly while grinding. The
dilutions were such that the antigen contained 2U/150 mg. per 0.1 c. c.
This was at first diluted twenty times before using but was eventually
used full strength. The antigen was heated to 56 C. for one hour.
In this preparation the strains of each type were pooled. The cultures
used were: I A, I B, I C, I D, I F, I G, and I li; II A and II B;
III A, III B, III C, and HI D.

Antigen 10. — The most important difference between this and the
foregoing antigens is that it was at no time raised above 37.5 C. The
strains of each type used were pooled and were as follows : I B, I C,
I D, I G, I H, I I and I J ; ir B ; III C and III D. After the second
washing with., salt solution the sediment was divided into known
amounts and kept on ice. The day of the tests, sufficient of sterile
solution of 2 per cent, bile plus 0.25 per cent, tricresol was added to
make 0.1 c. c. contain the equivalent of one twentieth of the growth
of pneumococci on 60 c. c. of broth.'" An incubated solution of 2
per cent bile plus tricresol was used as control.

In summary, the first seven antigens were simple suspensions of
pneumococci in saline solution or distilled water. Antigens 1, 3 and 6
were not autolyzed in the incubator, while antigens 2, 4, 5 and 7 were
incubated one week or more. In antigens 8 and 9 the pneumococci
were treated with bile and alcohol, and alcohol and ether, respectively.
Thus the pneumococcus protein was subjected to considerable violence.
Antigen 10 diflfered from all the others in that it was never raised
above 37.5 C, while the others were all held at 56 C. for one hour.
In the bile solution whatever endotoxins the pneumococci contain
should be liberated.

Tests. — A "test" consisted of multiple inoculations. In the first
seven antigens each strain of the different types was made up separately,
with the exception of Type IV strains which were pooled and served

10. Cole, R.: J. Exper. M. 16:644, 1912; 20:346, 1914. Avery, O. T.. Chicker-
ing. H. T., Cole, R. and Dochez, A. R. : Acute Lobar Pneumonia, Prevention and
Serum Treatment. Monograph No. 7, Rockefeller Institute for M. Res., 1917.
Weiss, C: J. M. Research 34:103, 1918.

228 ARCHIVES OF IXTERXAL MEDICINE

as a control for the reactions which might be specific for type. Thus,
with these antigens, a Type I pneumonia might receive from one to
four different strains of Type I antigen, besides one Type II, one
Type III and the pooled Type IV antigen. As it was very exceptional
to obtain a reaction with one strain and not with all of the same type
used (this occurred only in Case 7, which on the thirty-eighth day of
the disease responded positively to the homologous strain and negatively
to the heterologous strains of Type I), this method was abandoned
later, since it subjected the patient to the discomfort of multiple
inoculations.

Antigens 8, 9 and 10 were made from pooled strains of each of the
fixed types. Type IV was no longer used, since in specific type
reaction the two nonspecific type antigens acted as ample controls.
Thus with these antigens a "test" consisted of three inoculations,
one for each of the three fixed types.

With each new antigen various dilutions of each strain or pooled
type were injected in an effort to obtain the titre of the antigen for
both the specific type reactions and also for those reactions which did
not show type specificity. These titres were, unfortunately, found to
be very close to each other, and varied within considerable limits with
each preparation and individual.

As stated earlier, tests were performed with both the opalescent
supernatant fluid and the whole thoroughly mixed opaque antigen. In
general, the supernatant fluid gave fewer reactions that were nonspecific
for type, but it was also feared that in the same dilutions it might
also elicit specific type reactions with less constancy.

The tests were made on the flexor surface of the forearms. A 26
gage needle was used and 0.1 c. c. was injected intracutaneously. When
properly done a bleb appeared which persisted for some minutes;
there was pitting at the hair follicles, and no bleeding at the point
of inoculation.

Readings. — -The tests were read, as a routine, after two, eighteen,
twenty-four, thirty, forty-eight and seventy-two hours. Additional
readings were made on the positive reactions.

Reactions Appearing zi'ith Only One Type of Pncuuiococcits. — The
reactions in this group were at their height in from twenty-eight to
thirty-two hours. These showed a deep red, indurated, papular center
measuring from 2 to 4 cm. in diameter. Around this there was a
somewhat lighter areola measuring up to 8 cm. in diameter. In forty-
eight hours the papule still persisted, usually duller in color, the areola
had entirely faded and in about half the cases was replaced by a zone
of pigmentation. The papule in some cases persisted to the sixth day
after the test. In one case there was very fine scaling. In .some
cases this type of reaction showed nothing after eighteen and twenty-

BIGELOIV—IXTRACUTAXEOUS REACTIOXS 229

four hours. In others there was the reaction described below which
was fading in twenty- four hours. The reaction to but one type of
the antigen used was, in general, specifically positive for the type of
infection, and will be referred to hereafter as specific type reaction.

Reactions mith Tivo or More Types. — These reactions were at their
height in eighteen hours. They showed no distinct differentiation
between central papule and surrounding areola, and were usually
indefinite in outline, fading gradually into the surrounding normal
skin. The erythema was often mottled and of large dimensions up to
10 or 12 cm. in the longest axis which was always parallel to the long
axis of the arm. On fading, unless they later showed specific type
reaction, there was no pigmentation or scaling. By twenty-eight hours
there remained only a small area about 0.5 cm. across of local redness
around the point of inoculation. This reaction was always elicited
with more than one of the types of the antigen used and generally
with all when all types were used in the same strength. In an individual
showing this reaction the number of types in which it appeared seemed,
in general, to depend on the strength in which the type antigen was used
and in no case on its specificity for the type of infection present. As
will be discussed later these reactions were considered as a response
to some substance possessed in common by all types of pneumococci
or to some property of the antigen not even specific for the entire
pneumococcus group. This reaction which was elicited in common
by all types of pneumococci will be referred to hereafter as the common
reaction.

In an effort to increase the number of specific type reactions, the
strength of the antigens was varied with the object of reducing the
common reactions to a minimum and yet retain sufficient strength to
elicit the reaction to a single specific type.

Special reference must be made to the reactions obtained with
Antigen 10 in which the thermoiabile toxic substances from the pneumo-
coccus were not destroyed. When used in the strength recommended
by Kolmer* and in weaker dilutions, severe local reactions were
elicited, showing papules becoming pustular with wide zones of erythema
and induration up to 12 or 14 cm. in their longest dimension. This
was at its height in about thirty hours when the erythema faded, leaving
no pigmentation or scaling in any case. The pustule persisted for
weeks. This reaction was well marked in both controls and pneumonia
cases, the most severe being in a normal control. The control injection
of bile and tricresol showed only a small local papule, but none of
the violent erythematous and edematous reaction.

Results. — During the course of the work, 124 persons were tested,
as follows : ( 1 ) Lobar pneumonia, 104 cases ; forty-seven had a Type I
infection (specific antipneumococcus serum being used in twenty of

230 ARCHirES OF IXTERXAL MEDICIXE

these and not used in twenty-seven); eight had Type II infections;
seven had Type III pneumonia, one had Types II and III, twenty-nine
had Type IV, and in twelve the type was not determined (no atypical
Type II cases presented themselves) ; (2) controls, twenty persons,
of whom sixteen were patients showing for the most part acute febrile
conditions other than lobar pneumonia, and four were apparently
normal. On these 124 persons, 223 tests were performed, divided about
equally between the ten antigens. The earliest test was performed on
the third day of the disease and the latest on the eighty-first day.
The number of tests on each individual varied from one to five.

Of the 104 cases of lobar pneumonia tested, eleven (10.5 per cent.)
gave specific type reactions, forty-six (42.3 per cent.) gave a common
reaction and fifty-two (50 per cent.) gave no such reaction. Of the
twenty controls, none gave a specific type reaction, nine gave a common
reaction and eleven gave no reaction.

The eleven cases which gave specific type reactions were distributed
among the types of infection as follows:

T.\BLE 1. — C.vsES Ma.vifesti.vg Specific Type Reactions

Type Number Number Showing

of of Cases Specific Per Cent.

Pneumonia Tested Type Reaction Specific

This reaction was specific for the type of infection in the patient
as determined in the blood or sputum, or both, except with Antigen 9.
With this antigen, reactions were elicited in three patients, in each
case with the Type II preparation; while with Types I and III prep-
arations the reactions were consistently negative. Of these patients
one showed a Type II pneumococcus in the sputum, one a Type I
and one a Type IV. In the last two cases blood cultures were persis-
tently negative and repeated examinations of the sputum failed to
show a Type II organism. In the Type I patient, the test was repeated
with the same antigen five days later and was negative. In the cases
of Type I and IV pneumonias, the reaction with Tyi>e II antigen was
obviously nonspecific as to type of pneumocfKCUs. Although this
Type II preparation of Antigen 9 gave frequent negative tests in other
cases, it is not unlikely that its apparent specific reaction in one case
may be accounted for on the law of chance. Thus all three reactions
might well be considered as nonspecific for type, and as a response
to some form of protein common to all types of pneumococci, though
in that case the Type I and III preparations should also have shown
positive reactions, or to some other factor in the antigen rather than

BIGELOW—IXTRACUTAXEOUS REACTIOXS 231

to specific type sensitization in the patient. None of these three patients
gave a specific reaction with any of the other antigens used.

Deducting these three reactions from Table 1, we may reconstruct
it as follows :

Table 2

Type Xumber Number Showing

of of Cases Specific Per Cent.

Pneumonia Tested Type Reaction Specific

Although the number is small, it may be said that in these tests,
no one of the fixed types of pneumonia shows any marked prepon-
derance over the others as regards the ease with which intracutaneous
reactions, specific for the type of infection, could be elicited.

The relation between the specific type reactions in Type I ca.ses
and treatment with specific antipneumococcus serum is shown in
Table 3.

T.-XBLE 3. — Specific Type Reactions and Specific Serum Treat.mext

Total of Type :
Total of Type I

Treated
With serum

[ pneumonias tested

pneumonias with specific r
Total
Tested

47

Specific
2

Per Cent.
Specific

Thus, instead of finding more ease in eliciting this reaction fol-
lowing specific serum therapy, as might be expected from the report ^'
of the earlier appearance of antibodies in such ca.ses, a somewhat
smaller proportion of these reactions appeared among cases so treated
than among those receiving no serum.

Of the eleven cases showing specific type reactions, seven gave one
such test with one or more negatives; three gave two with one or.niore
negatives, and one gave three with one negative test. There was a total
of sixteen specific type reactions out of thirty-seven performed on these
eleven patients. The earliest specific test was performed on the sixth
day of the disease, or the day of the termination of crisis; the
latest was on the thirty-eighth day, or twenty-nine days after crisis.
The only reaction of this kind obtained before crisis was on the eighth
day of the disease, or three days before crisis. This, however, was
not specific for the type of organism found in the patient's sputum.
Dividing the disease into weekly periods, the s])ecific type reaction
occurs as shown in Table 4.

J. H.. Moss. \V. I., aii.l Bnnvn G. I..: .1. I-xpcr. M. 12:562. 1910.

ARCHIVES OF IXTERXAL MEDICINE
Table 4

Week
Of the disease

1st

2d
6

1

3d

3
4

4th
2

5th

6th

One reaction occurred before and one during crisis. Thus the
largest number of these reactions appeared during the second week
of the disease or the first week after crisis. With the crisis a
phenomenon of allergy', this is what would be expected since the seven
to fourteen day interval is that commonly required for antibody
formation (vaccinia ^- and serum disease ^^) after which period there
is a gradual reduction in the concentration of antibodies, and since
directly after the clinical manifestation, the antibodies may be in their
highest concentration,^'' as it is on these antibodies that the antigen
must depend for its reaction.

In these eleven cases the various antigens were used as shown in
Table 5.

Table 5

Number of antigen 1 2 3 4 Ic 5 6 7 8 9 10

Total tests 4534 3 62 3 14 4

Specific tests 0 5 0 3 1 2 1 1 0 3 0

It will be remembered that Antigens 2, 4, 5 and 7 were identical
with Antigens 1, 3 and 6, respectively, except that the former were
autolyzed in the incubator for a week or more, while the latter were
put directly on ice. Thus out of eighteen tests with autolyzed antigens,
eleven (61.1 per cent.) gave specific type reactions; while out of nine
tests with nonautolyzed antigens only one (11.1 per cent.) so reacted.
The superiority of the former type of antigen is well illustrated in
Case 11 in which autolyzed antigens gave specific type reactions on the
thirteenth and twenty-ninth days, and no such reaction on the twenty-
third day. It is very unlikely that in the interval of sixteen days
between the positive tests, the patient lost his specific type sensitiveness
and regained it again, but rather that the antigen used during this
interval was not suitable to elicit such a reaction. In this connection,
it is significant that when one case responded specifically on the thirty-
eighth day of the disease to an autolyzed antigen of culture I C

12. Rosenau. M. T. : Preventive Medicine and Hygiene, New York., D.
Appleton & Co., 1918."

13. Von Pirquet, C, and Schick. B. : Die Serumkrankheit. Leipsig, Deuticke,
1905.

14. Chickering, H. T.: J. Exper M. 20:599, 1914. Dochez, A. R.: J. Exper.
M. 16:655, 1912. Ricketts, H. T.: Infection, Immunity and Senim Therapy,
Chicago, A. M. A. Pres.s, 1906. Roscnow. E. C: J. Infect. Dis. 3:68.^. 1906.
Timnicliflf, R.: J. Infect. Dis. 8:302, 1911. Wolf, H. E.: J. Infect. Dis. 3:731, 1906.

BIGELOW— INTRACUTANEOUS REACTIONS 233

(homologous) the identically prepared, but nonautolyzed, antigen
gave no reaction.

Of the antigens in which the pneumococci were subjected to con-
siderable violence (Antigens 8, 9 and 10) No. 9 was the only one
showing any reaction to a single type and the lack of type specificity
of these reactions has already been discussed.

In the light of Weil's = work, the common reactions are especially
worthy of analysis. Of the 104 cases of lobar pneumonia tested,
forty-six (42.3 per cent.) showed the reaction with one or more
antigens, and of the twenty controls, nine (45 per cent.) showed such
reactions. They were elicited by the various antigens as shown in
Table 6.

T.-^BLE 6

RcaSs i

4
3

3
6
0

4

5
3
0

6

2
2

0
0

I

9
7
0

5

17

Reactions i

in?om"u.'":::;:::

3

The earliest reaction of this type appeared on the sixth day of the
disease and the latest on the forty-sixth day. In relation to crisis,
one appeared twenty-four hours before, two during, and the remainder
after crisis.

These reactions were at their height in eighteen hours, and were
fading in from twenty-four to thirty-six hours. They showed no
differentiation between central induration and areola. Though a central
macule, sometimes just palpable about the point of inoculation, might
persist for forty-eight hours or even longer, there was no pigmentation
or scaling. Weil describes what he considers a positive reaction to the
pneumococcus protein as follows :

Within the twenty hours following the injection further changes may occur
at the site of injection. A fairly well circumscribed area of erythema, with
slight infiltration and elevation of the skin surrounding the point of puncture,
may develop. If the infiltration is marked, a true papule results. These
changes may persist for forty-eight hours or more. . . . The normal indi-
viduals, or the diseased controls, may or may not present a reaction, depending
presumably on their previous sensitization by the pneumococcus or an allied
organism.

It would seem, then, that the reactions obtained by Weil and those
described here as common reactions were the same. Weil is not
disturbed by similar reactions in controls and feels that they are
specific for the pneumococcus protein. When a history of lobar
pneumonia cannot be obtained, he feels that an abortive and unrecog-
nized attack may be used as an explanation of sensitization. Another
explanation is that suggested by Gay ° for his "meningococcin" reaction
in normals with negative cultures, namely, that the presence of the
organisms on the mucous membrane in the past was sufficient to

234 ARCHIVES OF IXTERXAL MEDICIXE

sensitize. It is impossible to check up Gay's suggestion, because the
discovery of a carrier state in the past is beyond present day bacterio-
logic technic. At least, it can be said that a test in which 45 per cent,
of the controls react is of no service diagnosticall_\.

SUMMARY

Of 104 cases of lobar pneumonia, eleven (10.5 per cent.) gave one
or more intracutaneous reactions to only one type of pneumococcus
used, while forty-six (42.3 per cent.) reacted to two or more types.
Of twenty controls none showed the single type reaction, while nine
(45 per cent.) showed the multiple type reactions.

These two reactions are sharply dififerentiated both as to time and
character. The reactions elicited to a single type of pneumococcus were
(with the exception of those from Antigen 9) specific for the type of
organism isolated from the patient. The reactions, elicited by multiple
types of pneumococci in 42.3 per cent, of the cases of lobar pneumonia
and in 45 per cent, of the controls were not specific for the type of
pneumococcus causing the disease. To determine whether they are
specific for pneumococcus protein or not a control antigen made from
other bacteria should be used.

In 10 per cent, of the cases treated with Type I antipneumococcus
serum, specific type reactions were obtained, and in 14.8 per cent, not
so treated there were similar reactions. No one of the fixed types
showed any marked preponderance of specific type reactions. The
longest period over which the specific type reaction was obtained was
seventeen days. The largest number of the specific type reactions
occurred during the second week of the disease (six cases) and the
first week after crisis (seven cases).

With antigens prepared from simple saline suspensions of pneumo-
cocci, 61.1 per cent, of the tests, performed on the patients showing
the specific type reactions, were positive when the antigen used had
been autolyzed in the incubator for a week or more, and 11.1 per cent,
of the tests with nonautolyzed antigen were positive.

Of the antigens in which the pneumococci were treated with bile,
alcohol, etc.. only one gave reactions to a single type and tiiesc were
probably all nonspecific for the type of infection present.

No reactions comparable to those reported by Weiss and Kolmer
with their "pneumotoxin" were obtained with a similar preparation,
nor was there any specific absence of reactions as might be expected
from an analogy to the Schick test.

CONCLUSIONS

1. Intracutaneous reactions specific for the type of pneumococcus
causing lobar pneumonia may be obtained in certain cases.

B!GELOU-—IXTRACUTA.\EOUS KEACTIOXS 235

2. The reaction has not been demonstrated sufficiently early to be
of service in directing specific serum therapy.

3. The largest number of reactions occur during the period when,
on the assumption of allerg}% the highest concentration of antibodies
would be expected.

4. The sensitization responsible for the specific reactions may
persist more than two weeks.

5. The most satisfactory antigen for obtaining specific type reactions
is made by autolyzing saline solution or distilled water suspensions
of the various types of pneumococci.

6. A reaction, differing from that which is specific for type in
time and character, may be obtained in a considerable number of cases
of lobar pneumonia and controls. This appears with more than one
of the pneumococcus type antigens and is in no way specific for the
type of organism causing the infection. Whether this reaction is
specific for a common factor in all types of pneumococcus protein
or whether it is in no way specific for the organisms composing the
antigens has not been demonstrated.

CLINICAL STUDIES ON THE RESPIRATION

VIII. THE RELATION OF DYSPNEA TO THE MAXIMUM MINUTE-
VOLUME OF PULMONARY VENTILATION *

CYRUS C. STURGIS, M.D.,

FRANCIS W. PEABODY, M.D.,

FRANCIS C. HALL, M.D.

AND

FRANK FREMONT-SMITH, JR.. M.D.

Failure to obtain the required amount of oxygen or to remove a
sufficient quantity of carbon dioxid from the body tissues results in
the subjective sensation of dyspnea. The maintenance of a gaseous
exchange which is adequate for the needs of the metabolism depends
on the coordination of several factors, and insufificiency of any one
of these may produce dyspnea, but we proposed here to consider only
the subject of the pulmonary ventilation. With the body at complete
rest the demand for oxygen is supplied and the excess carbon dioxid
is removed when an individual is breathing from 4 to 5 liters of air
per minute. Physical exertion requires a greater gaseous exchange and
this is met by an increase in the amount of air breathed which is
nearly proportional to the rise in metabolism. The degree, therefore,
to which the metabolism may be increased by physical exertion without
the production of dyspnea will depend, in part at least, on the ability
to increase the minute-volume of the respiration so that it keeps pace
with the metabolism. Aside from any other factors, the maximum
minute-volume which an individual is capable of maintaining thus
sets a definite limit to the amount of exercise which he can carry on
without dyspnea, and the determination of the maximum minute-
volume becomes a point of considerable practical interest. The maxi-
mum minute-volume depends, of course, on the ability to increase the
rate, and more especially the depth, of breathing, and it is normally
subject to variations depending on such factors as the vital capacity
of the lungs and physical training, while in pathologic conditions it
may be affected by numerous influences. The diflference between the
amount of air breathed per minute when lying quietly at rest and the
amount breathed during the greatest exertion, or the maximum minute-
volume, constitutes the pulmonary reserve which enables an individual
to provide a greater supply of oxygen and remove larger quantities of

From the Medical Clinic of flie Peter Bent Brigham Hospital.

STURGIS ET AL.— RESPIRATION 237

carbon dioxid from the body when the demand is made. An individual
with a large pulmonary reserve may be capable of violent exercise with-
out dyspnea, while a patient with a very small reserve, from any cause,
as, for instance, from heart disease, will become dyspneic on slight
exertion. Hence the tendency to dyspnea depends, in part at least,
on the extent of the pulmonary reserve.

The first experiments to be reported consist of an attempt to deter-
mine the maximum minute-volume and the pulmonary reserve in normal
young men. The second series of experiments was devised to illustrate
the effect on the pulmonary reserve of a decrease in the vital capacity
of the lungs, with a consequent diminution of the maximum minute-
volume.

■ THE MAXIMUM MINUTE-VOLUME IN NORMAL YOUNG MEN

To determine the maximum minute-volume which can be main-
tained during exercise, a study was made of twelve normal, young,
male adults. The volume of air breathed was measured by conducting
the expired air through a Bohr meter while the subject was riding on
a stationary bicycle. During the period of exercise, the subject was
instructed to ride hard until he was exhausted, and toward the end of
the experiment he was encouraged to "spurt" for thirty seconds. The
duration of the experiment varied from two and one-half to fifteen
minutes, depending on the endurance of the individual. Dyspnea and
muscular fatigue were both factors in making the subject stop riding
and it was often difficult to determine which was the more important.

The results of the experiments, which are summarized in Table 1,
show that in the twelve normal subjects the average minute-volume
during the last one and one-half minutes of the observation, when the
exercise was most violent was 60.3 liters. For periods of thirty seconds
somewhat higher minute-volumes were often obtained, but these were
maintained for so short a time that they are of less practical significance
from the present point of view. No observations were made on the
minute-volume of these subjects when lying down at complete rest,
but general experience indicates that the average would not be far
from 5 liters per minute. On this basis, it is clear that the pulmonary
reserve is such that the minute-volume during violent exercise may be
raised to about twelve times its resting value.

It is of interest to analyze the high minute-volumes in order to find
out how they have been produced. There is always an increase in the
rate of respiration and this consists of an average rise to 34 per minute
during severe exercise. During the last minute and one-half of exercise
the lowest rate observed was 26 and the highest 45 per minute, but
in eight of the twelve subjects the rate was between 31 and 37. It
would appear, therefore, that a rate of respiration of about 35 per

238 ARC HIVES OF IXTERXAL MEDICI XE

minute is usually attained during severe exertion. The other factor
in producing the high minute-volume breathed during exercise is the
increase in the depth of respiration. The depth of breathing will,
of course, be governed to some extent by the vital capacity of the
lungs, but it is quite obvious that when breathing rapidly it is impossible
for the volume of- each respiration to be equal to the total volume
of the vital capacity. In the subjects investigated it was found that
the volume of each respiration, during the period when they were
breathing their maximum minute-volume, varied between 23 per cent,
and 45 per cent, of their vital capacity, with nine of the twelve subjects
using between 27 and 39 per cent. In general, therefore, an average of
about 33 per cent, of the vital capacity was made use of with each
respiration during the period of greatest exertion. It is interesting

TABLE 1. — SuMM.^Rv OF E.xperiments on Twelve Subtects

Highest

Dura-

Pulmon

ary Ventilation for

Percentage

Vital

Minute-

tion

Last IH Minutes oj Exercise

ol Vital

No.

Subject

Capacity,

\^Z

r^U.

Capacity .

Minute-

Respi- Volume

Seconds,

cise,

Volume,

ratory

per Respi-

Respira-

Liters

Min.

Rate

ration, C.c

tion

1

6. E. H.

7,180

84.00

7

79.0

39

2,050

28

A. M. G.

6,200

67.2

60.8

26

2,339

K. T. R.

5,530

60.4

5

58.9

31

1900

4

J.B.M.

5.000

60.0

2%

50.6

37

1,367

5

T. E. B.

5,180

63.0

58.7

37

1,586

31

6

W. 0. R.

5,400

66.0

6%

56.6

45

1,258

23

r. 0. H.

80.0

2,285

J. W.

5,060

74.0

15

69.5

35

1,985

9

H. C. D.

5,210

54.6

4^4

62.6

37

1,422

27

W. T. V.

5,180

50.4

5H

47.6

1,762

34

n

J. M. Mc.

4.210

66.4

7%

60.6

32

1,895

45

12

F. R. B.

5,080

50.6

48.2

31

1,555

31

Average

eo.s

34

33

that Barr and Peters ' found that normal .subjects and patients with
heart disease, made dyspneic by rebreathing carbon dioxid, were able
to use between one-third and one-half the volume of their vital capacity
at each respiration. If one accepts the usual figures associated with the
maximum minute-volume in this group of subjects, of a rate of 35
per minute and a volume per respiration equal to 33 per cent, of the
vital capacity, it is possible to compute with tair approximation the
theoretical maximum minute-volume of any individual if the vital
cajjacity is known. Thus, if the vital capacity is 4.600 c.c, the maximum
minute-volume is ^^^- X 35, or 53.7 liters. Figures obtained in such a
manner, have, of course, only a relative value, and there will necessarily
be considerable individual variation from such theoretically derived
maximum minute-volumes depending on the extent to which the indi-
vidual under consideration is able to increase the rate and depth of

irr, D. P.. and Peters

P.. .Tr

.Aliii. J. Physiol. 54:,M5. 1920.

STURGIS ET AL.—RESPIRATIOX 239

respiration. If the above formula be applied to the twelve subjects of
these observations, it is found that the actual minute-volume of air
breathed during the minute and one-half of greatest exertion was
between 79 and 129 per cent, of the calculated maximum minute-
volume, but in nine of the twelve subjects it was between 79 and 98
per cent, of the calculated maximum. The average minute-volume of
the twelve subjects is 96 per cent, of the calculated maximum. Accept-
ing the fact, therefore, that such calculations can give only a rough
approximation of the truth, it is nevertheless clear that by their use
it is possible to determine, with a degree of error that is unimportant
for practical purposes, the maximum minute-volume of respiration
which any individual can maintain, if the vital capacity of the lungs
is known. Since most persons are able to increase the rate of respira-
tion to about 35 per minute, or to the usual maximum rate during
severe exertion, it is obvious that the fundamental factor which influ-
ences the maximum minute-volume is the vital capacity of the lungs.
The clinical significance of these observations will be discussed later.

THE EFFECT OF REDUCTION OF VIT.VL CAPACITY OF THE LUNGS

ON THE MAXIMUM MINUTE-VOLUME OF PULMONARY VENTI-

L.\TION AND ON THE PRODUCTION OF DYSPNEA

The above observations having called attention to the importance
of the maximum minute-volume of the respiration in its relation to
bodily activity and the production of dyspnea, and also to the importance
of the vital capacity of the lungs in determining the maximum minute-
volume, the experiments to be described now were devised as an attempt
to illustrate the relation of variations in the vital capacity of the lungs
to the production of dyspnea.

It was found possible to cause an artificial reduction of the vital
capacity to about one-half the normal value by means of a heavy canvas
swathe strapped tightly round the chest, and in two healthy subjects
the reaction to exercise was studied both before and after reducing the
vital capacity in this way. Observations were made during the fasting
state, on the respiration rate, volume per respiration, minute-volume,
oxygen consumption, and carbon dioxid production per minute while the
subject was lying at complete rest (Period I), standing at rest (Period
II), walking sixty .steps upstairs on a treadmill in sixty seconds
(Period III), and over two consecutive five-minute periods standing
at rest immediately after the walk (Periods IV and V). The amount
of exercise involved in the walk upstairs was, of course, .slight and
caused little shortness of breath. In order to increase the exertion and
produce more dyspnea a pack weighing 50 pounds was then put on the
subject's back and he again walked sixty steps up.stairs in sixty seconds
(Period VI), this exercise being immediately followed by two five-

240

ARCH 11 'ES OF IXTERXAL MEDICINE

minute periods of observation while standing at rest (Periods VII and
VIII). In each experiment the subject had been fasting fourteen hours
so that the effect of food on the metaboHsm was eliininated. The
expired air was collected in two 100 liter Tissoi spirometers by means
of a half-mask fitting tightly over the nose and mouth and connected
to the spirometers with rubber tubing of large caliber. The oxygen
and carbon dioxid percentages in the expired air were determined by
analysis with the Haldane gas analysis apparatus. Duplicate analyses

TABLE 2. — Summary of Experiments ox Subject S. G.

Subject. S. G.: Age, 24 years; weight, 68.9 kg.; height, 171.4 cm.; surface area, 1.80 sq. m.

Experiment A: Vital Capacity = 4,600 C.c.

V. O.
Miximum Minute-Volume — x 35 = 53.7 liters.

VoUim

Minute-Volume

Respir.i

tion

Respi-
ratory

1 Percen-
tage of

sumption,

Period

Percen-

tion.

Bate

tage of

Liters

C.c.

Vital

Minutc-

ity

Volmne

t Basal

8.9
11.2

500
563

12

4.45
6.30

8
12

^l

189

II. Standing at

227

III. Wallfing 60

19.0

16

14.30

27 no

■m

dyspnea

765

16

V. Best of 5 min.

11.8

537

12

6.35

VI. Walking 60

23.0

792

17

18.20

steps carrying
W pound pack

dyspnea

VII. R«stof5min.

13.0

1,120

24

VIII. Rest of 5 min.

13.0

540

12

7.04

13

Experiment B. Vital Capacity = 2,100 C.c.

V. C.
Maximum Minute-Voliune = x 35 =r 24.5 liters.

353

17

6.15

25

250

201

II. standing at

19.3

386

18

7.44

30

III. Walking 60

28.0

5«0

27

15.70

826

steps

dyspnea

IV. Rest of 5 min.

26.2

615

24

13.50

IS^

V. Rest of 5 min.

23.1

435

21

10.00

VI. Walking 60

28.0

696

33

19.50

80 more

'tO pound pack

21.0

705

34

J^

vni. Rest of 6 min.

20.3

490

23

9.83

2-6

were done on samples from each period and the results were not
accepted unless they checked within 0.03 per cent, of the average for
oxygen and 0.02 per cent, of the average for carbon dioxid. The
respiratory rate was determined by means of a pneumograph leading
to a tambour which recorded the respiratory movements by means of
a lever marking on a smoked drum.

STURGIS ET AL.— RESPIRATION

TABLE 3. — Summary of Experiments om Subject F. F-S.
Subject, F. P-S.: Age, 25 years; weight, 72.1 kg.; height, 179.2 cm.; surface area, 1.90 sq.
Experiment A: Vital Capacity = 4,200 C.c.
Maximum Minute-Volume :

V. O.

Volume per

Minute-Volume

Respiration

COs

Rcspi-
ratory

Percen-
tage of

Con-
sumption,

Period

Percen-

tlon.

tage of

Liters

C.c.

ity

Volume

I. Basal

11.8

S13

7

3.69

8

202*

145

n. Standing at

14.0

420

10

5.87

12

240

202

lU. Walking 60

23.0

687

16

15.80

32 no

9.70

V. Rest ol 0 min.

14.0

450

11

6.30

13

26.0

705

17

18.30

37 slight

mo

steps carrying

dyspnea

50 pound pack

VTI. Rest of 5 min.

16.8

625

15

21

VIII. Best of 5 min.

14.5

457

11

6.63

14

Experiment B: Vital Capacity = 2,550 O.ct
\
Minute- Volume = -

35 = 29.8 liters.

302

U

5.57

18

224

IS?

ir. standing at

18.2

381

15

6.M

III. Wailcing CO

3i.O

610

24 .

18.90

64 mod-

1090

752

steps

TV. Best of 5 min.

23.2

510

20

11.90

40

18.8

430

17

8.08

27

266

VI. Walking 60

35.0

615

24

21.60

72 more

steps carrying

dyspnea

27.8

530

21

14.70

518

VIII. Best of 5 min.

22.4

375

15

8.40

28

• There was no apparent reason to account for this subject's abnormally low oxygen
consumption. He was an active, alert, fourth year medical student with none of the signs
or symptoms suggesting hypothyroidism. His basal metabolism on this occasion was —24
per cent, of normal when compared to the standards of DuBois.

♦ The vital capaetly was 2,700 c.c. in the basal period. In periods TT to VIII it was
decreased to 2,M0 c.c. by tightening the swathe. This alteration results in a theoretical
maximum minute-volume of 31.5 liters in the basal period and i!9.8 liters thereafter.

ANALYSIS OF RESULTS IN TABLES 2 AND 3

Tables 2 and 3 give the results of the two experiments. In both of
them Experiment A was performed while the vital capacity was normal,
while in Experiment B the vital capacity was reduced approximately
50 per cent, by means of the swathe. In Experiment A, Table 2, when
the subject, S. G., was lying quietly at complete rest he was breathing
4.45 liters of air per minute. His maximum minute-volume, com-
puted according to the formula sugge.sted above, is 53.7 liters, so that
during this first period he was breathing only 8 per cent, of his
maximum minute-volume. In period III. while walking sixty steps in

242 ARCHIJ'ES OF IXTERXAL MEDICIXE

sixty seconds, he breathed 14.30 liters per minute, or 27 per cent, of his
maximum minute-volume and yet experienced no dyspnea. In Period VI,
while walking sixty steps upstairs in a minute and carrying a 50 pound
pack on his back, the minute-volume rose to 18.20 liters, or 34 per cent,
of his maximum minute-volume and slight shortness of breath was
noticed. In Experiment B, Table 2, the vital capacity of the lungs
was reduced to 2,100 c.c. in the same subject and the calculated
maximum minute-volume was thereby diminished to 24.5 liters. Under
these conditions, although the subject was only breathing 6.15 liters
per minute in Period I, while lying at complete rest, he was. neverthe-
less, using 25 per cent, of his maximum minute-volume. In Period III,
when walking upstairs, he used 64 per cent, and was moderately
dyspneic, and in Period \T, when walking upstairs with a pack, he
used 80 per cent, of his maximum minute-volume and had marked
dyspnea. In general, therefore, there is a striking correspondence
between the degree of subjective dyspnea and the extent to which the
maximum minute-volume is eiicroached upon. The actual minute-
volumes in the corresponding exercise periods (Periods III and \T)
in the experiments with and without the swathe were not very different
but the degree of dyspnea was much greater in the experiment in
which the vital capacity was reduced. Experiment B, and in which a
larger proportion of the theoretical maximum minute-volume was made
use of. Exactly the same relationships are found in the case of sub-
ject F. F-S. which are given in Table 3. Here, again, the subjective
sensation of dyspnea varied, not with the actual minute-volume breathed,
but with the extent of utilization of the maximum minute volume or,
to express it in another way. with the decrease in the pulmonary
reserve.

In both sets of observations shown in Tables 2 and 3 tlie oxygen
consumption was approximately the same in the corresponding periods
of Experiments A and B. There is a general tendency to a slightly
higher metabolism in the experiments with reduced vital capacity and
this is probably tiue largely to the discomfort incident to the very
tight swathe and to the greater exertion associated with more rapid
breathing. The minute-volumes for the corresponding periods are
almost uniformly higher in the experiments with reduced vital capacity.
This depends on the fact that in order to obtain the volume of
alveolar ventilation necessary to meet the needs of the metabolism it
was easier to increase the rate of respiration than it was to increase
the depth. Under such circum.stances the "dead space" of the upper
respiratory tract becomes an important factor and the total minute-
volume of pulmonary ventilation must be increased in order to produce
the same volume of alveolar ventilation.

STURGIS ET AL.— RESPIRATION 243

DISCUSSION

The first observations described were made on twelve normal young
men who rode on a stationary bicycle until they were forced to stop
on account of complete exhaustion. During the last one and one-half
minutes of the ride, when the exercise was most violent and the
dyspnea great, the average minute-volume of air breathed was 60.5
liters, or approximately twelve times the average minute-volume of
such a group of men when they are lying down at complete rest. These
figures give a general indication of the normal pulmonary reserve, by
virtue of which the individual is able to increase the pulmonary ventila-
tion and keep it adequate to the needs of the body when the metabolism
is raised far above its resting or basal value by severe exercise. An
analysis of these high minute-volumes shows that they were obtained
by increasing the rate of respiration up to an average maximum of
about 35 per minute, and by increasing the depth of each respiration
up to a volume which approximates one third of the vital capacity of
the lungs. On this basis, it was suggested that the maximum minute-
volume which can be maintained for more than a very short period
can be calculated in any given instance with an accuracy which is
sufficient for practical purposes, by multiplying one-third of the vital
capacity of the lungs by 35. Such a calculated or theoretic maximum
minute-volume has considerable clinical significance for it tells about
how far the person concerned is able to increase his minute-volume,
and consequently how great an increase of metabolism he can meet
with a pulmonary ventilation which will ensure proper aeration of the
blood in the lungs. If the minute-volume of respiration can only be
raised to three times the volume that a given subject breathes while
at rest, then he will not be able to meet much more than a three-fold
increase of metabolism, and he will become dyspneic walking slowly
upstairs, but if the minute-volume can be raised to six or seven times
above the resting value then the subject will be able to carry on most
of the activities of a normal life without subjective dyspnea. The
maximum minute-volume calculated in this manner is thus a guide to
the amount of physical exertion that any individual may be expected
to undertake.

The maximum minute-volume that a person can breathe depends on
the rate and depth of the respiration. The observations reported appear
to show that the highest rate that is compatible with efficient respira-
tion is about 35 per minute, and practical experience indicates that
most persons, whether or not they are in normal condition, can raise
their respiration rate to this figure. The important factor in producing
variations in the maximum minute-volume is thus the vital capacity
of the lungs, which determines the pos.sible depth of breathing. This

244 ARCHIVES OF IXTERXAL MEDICINE

is, of course, subject to marked changes in normal, and to a greater
extent in pathological conditions.

In the second series of observations reported, two subjects were
studied while walking upstairs on a treadmill. One set of experiments
on each of them was carried out in the normal state, and one set when
the vital capacity was reduced to about one-half by means of a tight
chest swathe. The wholly artificial condition this produced is not
unlike that in a case of pleurisy with effusion and simulates in some
degree other conditions, such as heart disease, in which the vital
capacity of the lungs is below normal. The calculated maximum
minute-volume was, of course, much less in the experiments with the
chest swathe than in those in which the vital capacity was normal and
it was found that the tendency to dyspnea varied closely with the
percentage of the calculated maximum minute-volume that was being
used in respiration. When the minute-volume breathed was only 25
per cent, of the maximum, the subject was not conscious of his respira-
tion ; when it was 50 per cent, he noticed that he was breathing deeply,
and when it was 75 per cent, of the maximum he was frankly short of
breath. These experiments, therefore, are of interest in that they
bring out the importance of the conception of the maximum minute-
volume of the pulmonary ventilation as one of the factors which
determine the occurrence of dyspnea in various clinical conditions.
This factor can be easily expressed in a sufficiently accurate quantita-
tive manner and it has a broader physiologic significance than has the
determination of the vital capacity of the lungs alone, for the minute-
volume of pulmonary ventilation bears a close relationship to those
fundamental processes of the body which go to make up what is known
as the metabolism. The essential cause of the variations which may
occur in the maximum minute-volume is an alteration in the vital
capacity of the lungs and this may, therefore, be regarded as a practical
index of the maximum minute-volume.

POSITION AND ACTINTriES OF Till-: I)1AI'I1RA(,M
AS AFFI-:CTEL) BY CHAXCiES OF FoSITKl-:*

ROY D. ADAMS. M.D.

HKXRV C. PILLSHLRN. Ml).
Major, M. C. U. S. Army

WASHINGTON, D. C.

Most Studies of the diaphragm are made with the subject standing,
sitting or lying prone or supine. Such consideration ignores the fact
that in health, as well as in disease, much time is spent lying on the
si<le. and that the position and activities of this muscle and adjoining
organs are greatly changed by shifts of posture. .Although the literature
is not without more or less detailed reference to these changes, their
magnitude and clinical importance are popularly underestimated. .-Ks
tliey affect directly physical signs, and as convenience or neces.sity
frequently demands examination of a patient lying on one or the
other side, it follows that an understanding of the action of the
dia])hragm in lateral positions of recumbency possesses an inten-st of
practical value.

The relaxed diaphragm, in health, assumes a height and position
dependent on forces e.xerted from below as well as from above. Recum-
Ijent on the back or face, gravity ojjerates on the viscera in such a
direction as to affect the diaphragm least. In this position the organ
arches high into the chest. When standing or sitting, gravity comes
into action on the thoracic organs above and the abdominal viscera
below in such a manner as to push and draw the diaphragm down-
ward. Its level is lower in the sitting position because of relaxed
.ihdominal walls and consequent opportunity offered the viscera to sag
forward as well as downward.

While a number of normal subjects form the ba.sis of this discussion,
in order to reduce the problem to its simplest terms, effort has been
made to present an average specimen. Such a process is not without
acknowledged error, in that wide variations from a so-called average
present themselves. However, as variations in different types, in
respect to the points under discussion, are of degree rather than kind,
tile error is not of a serious nature.

In our studies of nf)rmal chests, a fixed point, arbitrarily selected
on the first lumbar vertebra, has been utilized. \'ariations in height of
ihe diaphragm are measured on a line drawn from this point parallel

* Read by invitation before the Medical Section of the College of Physicians
>f Philadelphia, April 25, 1921.

246 .tRCIIII F.S OF JXTF.RX.IF MFPICFXF

to the verte1)ral column. In a state of normal inspiration, the distances
ahn\i' the marker of the right and left domes respectixelv are;

Standing : 6.5 and 5 cm.

Sitting: .t.5 and 3 cm.

Prone on abdomen: 12 and 0 cm.
It is tlnw ^een that the right dome is .^..^ cm. and the left dome
(i cm. higher when the suhjecl is reclining than when he is sitting.
In all three ]iositions, the excursion during normal, mi.xed breathing
is about e(|ual on the two -ides and \aries from 1.5 to 2.? cm. in
extent.

With assumption of the right prone position, the upper or left leaf
of the diaphragm descends, opening to a considerable e.xtent the left
costophrenic angle, while the dependent or right leaf takes a higher
level, at the same time increasing the extent of costophrenic contact.
During quiet breathing, the excursion of the dependent half of the
muscle is greater than that of the upper, the ratio being approximately
2 to 1. Recumbency on the left side reverses conditions, the right half
of the diaphragm assuming a lower level, and exhibiting less excursion
than the dependent left half. The explanation of these changes is
simple on a purely mechanical basis. Both thoracic and abdominal
viscera, especially the latter, because of their great weight and mobility,
sag toward the dependent side, thus opening the upper costophrenic
angle and straightening the upper diaphragm. In this position, the
effect of equal contraction of the halves of the muscle will differ on
the two sides. The dependent half, being highly arched, will descend,
whereas the upper, approaching more nearly a |ilanc. will tend rather
to antagonize the intercostal muscles. In otlier words, the ujiper half
works to a better mechanical, but poorer resjiiralory ;idvantage than
does the dcnendcnt half.

Ni;i(,iii;(iKiX(; ircans

The effects of lateral rccuniliency on the positions and activities of
adjacent organs arc mrniifolij. llii\\e\fr, we here conlinc oiu" remarks
to three, namely; ( 1 i tlic |>o^iiion of ilic hcari ; (2i marginal sounds
or rales; (.i) breath somuK at the bases.

/. Thr llrarl. In .ill positions the roentgenograms on which our
results are based wimc tal<en at a distance of live feet from the plate.

more than 1 cm. from tllo^c (if the sulijccl. so that errors in detail
are com])arative1y slight. The cardiac shifts in position are noted in
figures which represent the distance of the apex from the midline
of the spine. In some instances, the apex is near to or remote from
the spine because of lateral tilting of the cardi.nc axis. In others, a

m^-

#^

248 ARCHirES OF IXTERXAL MEDICI XE

sagging mediastinum allows the entire cardia to fall toward tlie
dependent side. Rotation is a third factor in the production of varia-
tions in heart shadow. Finally, two or more ot these factors may be
found combined.

The distance of the apex from the midline, in the average chest
used for detailed illustration, in the sitting position is 8 cm. ; prone on
the abdomen, it is 10; both measurements taken at the end of a normal
inspiration, a difference of 2 cm, effected b} simple shift from a
vertical to horizontal plane.

At the end of normal expiration, with the subject on the right side,
the apex is 6 cm. from tiie midline. .\t the same respiratory phase,
prone on the left, it i> 12.3 cm., a total difference, allowing for slight
error of oblique rays of about 6 cin. The difference in the same
positions, at the end of full inspiration is much less marked, being
only 1.5 cm. This is as one would naturally expect because of the
firm pericardiophrenic attachment and the consequent straightening of
the mediastinum incidental to descent of the diaphragm.

The distance of the left border of the heart from the midline at
the end of normal inspiration, with the subject standing, is 8.5 cm.,
sitting, 8 cm.; prone on abdomen, 10 cm.; recumbent on the right side,
6.5 cm.; prone on left side, 11.7 cm. That is to say, with the ascension
of the diaphragm consequent to a change from tlie vertical to horizontal
position on the abdomen, the apex shifts to the left a distance of from
1.5 to 2 cm. Turning on the left side produces a further movement
to the left of a little over 1.5 cm., while recnmbencv on the right side
is accompanied by a shift to the right n\ .v5 cm. The left border,
then, in the right prone position, is only frnni 1.5 in 2 cm. to the
right of the point which it occupied in the upright jjosition, whereas
prone on the left .side it is removed from this point by a distance about
twice as great. However, the subject ha\ing assumed a recumbent
position may accomplish a greater cardiac >hift by turiiin<; njion tlic
right side than by turning to the left.

2. Marginal Sniiiids. ]\\ llic term iiiarLjinal snumN w r lefcr In tli.il
group of sounds which i> heard ci\er the base cif the lung (luring; deep
inspiration. They vary in different individuals from dry crepitations nr
crackles to sounds resembling fine or medium size moist rales. Such
sounds may be elicited in ahrost any imrnial individual capable of
abdominal breathing. riie\ arc beard n\er tin- nh^c n\ descending
lung and therefore assimK' the ferni of a \\a\c rather than n\ a IincmI

space by diapliragniatic cnnlraction. It is greatest in the axilla, tapering

.\D.\MS-PILLSBrRY—niAI'IIR.\GM .IXD POsrCRE

249

I'ositions of recumbency are especially favorable to the production
of marginal sounds because of the high position of the diaphragm and
consequent increase in the complimental space. Right lateral recum-
bency further increasing the right complimental space, amplifies the
extent of marginal sounds on the right side. .\t the same time, by
opening the left space, this position diminishes or obliterates the .^oimds
on the upi-er ^ide. Reversal of position, reverses conditions. The
sounds may be produced or dissipated at will by having the subject lie
upon one or the other side.

Fig. 4.— Prone on left side, end of normal 1 ig. 5.— Same suljjcct prone on left side,

expiration. Diaphragm high on left side, end of full inspiration. Diaphragm levels

right costophrenic angle widely open, heart equal on two sides and heart occupying slight

prolapsed far to left. left position.

.\otc: .Showing wide excursion of dependent left diaphragm ai
from parietal pleura Corresponding to area of marginal sounds oi
the change in axis of heart and position of apex.

(Because of the necessity for economy in reproduction of photographs, the illustrations in
Ihc corresponding respiratory phases with the subject prone on the right side, are omitlc<l.
Reversal of position, however, produces a corresponding reversal of position of the organs.
With deep inspiration, there is a correspondin; wide excursion of the dependent diapliragm,
together with a cardiac shift toward Ihc left.)

1 wide extent of separation
the dependent side. Nolc

The tendency of an individual in dee]) breathing is always toward
over inflation, so that after a number of insjiirations the diai)hragm
becomes for awhile establi.shcd at a lower level. Its e.xcursion is
thereby considerably curtailed and marginal sounds disappear, to recur

250 ARCHIVES OF IXTERSAL MEDICIXE

only after rest sufficient to permit the lung elasticity to reassert itself.
This situation does not arise, or it may be obviated, when the subject
instinctively, or as a result of instruction, forcibly exhales sufficiently
to prevent over inflation.

In explaining the occurrence of marginal sounds, it is i;ecessary
to bear in mind the fact that at their point of audibility, three events
are transpiring. The phrenic pleura is separating from the costal ; tlic
vi.sceral jjleura is gliding downward to replace the phrenic, and finally,
air is entering the margins of the lungs.

That the second of these phenomena is not productive of marginal
sounds may be assumed from the fact that viscerocostal friction is
elsewhere inaudible in the chest. Moreover, marginal sounds are heard
only during inspiration.

The view most generally accepted is based on inflation of the
marginal air spaces, a theory dependent on complete or incomplete
atalectasis of the lung. There are serious objections to such an explana-
tion. In the first place, marginal sounds are heard in almost any
normal subject capable of deep breathing. They may be repeatedly
dissipated or produced at will by over inflation or strong deflation of
the lung. They may likewise be made to appear and disappear by
changes of position in lateral decubitus. Most convincing of all is tlie
fact that they occur over a moving line keeping pace with l.itten"s
Shadow and correspimding, not f)nly to the downward moving margin
of the lung, but also tn the line of cleavage of the diaphr;igmatic from
the parietal pleura.

I'.xplanalion for tlic prnchunion n( marginal sounds ba>ed on sej)-
arati(Mi of the ])hrcnic frnm tlic parietal ])leura is jiroviiled for in two
ways. In tlie first place, otiier ex|)lanations. under close scrutiny, fail
to withstand criticism. In the second ])]ace, the "peeling off" of the
diapliragni from the costal jileura contril)utes a mechanism unitiue in
pleural o.ntarl and, at lea>t, iheoretically vati-lying. I'urther, ii ha>
been fotnid thai varying condilidii- of moi^lurc■ and dryni--s cf the
pleural surfaces bear ;i direct relation to the (inality and inlcnsiiy ot
marginal sounds.

The points \\e wish to emphasize in connection with marginal rale^
are: ( I i riRv may lie elicited in the normal subject. (2) They are a
tribute t(i the muliilits of the dia])hragm. (3) They arc reversilile in
positidii and extent. (4) They are audible at the jxiint of cleavage
of the diaphragm from the che>t wall and may therefore be heard
early ky late in deep ins]iiration according a> one listens at the lop or
Ixittom of the costo|)lirenic rpace.

.V l!i;\:fli Somiil.s a' Ihc Ha.us. .\u-cuitatioii n\ the b-ick fvoni the
scvemh vertebra tn the lia>e, with the vul>ie/l prone on the side.

ADA.^ts-PiLLSBrRy—r>i.ir/iR.u;\i .i\n rosrrR!: 251

reveals a more intense resiMralnry murmur on the dependent side. Tiie
difference is more marked in ilu- i-.\|iir;it(iry than in the inspiratory
phase. Relative intensificatiim of hreath sounds on tiie lower side is
due to two factors; first, the more active ventilation of the lower ])art
of this lung as revealed by the greater diaphragmatic excursion. In
the second place, the relative compression and actual relaxation (if tlie
lower side furnish conditions favorahle to better transmission of sound.

Fig. 6. — Subject i)ri>ne on right side, end of
normal inspiration. Showing high position of
right doine of diaphragm, low position of left
dome, also wide cardiac and mediastinal
shadow to right of vertebral column.

-Vote the vast difference in levels of diaphr
from right to left prone position.

Fig. 7. — Same subject prone on left side, ( n(
of normal inspiration. Sliowing high posUioi
of left dome and low position of righl
Little cardiac and mediastinal shadow seen ti
riglit of verteliral colnmn.

bv

Inasmuch as the inspiratory murmur is |iro<hiced close lo the ear. while
the expiratory .sound is initiated at a point reujote from tiie surface
of the lung, it follows that in the inspiratory phase, ventilation is tiie
more important agent, whereas in exi)iration, conduction ])rohalily
])lays a great role in augmenting sounds on the lower side. .\ccomp;my-
iiig intensification of the hreath sounds just referred lo. there is an
increa-ed intensity of the whispered sound. s])oken word .-ind of
tactile fremitus.

252 ARCHll-ES OF IXTERX.IL MEDIC/XE

SUMMARY

To summarize; The diaphragm is highest in the prone position;
intermediate, with the subject standing erect ; lowest in the sitting
l)osture. Its excursion in these positions is equal on the two sides.
Prone on the right side, the right dome is higher than the left and its
excursion is in excess in the proportion of 2: 1. Re\ersal of jMisition.
reverses the height and excursion of the two sides.

The position of the heart, accompanying changes in position of the
diaphragm, is subject to a wide range. The extreme excursion is 6 cm.

Marginal sounds are heard over the healthy lung and are not
incidental to or dependent upon pathology in the lung or pleura. They
are heard over a broad area on the dependent side in lateral decubitus
because of the greater extent of the complim^ntal space. These sounds
are heard best dtiring vigorous inspiration following forceful expiration.

In lateral recumbency, the dependent lung is relatively rela.xed.
Its diaphragmatic ventilation is in excess of that of the u\i\)ev lung.
Breath, voice and whisper sounds and tactile fremitus are all increased
as compared with the opposite side.'

1. Bushnell, George E. : Some Rxtra Pulinonarv Sounds Which Simulate
Riles, Med. Rec. (Jan. 20 1 1512, Vol. 81, Page 101-107 inc.; Marginal Sound.s
in the Diagnosis of Pulmonary Tuberculosis. Med. Ret. ( Dec. 21 ) 1912. Vol. iS2,
Page 1109.

Gerhardt: .'Kuskultation unl Perkussion.

Hoover, C. F. : The Functions of the Diaphragm ;inrl their Diagnostic
Significance, .Arch. Int. Med. 12:214 (.Vug.) I9U3.

Morrison; Lancet 1: (May 6) 1911. Pa.ge 1202.

AURICULO\"ENTRICULAR RHYTHM AND DIGITALIS*
HENRY B. RICHARDSON. M.D.

NEW YORK

Auriculoventricular rhythm, a condition in which the A-\ node
assumes rhythmicity independent of the sinus node, is due to one of
two causes ; depressed irritability of the sinus node or enlianced
irritabihty of the A-\' node. Experimentally both types have been
produced, the former by tying off ( Engelmann ^ ) , injuring (Hering^).
or cooling (Ganter and Zahn '■') the sinus node, the latter by stimulating
the A-\' node either directly (Lohmann*) or through the left acceler-
ator (Hering;^ Rothberger and Winterberg •*). Meek and Eyster '
reviewed and extended the experimental work. By means of electrodes
connected with a string galvanometer and applied direct to the heart
they obtained evidence that in A-\^ rhythm the electrical activity
originates in the A-\' node.

Clinically, both types of the A-V rhythm have been observed, the
type in which the irritability of the sinus node is depressed, character-
ized by a slow rhythm, and the type in which the irritability of the
.\-\" node is enhanced, characterized by a relatively rapid rhythm.
The former, though rare, is much the more common of the two.
White ^ distinguishes further between A-V rhythm and simple
ventricular escape. In the former the excitation wave arises from
the A-V node alone, in the latter from both nodes simultaneously. He
bases the distinction on the forfn of the P wave which is known to be
inverted or upright according as the excitation wave arises near the
A-V node or near the sinus node. A-\' rhythm is. therefore, char-
acterized by an inverted P wave, simple ventricular escape by an
upright P wave. Of the two conditions A-V rhvthm is much the
more permanent.

The accompanying table indicates the variety of the conditions
in which A-V rhythm occurs, together with the outcome of the cases.
No death occurred as a result of A-V rhythm alone. In the fatal
cases the postmortem examination revealed ample causes for death

* From the Cardiovascular Service of the Presbyterian Hospital.

1. Enffelmann, Th. W. : .^rch. f. Anat. u. Physiol. Physiol. Abth., 505. 1903.

2. Hering. H. E.: Arch. f. d. ges. Physiol. 136:466. 1910.

3. Ganter. G.. and Zahn. A.: Arch. f. d. ges. Pliysiol. 145:335, 1912.

4. Lohmann, A.: Arch. f. Anat. u. Physiol.. Physiol. Abth.. 431. 1904.

5. Hering. H. K.: Zentralbl. f. physiol. 19:129 (June) 1905.

6. Rothberger. C. J., and Wintcrberg, H. : Arch. f. d. ges. Physiol. 135:559,
1910.

7. Meek. W. J., and Evster. J. A. E.: Heart 5:227. 1913.

8. Whi(e. P. D.: Arch. Int. Med. 18:244 (Sept.) 1916.

254 ARCHIVES OF IXTERSAL MEDICINE

quite apart from the A-\' rhythm. In the other cases the arrhythmia
was either transient or relatively short, with one exception, the case of
Williams and James," in which it was observed for fourteen months.
During this period the patient improved. It is evident, therefore, that
in and by itself A-\' rhythm is not dangerous to life. When asso-
ciated with severe infections or cardiac disease it is the associated
disease which determines the prognosis.

The case reported here is of especial interest in view of its relation
to digitalis. Such a relation has been reported but rarely. Norrie
and Bastedo i" observed it in a case of heart-block following the
administration of digitalis. Cohn and Fraser,^^ also Cohn ^^ give
a clear description of A-V rhythm as an eflfect of digitalis. White"
observed it in a case of auricular flutter in which the administration
of digitalis was followed first, as expected, by auricular fibrillation
and then by A-V rhjthm instead of normal rhythm. Later ^ he
recorded two cases of A-V rh\thin following digitalis, the first a slow
rhythm arising, in part at least, from depression of the sino-auricular
node, the second a more rapid rhythm resulting from irritation of the
A-V node. Egglestion '* mentions A-\' rhythm as one of the toxic
effects of digitalis.

REPORT OF CASE

History. — .\ woman, aged 47. housewife, was lirst admitted to the Presby-
terian Hospital, March 10, 1920. Her father and mother had died aged 35
and 25, respectively, and four of her brothers and sisters had died in infancy.
Marital history included one miscarriage and two stillbirths. At the age of
18 she Iiad an attack of acute arthritis in the left knee. Her symptoms began
four weeks prior to admission and consisted of epigastric pain radiating
to the left arm, dyspnea and edema.

Physical Exaniiiiatioii. — This revealed cyanosis, emaciation and dyspnea. The
heart was enlarged and a blowing systolic murmur was heard at the apex.
Signs of a moderate amount of fluid in the right chest were present, together
with congestion at the bases of the lungs. The liver was much enlarged and
there was moderate edema. Blood pressure: 150/98. The roentgenogram
showed a cardiac shadow extending 6 cm. to the right and 9.75 cm. to the
left of the midline.

The Mood Wassermann \vas negative. Blood urea was 0.78 gm. per liter;
phthalein excretion, 19 per cent.

Clinical Diagnosis. — Chronic cardiac valvular disease; mitral insufficiency;
cardiac insufficiency ; caries of teeth.

Treatmenl. — She received 45 minims of the tincture of digitalis daily from
March 17 to March 25.

9. Williams, H. B., and James. H. : Heart 5:109, 1913.

10. Norrie and Bastedo: St. Luke's Hospital, New York M. & S. Rep.
No. 2. p. 101.

11. Cohn and Fraser: Internal. Med. Congr., Sec. 6, Pt. 2, p. 258. 191.1.

12. Cohn. A. E.: J. A. M. A. 65:1527 (Oct. 30) 1915.

13. White, P. D.: Arch. Int. Med. 16:517 (Oct.) 1915.

14. Eggleston, Carv : Am. J. M. Sc. 160:625, 1920.

RICHARDSON— RHYTHM AXD DIGITALIS

255

Clinical Course. — She was readmitted May 12, 1920, after rapid return of
decompensation. Physical signs were essentially the same, cardiac borders
somewhat wider but liver not palpable. An irregularity of the pulse was
noted. Blood pressure: systolic, 180; diastolic, 120. To the previous diag-
nosis were added hypertension and chronic nephritis. She was admitted for
the third time September 3. For a while she improved gradually, but then
developed ascites and was twice tapped. She died October 7 of pneumonia.

In taking the graphic records the galvanometer was standardized so
that the introduction of 1 millivolt of current caused a deflection of the
string 1 cm.

Fig. 1.— May 14, 1920. Lead I. Control record taken li
of digitalis. Sequential rhythm. P-R interval CIS. Re
tricular extrasvstoles recorded. Time, Mt, sec.

nre administratii
1 = 80. Two ve

■p^= : --^----p^-f^- '-

I

/' -1 "^

■^•■^

^.^.^

'

- i._

. ._j_^

Fig. 2.— June 1, 1920. Lead L Transition from sino-auricular to A-V
rhythm. Lead IL A-V rhythm fully developed. P-R interval has changed
from -1-0.18 to —0.18. Rate, 43.5.

The electrocardiograms taken on the first admission were essentially
the same as shown in Figure 1. At this time the effect of digitalis
was limited to minor changes in the T wave. On the second admission,
the effect of digitalis was quite different. Figure 1 presents the control
record taken on the third day. The P wave precedes the R wave

256 ARCHIJ-ES OF IXTERXAL MEDICIXE

by a normal interval of 0.18 seconds, indicating a sequential rhythm,
that is, one in which the pacemaker is located in the sinus node. P is
of normal contour except that it is inverted in the third lead. The
Q. R. S. complex is essentially normal except for a deep notch in
Lead III. Frequent extra systoles occur. Digitalis was then given in
the form of the tincture, 30 minims a day for five days, and 20 minims
a day for two days. Figure 2 is the curve recorded just after cessation
of digitalis. It is characterized by a marked bradycardia and a striking
variability in the time relation of the P and R waves. Reading from
left to right the P wave gradually approaches the R wave until it merges
with it. This change is associated with a decrease in rate from 48
to 43. In Lead II the change of the position of P has continued until
P appears after R. This wave is a clear example of auriculoventricular
or nodal rhythm.

On the third admission no record was taken until after a course of
digitalis consisting of 2 c. c. digifolin and 172 minims of the tincture
in the course of eight days. This record (Fig. 3) shows essentially
the same mechanism as Figure 2 ; bradycardia and A-V rhj'thm. Two
days later only a single delayed nodal beat was recorded (Fig. 4) ;
sequential rhythm otherwise present, though with a marked variation
in shape and rate of P wave. One week later an occasional delayed
nodal or A-V beat was recorded and two weeks later a reversion
to a more rapid sequential rhythm with occasional extrosystoles of
ventricular origin (Fig. 5).

There can be little doubt that the mechanism represented by Figures
2 and 3 is that of A-V rhythm. A possibility to be ruled out is A-V
block with complete dissociation. In this case the auricular wave must
appear sooner or later in the middle of diastole. This it failed to do
in any of the several curves in which A-V rhythm was recorded. It
is conceivable also that the abnormal P wave belongs to the following
ventricular complex, in other words that there is a P-R interval of
over a second. Such a delay in conduction without block is hardly
probable.

The gradual transition from normal to A-\' rhythm is jiuzzling.
Two explanations have been advanced. According to the first the
transition represents a stage during which both nodes are active. The
auricle, receiving its stimulus from the sinus node, contracts in the
normal way and its contraction is recorded as an upright P wave. The
stimulus arising from the A-V node activates the ventricle prematurely.
Hence the combination of an upright P wave and a diminished P-R
interval. .A^ccording to the second explanation the pace-maker simply
migrates first to the auricular end of the A-\' node and then gi-adually
downward through this structure. .'\s most of the delay in conduction
takes place in the A-\' node tliis migration is accompanied bv all the

R/CHARDSOX-RHrTHM AXD DIGITALIS 257

changes from positive to negative P-R interval. The auricle, receiving
its stimulus from below, would be expected to contract in such a
manner as to produce an inverted P wave. Ganter and Zahn ^ and
Lewis ^= favor the former explanation, Hering,^ Weil ^° and Meek
and Eyster ' the latter. In the present case the fact that the P wave
remains upright throughout the transition tends to favor the hypothesis
of an interference between two nodes, both active.

Fig. 3.-Sept. 14, 1920. Lead HI. A-V rhythm. P wave merged with ven-
tricular complex. Rate = 38.3.

f) IX ni~ l • .' ^-°- ^*=="1 I"- Sequcitial rhythm. Kate = 67.5. PR =
?n ,h.^v J'rT-'^',r''^'y''°'^ ^hown. (The inverted P wave was present
.n the th>rd lead m all records and is of no apparent significance.^

907?'Dr'"v/rrulfn' ^AK^'"'- "( ""■ ^-"'""•'^»' E.-amination.-^ccropsy
H«r, TU ,Abdommal cavity 1,000 c.c. clear fluid.

J5- Lewis, T.: Heart 5:247, 1913.

16. We.l, A.: Deutsch. Arch. f. klin. med. 116:486. 1914.

258 ARCHIVES OF I.XTERXAL MEDICIXE

epicardium over the right auricle is roughened and the auricle is greatly dis-
tended. The heart weighs 420 gm. The right auricle contains postmortem
clot and in the appendage and clinging to the wall of the auricle are ante-
mortem thrombi. Just at the opening of the superior vena cava in the auricle
the endocardium is quite thickened and opaque over an area about 2 cm. in
its greatest diameter. There is also some thickening of the intima of the
cava at this point. The foramen ovale is closed. The tricuspid valve shows
a little thickening of the cusps but it is quite flexible. The columnae carnae
stand out quite prominently. There are seen beneath the endocardium a few
yellow areas which appear fatty. The pulmonic valve is thin and delicate.
The left auricle contains only postmortem clot. The endocardium is slightly
thickened over the entire auricle. The mitral valve leaflets are slightly thick-
ened, but flexible. At the apex of the left ventricle are a few small grayish
red thrombi which have softened centers. These are adherent to the wall of
the ventricle, in the spaces between the columnae carnae. On the anterior
and left posterior leaflets of the aortic valve are small clumps of gray trans-
lucent vegetations which break oft' readily. They are situated over the corpus
aurantii. The coronary arteries show a few yellow fatty plaques, without calci-
fication or obvious stenosis of the lumen. The myocardium is grayish red in
color with here and there lighter areas due to connective tissue.

Anatomic Diagnosis. — Scars of endocardium of right auricle; cardiac hyper-
trophy ; fibrous myocarditis ; acute cardiac valvular disease, aortic ; thrombi in
right auricular appendage and left ventricle ; chronic passive congestion of
liver and spleen ; ascites ; chronic diffuse nephritis ; confluent lobular pneu-
monia ; acute fibrinous pleurisy ; acute fibrinous pericarditis ; arteriosclerosis,
slight.

Microscofyic Examination. — Serial sections of the sinus and A-V nodes
together with the bundle of His and a portion of the branches were made
and examined by myself. The -sections were from 15 to 17 microns in thickness
and every fifth one was examined. The sinus node was situated between
superior and inferior vena cava, about 2 cm. from the orifice of the former.
It was situated in a sulcus just beneath the endocardium. It measured 5.1
by 10 mm. and was about 1.5 mm. deep. The muscle cells of the node appeared
normal. The connective tissue stroma comprised usually about one third,
sometimes about one half of the nodal tissue. This amount does not appear
in excess of normal. The arteries showed no sclerosis. Beneath the epicardium
was a layer of fat infiltrated in the region of the node with cells of which a
majority were lymphoid, the rest polymorphonuclear. The infiltration did not
extend into the node. Microscopical examination of the sinus node therefore
revealed no abnormality except a subendocardial infiltration which did not
extend to the node itself.

The auriculoventricular node was found in the auricular tissue near the
septal cusp of the mitral valve. It appeared normal as to condition of muscle
fibers, arteries and proportion of connective tissue. Its greatest length was
17 mm., greatest width 7.5 mm., and depth 1 mm. The stem showed no
abnormalities. The right and left branches were traced continuously to a
point lO.S mm. below the beginning of the node. No abnormalities were found.

As to the cause of the A-V rhytlim, three possibilities must be
considered : first, an organic lesion in the conduction .system ; second,
an endogenous factor which failed to produce demonstrable pathologic
lesions, and third, digitalis. The first was ruled out by the examination
of serial sections. The .second remains a possibility, though an unlikely
one. Arrhythmias, heart block, for instance, can occur without digitalis
or pathologic lesions, but the absence of the latter markedly diminishes
the probability of an endogenous cause. Tn the thir<I case, the relation

RICHARDSOX— RHYTHM AND DIGITALIS 259

Clinical Course of Auricl-loventricular Rhythm

Observer

Diagnosis or Pathology

Outcome

Cause of Death

Belski (9)

(1) Acute rheumatic fever

(2) Typhoid fever

Recovery

Recovery, arrhythmia

for 23 days
Lasted 1 day
Recovery, arrhythmia

lasted 4 weeks

lasted 2 days
Recovery, arrhythmia

(4) Scarlet fever

(3) Acute rheumatic fever

Cowan

(1) Acute ulcerative endocarditis:

profound inflammatory dis-
turbance in A-V node

(2) Acute endocarditis; acute in-

flammatory process involv-
ing A-V node

(3) Acute endocarditis: subacute

pericarditis; infiltration of
A-V bundle and node

Died

Endocarditis; pneu-
monia

Fleming and

Kennedy

(lO)

Died

Died

Endocarditis: peri-
carditis

WiUiams and
James (11)

Cardiac arrhythmia: A-V
rhythm

Duration 1 year; im-
proved

Hume (12)

(1) Diphtheria, bronchopneumo-

nia, pleural effusion; A-V
node normal

(2) Diphtheria, S-A node the seat

of an acute inflammation

Died

Diphtheria: bron-
chopneumonia:
pleural effusion

Diphtheria

Died

(1) Carcinoma of stomach; peri-

carditis: arteriosclerosis of
branch of coronary leading
to region of sinus

(2) Syphilitic aortitis; thrombosis

of left coronary artery

acn; pencaroitis;
arteriosclerosis

Aortitis: aneurysm

Discharged from hos-
pital after 11 days

White (15)

Auricular flutter: auricular flbril-
lation: A-V rhythm

Discharged from hos-
pital 19 days after
admission

Recovered after few
days

Left hospital after 12
days

PusseU and

A-V rhythm; paroxysmal tacliy-
cardia

Died

Cardiac decompen-
sation

WoUerth (17)

Presbyterian
Hospital
46935

Chronic myocarditis; cardiac in-
sufficiency; cardiac arrhythmia:
hypertension

Discharged improved
after 5 weeks

Presbyterian

Acute rheumatic fever; acute rheu-
matic endocarditis: chronic car-
diac valvular disease; mitral in-
sufficiency and stenosis; A-V
rhythm resulting in incomplete
A-V dissociation with ventricu-
lar rate in excess of auricular

Discharged from hos-
pital after 1 month,
recovered from acute
illness; electrocar-
diogram normal 1
month after dis-
charge

between the administration of digitalis and the onset of the A-V
rhythm is so striking as to point with considerable emphasis to this
last explanation. Electrocardiograms demonstrated sinus rhythm prior
to a course of digitalis medication and A-V rhythm just after it; on
another occasion A-V rhv-thm just after a course of digitalis and sinus
rhythm after the eflfect had worn off. The clinical and pathologic
observations combine to indicate a causal relation between digitalis
and A-\' rhvthni.

260 ARCHIVES OF JXTERXAL MEDICINE

In neither of our cases did stimulation or depression of the vagus
nerve by pressure on the vagus nerve and atrophin, respectively,
produce any definite efiFect. This is in accord with the findings of
several other observers on the more permanent type of A-V rhj'thm,
and would seem to indicate that the region affected lies within the
heart rather than in the extrinsic cardiac nerves."

SUMMARY

Auriculoventricular rhythm is not in itself fatal, but is frequently
associated with severe infections or severe acute and chronic cardiac
disease.

A case is described in which clinical and pathologic observations
combined to indicate a causal relation between the administration of
digitalis and auriculoventricular rhythm.

I wish to take the opportunity of expressing my thanks to Drs. T. Stuart
Hart and Dr. Warfield T. Longcope for their helpful interest in the work, and to
Miss Rose Richter for her advice and cooperation in the preparation of sections.

17. The following references also bear on this subject :

Belski, A.: Ztschr. f. klin. Med. 67:515, 1909.

Cowan, J.; Fleming, G. B., and Kennedy, A. M. : Lancet 1:277, 1912.

Hume, W. E.: Heart 5:25, 191.3.

Laslett, E. E. : Heart 6:81, 1915.

Neuhof, S.: Arch. Int. Med. 15:169 (Feb.) 1915.

Fussell, M. H., and Wolferth, C. C. : Arch. Int. Med. 26:192 (Aug.) 1920.

A CASE OF DISSEMINATED MILIARY TUBER-
CULOSIS IN A STILL-BORN FETUS*

R. C. WHITMAN and L. W. GREENE

BOULDER, COLO.

Every layman knows that tuberculosis tends to cling to families.
"The consumptive," says Hippocrates, "is born of a consumptive,"
and the search for an explanation of this fact has long engaged
general interest. The earliest and most obvious explanation assumed
an hereditary transmission, but the notion of an inherited tuberculosis,
in a mendelian sense, could not, of course, survive the discovery of
the cause of the disease. Other notions which have enjoyed more
or less vogue at various times are, for example (a), germinal trans-
mission by ovum or spermatozoon, such as is known to occur in silk-
worm pebrine. A certain amount of experimental evidence (which will
be discussed in some detail later), supports this view, but the theory
is open to criticism on various grounds, among others, that tuberculous
infection of either germ cell would probably render it incapable of
function, or, at least, lead to the early death of the embryo, (b) It is
often said that the individual inherits not the disease itself, but a
special predisposition to it. But this assumed predisposition has
remained so vaguely defined as to represent little more than a form
of words. If it is supposed to be a definite mendelian character, it
is open to the same objection as the theory of hereditary transmission
of the disease itself, and if it is assumed to be specific it runs counter
to all we know about immunity ; for immune bodies or toxins which
might find their way from the maternal into the fetal circulation, would
serve to protect the child, by conferring passive immunity or stimulating
the production of an active immunity, respectively, unless, indeed, we
are ready to believe that every infection is an anaphylactic phenomenon.
Only if aggressins alone of the tubercle baccillus should be absorbed
by the fetus would there be a specific lowering of resistance to the
disease. These, being foreign substances, would tend to be eliminated
very rapidly after birth. Moreover, the selective absorption of aggres-
sins alone would probably be the rarest of occurrences. For example.
Coca ' found that anaphylactic sensitiveness is not inheritable in the
true sense; but the mother may transmit specific antibodies from her
blood to that of the fetus through the placenta. Krause ^^ observed

♦From the Henry S. Denison Research Laboratories and the Department of
Pathology, University of Colorado School of Medicine.
1. Coca: J. Immunol. 5:363, 1920.
la. Krause: Johns Hopkins Hosp. Bull. 22:250, 1911.

262 ARCHIVES OF INTERNAL MEDICINE

that such maternal transmission occurs, but is always more or less
irregular and inconstant, and the amount of immunity decreases with
age and weight of the animal ; the '"inheritance" is probably always
one of antibodies. Finally, if the predisposition is merely a non-
specific condition of lowered resistance, it flies in the face of daily
experience. Such patients react to other infections just as those
normally constituted, and often exhibit marked physical and mental
vigor. The stigmata of the tuberculosis candidate can be explained
more rationally by regarding them as interferences with growth due
to an accomplished infection rather than as earmarks of an inherited
constitutional inferiority predisposing to a later infection, (c) Proba-
bly the most widely accepted notion is that the child of a consumptive
family inherits neither the disease itself, nor any peculiar tendency to
it, but merely an increased exposure to infection. There is little doubt
that this is the effective factor in many cases, (d) There is, however,
one other possibility, which forms the topic of the present discussion,
viz., that the child inherits the disease not in a mendelian sense, but
in the sense that constitutional syphilis is inherited ; that is, it is infected
during intrauterine life. Such infection, in the case of syphilis, we
regard as a universal law, in tuberculosis we regard it as rare. What-
ever disparity exists (and the disparity, we are inclined to believe, is
not so great as is generally thought to be the case), may be explained as
being due to the markedly greater invasive tendency of syphilis.

These several explanations of the origin of familial tuberculosis
are discussed at greater length, but in the same general tenor, by
Sitzenfrey - and by Pehu and Chalier.^

We are not concerned, in this paper, with minutiae of jjathologic
anatomy and histogenesis, but only with the epidemiologic, or, as
Warthin and Cowie * so aptly express it, the sociologic and eugenic
significance of prenatal tuberculosis. The cases are now so many that
a detailed enumeration of them has become, in a sense, of as little
practical importance as such an enumeration of appendicitis cases would
be. We have been content to collect without distinction as to source,
the cases listed in the bibliographies of Warthin and Cowie.' Sitzen-
frey,- Pehu and Chalier,"* Lanz,'' Weber ' and Chome,' care being taken
to eliminate duplicates. The distinction drawn by all previous writers
between authentic and non-authentic cases, we have adhered to as

2. Sitzenfrey : Die Lchre von der kongcnitaler Tuhcrkulose, etc., Berlin,
1909.

3. Pehu et Clialier : .'\rcli. d. Med. d. enf. 18:1. 1915.

4. Warthin and Cowie: J. Infect. Dis. 1:140, 1904.

5. Pehu ct Chalier: Arch. d. med. d. enf. 11:1, 100, 1908; 17:921. 1914.

6. Lanz: Arch. f. Gynik. 104:214, 1915.

7. Weber: Brit. J. Child. Dis. 13:321, 359, 1916.

8. Chome: Arch. mens, d'obstr. et de gynec, Paris 7:294, 1918.

U'HITMAX-GREEXE— FETAL TUBERCULOSIS 263

closely as possible, in view of the fact that, naturally enough, there
is no general agreement as to certain of the cases. The four or five
cases listed by Warthin and Cowie in the doubtful class, as "probable"
or "very probable" we have placed in the list of cases accepted as
genuine (where, indeed, they are also listed by others), in the convic-
tion that a truer notion of the actual situation is thus gained. To the
total thus obtained we add with greater detail the small number we
have found, not included in any of the lists above mentioned, and our
own case.

List of Published Cases of Pren-at.\l Tuberculosis

Authentic Doubtful

Congenital tuberculosis of fetus and placenta 38 509

Tubercle bacilli but no histologic changes :

Fetus and placenta 21 10

Fetus only 3

Same in fetus with histologic tubercle in placenta 4

Tuberculosis of placenta, with bacilli and histologic tubercle 44

Bacilli but no histologic tubercle, placenta only 3

Total 113 519

To this total of 113 authentic and 519 more or less doubtful cases
of prenatal tuberculosis we have been able to add a number of cases.

Leuenberger ^ reports two cases. In the first, the mother suffering
from an old tuberculosis of the kidney, ureter and bladder died of acute
miliary tuberculosis and tuberculous meningitis in the eighth or ninth
month of pregnancy. The placenta showed no macroscopic evidence of
tuberculosis, but in sections there were found many typical tubercles,
uniform in size and smaller than those in the maternal organs. Tubercle
bacilli were found by direct examination in the capillaries of the fetal
liver, but there were no histologic tubercles. Guinea-pig inoculation
tests of the fetal blood and liver gave positive results. In the second
case, the mother died of cavernous consumption following spontaneous
abortion. Bacilli but no histologic tubercles were found in the fetal
liver. Guinea-pig tests of the fetal blood were positive ; of the
fetal liver, negative. The placenta likewise contained no histologic
tubercles, but bacilli were found in the intervillous spaces, especially at
the edges of the white infarcts. A very few bacilli could be traced
through the wall into the capillaries of the villi.

Sugai and Monobe,^" using direct search in sections, found tubercle
bacilli in three out of seven placentas from tuberculous mothers. (By
the same method they found leprosy bacilli in all of twelve placentas
from leprous mothers).

9. Leuenberger: Beitr. z. Geburtsh. u. Gynak. 15:456, 1910.

10. Sugai and Monobe : Zentralbl. f. Bakteriol. Orig. 67:232, 1912.

264 ARCHIVES OF IXTERXAL MEDICINE

Chome * reports the case of a mother who died of chronic pul-
monary tuberculosis twenty-four hours after delivery at term. The
infant had no contact with the mother after birth, and was placed
at once in a tuberculosis-free environment. The child was sick from
birth, coughed on the seventh day. and when seven weeks old, enormous
numbers of tubercle bacilli were found in the stools. At eight weeks
it weighed less than at birth. There were rales in the lungs, and
inoculation of guinea-pigs with material from feces gave a positive
result in six weeks. The intradermal tuberculin test was negative.
The child died at three months. At the necropsy there was found
miliary tuberculosis of the lungs, with croupous pneumonia in the
stage of red hepatization. The peribronchial and mesenteric lymph
nodes were large and caseous. The middle ear on both sides contained
pus and tubercle bacilli. The miliary tubercles in the lungs con-
tained incredible numbers of tubercle bacilli. The liver contained
early miliary tubercles with giant cells and a few bacilli. The spleen
was packed with tubercles. The suprarenal medulla was caseous and
contained many bacilli. The placenta was not examined. Chome
thinks that intrapartum infection in this case is excluded by the fact
that the amnion ruptured only at the moment when the head disengaged.
He suggests, as we do, that the lung may be a favorable culture medium
for the tubercle bacillus because of the abundance of atmospheric
oxygen and blood.

In addition to the above we have found the following titles to
which we have not been able to get access : Kiralyfi's ^' article on con-
genital tuberculosis has a Magyar text which we cannot read, nor
have we been able to find anyone who can translate it for us. Lud-
wig's '^ article on tuberculosis of the placenta and congenital tuber-
culosis, and Kerscher's " monograph we have not been able to obtain.
Kerscher's monograph is the only one whose title gives promise of
adding more material.

Our case adds one more clear instance of transplacental infection.
The following is a brief abstract from the protocol of the necropsy,
which was performed Oct. 24. 1919, by Prof. E. R. Mugrage, to
whom we here express our grateful appreciation.

KKPORT OF CASE

The body is that of a still-born Mexican female infant, stated to be in
the ninth month of gestation. (The record in the University Dispensary shows
a nine months' pregnancy.) The body is 45 cm. long and weighs 3,402 gm.
Body heat is still present. There are several scars, oval in shape and some-
what depressed, varying in size from a pinhead to 5 cm. in diameter, scattered

11. Kiralyfi: Orvosi hetil. 49:568, 1905.

12. Ludwig: Gynackol. Helvet. 14:25, 1914.

13. Ktrscher : Kasuistischcr Beitr., etc., Stuttga

IIHITMAX-GKEEXE— FETAL TUBERCULOSIS 265

over the anterior aspect of both shoulders, and on the left cheek. The pleural
cavities show widespread, firm, bandlike adhesions. The lungs sink in water
and are studded with innumerable calcareous nodules. The kidneys are
thickly set with small, grayish white nodules uniformly distributed throughout
the cortex and medulla. The adrenals are much enlarged and the medulla
filled with a large amount of caseous material. Macroscopic lesions are also
present in the left ovary, both lateral ventricles, and in the anterior part of
the thalamus. The other organs do not contain lesions visible to the naked
eye. The fragments of the placenta, which were saved, contain numerous
caseous nodules measuring 4 or 5 mm. in diameter.

Microscopic Examination. — (The structure of the tubercles in the lungs only
is described. The lesions in the other organs are essentially identical.)
The lungs are thickly set with tubercles varying greatly in size and stage of
development. The earlier lesions consist of collections of epithelioid cells.
which are large and circular at the center of the tubercle, becoming more and
more flattened toward the periphery, until at the edge they are long and
spindle shaped. Giant cells are rare but occasionally are found. Lymphocytes
form a zone about the periphery, and some tubercles are completely infiltrated
by small round cells. About the tubercles the air cells are filled with hyaline
coagulated exudate, with a considerable number of polymorphonuclear cells.
The larger tubercles are caseous at their centers. Many calcareous nodules
are also present, probably representing a still later phase of the process.

The spleen and pancreas contain numerous calcified tubercles, and the
suprarenal and ovary contain caseous tubercles. The kidney contains many
tubercles, frequently caseous. After prolonged search through many sections,
a few tubercle bacilli were found in the kidney lesions. They could not be
demonstrated in any other organ. The liver is markedly fatty and contains
an occasional small tubercle. The floor of the lateral ventricle contains an
occasional caseous tubercle. The placenta shows numerous necrotic areas and
tubercles with caseous centers.

It is to he regretted that no animal inoculation tests could be made
owing to the tissue being fixed prematurely ; but the diagnosis is
firmly established by the character of the histologic changes, and the
finding of tubercle bacilli in the kidney. The fact of still-birth at
term precludes intrapartum infection.

History of the Mother. — Mexican, 36 years old, has had five children, four
of whom died of influenza in 1918. Denies miscarriages. The examination
made at the dispensary previous to confinement revealed no active tuberculosis
but one made at the time of delivery by Dr. H. O. Calhoun (to whom we
express our sincere thanks) revealed an area of dulness with rales at the
right apex. Thirteen months later (November. 1920) she is apparently healthy
and well nourished but stilT has a dull area with rales as before. She raises
little or no sputum.

In view of all the difficulties which, except in rare cases, attend the
search for tuberculosis in fetus and placenta, and the fact that they
are rarely subjected to exhaustive examination unless there is some
special reason therefor, this collection of cases is by mere weight of
numbers an imposing one; imposing enough to lift prenatal tuberculosis
out of the class of medical curiosities and to stimulate consideration of
the broader aspects of the question. But if. as is maintained by many,
prenatal tuberculosis is inevitably fatal within a few weeks or months
after birth, the condition becomes again merely an interesting phenonie-

266 ARCHIVES OF IXTERXAL MEDICINE

non of little practical importance. The question of its significance in
human pathology- is intimately bound up with the question as to the
validity of the old Baumgarten theory of latency. By the very nature
of the case, the theory is hardly susceptible of objective proof or dis-
proof, and can only be approached indirectly. We shall review as
briefly as possible the arguments which Baumgarten himself has made
familiar, and submit certain further considerations which seem to us
not devoid of weight.

The possible common portals of entry are (a) the respiratory
tract, by the inspiration of bacilli suspended in dry dust, or, still wet,
in droplets of saliva and mucus, coughed, sneezed or otherwise expelled
from the mouths and noses of consumptive persons; (b) the digestive
tract, by swallowing bacilli reaching the mouth as above, or in food
and drink; (c) the tonsils, similarly. Baumgarten "=" has argued that
in the first place, the real or apparent primary localization of the dis-
ease in the lungs is absolutely no proof of an aerogenic origin. The
relative frequency of pulmonary tuberculosis, as compared to other
localizations, is due merely to a special predilection of the organism for
the lungs. The extraordinarily tortuous character of the path to be
traveled along the bronchi, the defense against entrance afforded by
mucin and cilia, and the very slow progress into the lung of any given
unit of air, probably render the bronchial tree the most difficult of the
routes into the lung. Animals are rarely infected by breathing bacillus
laden dust, perhaps due to attenuation caused by drying. Breathing
moist bacilli in spray produces a higher percentage of infections ; usually
caseous pneumonia or bronchopneumonia, rarely bronchiolitis or per-
bronchiolitis. But the same processes can be produced in typical form
by hematogenous infection. On the whole, the droplet method of
infection, Baumgarten thinks, has more in its favor that the dry
dust method, but this is far from proving that the former is the most
common mode of infection under normal conditions.

Feeding experiments are much more successful, especially when the
bacillus is administered in small doses over a long period, thus simulat-
ing as closely as possible the natural conditions of infection, rather
than in only one or two large doses. Moreover, calves, which never
become infected by breathing infected dust, etc., do so regularly in
the feeding experiment. Necropsies on such animals, performed at
successive stages of the process, prove that the (bovine) bacillus may
pass the intestinal wall without producing any local lesion, and ran be
found in the mesenteric glands, even when these are not enlarged.
Later the bacilli reach the thoracic organs, usually only the peribron-

13a. Baumgarten: Samml. klin. Vortr.. No. 281. 1882; (Ii
I). 1955); Deutsch. med. Wchiisclir. 35:1729, 1909.

WHITMAX-GREENE— FETAL TUBERCULOSIS 267

chial glands, but sometimes the lungs also. Since no direct path from
the mesenteric to the bronchial glands exists, the transfer must be
assumed to take place via the thoracic duct or portal vein.

Baumgarten insists that these facts can not be transferred unhesi-
tatingly to human pathology. Primary tuberculosis of the mesenteric
glands is rare in man, and a primary lesion due to the bovine type is
rarer still. Even when the bovine type does reach the glands it grows
but meagerly and soon dies out. It is rarely found in other lesions,
especially the lungs, and he does not believe that an infection due to
the bovine type could lose this character by conversion of the bacillus
into the human type, since this would require passage of the strain
involved through several human hosts. Human tissue, on the other
hand, is a favorable medium for the human type of bacillus, so that
these objections do not apply to infection with the latter. In the
vast majority of cases of combined pulmonary and intestinal tuber-
culosis the pulmonary lesion is obviously the older. But just as this
fact does not prove aerogenic origin, so involvement of the intestines,
even if very common, would not prove that the intestine serves as
a frequent portal of entry. If that were the case, the bacillus must
be assumed to have passed the intestinal wall and mesenteric glands
without injuring them, and this, he insists, the human tyjie, in contrast
to the bovine type, does not do.

Baumgarten's argument, however, is open to criticism. His
thesis requires that, while postnatal infection through the intestinal tract
is looked upon as unable to reach the lung without leaving indications
of its passage along the route followed, prenatal infection through the
placenta and umbilical vein shall be assumed to be able to do just
that. Schmorl and Kockel " long ago showed that the bacillus is found
more abundantly in the fetal liver than anywhere else in prenatal tuber-
culosis, and often unaccompanied by histologic changes. Conceivably
this might be due to infection havi'ng occurred so short a time before
the examination was inade that no opportunity was afforded for the
customary changes to develop. Be this as it may, there is no question
that lesions in the liver, other than those obviously arising in the last
days or weeks of life, are very rare in pulmonary tuberculosis. Baum-
garten's argument on this point therefore cuts both ways.

To the above brief and very incomplete sketch of Baumgarten's
argument in favor of latent and prenatal infection we would add the
f(jllowing considerations :

1. There is a general tendency to a])[)ly to man the ideas of von
Behring '^ regarding the age period of infection in cattle. Quite

14. Schmorl and Kockel: Beitr. z. path. Anat. u. z. allg. Path. 16:313, 1894.

15. Von Behring: Deutsch. mcd. Wchnschr. 29:689, 1903; 30:193. 1904
abstr. Zentralbl. f. Bakteriol. Refer. 34:136, 729, 1903.

268 ARCHIVES OF INTERSAL MEDICIXE

recently, for example, von Jaksch "^ has complained with some petulance
of a failure in certain quarters to apprehend the "fact" that the human
animal is insusceptible to tuberculous infection (aside from such
purely local manifestations as the anatomist's tubercle) after the age of
six years. Griffith ^' puts it somewhat differently when he defines
tuberculosis as a disease of childhood. But if we accept the theory
that tuberculous infection, either always, or only usually, occurs either
before birth, or during the years of infancy and childhood, we cannot
escape accepting "in principle," as the diplomats say, the theory of
latency ; because in the vast majority of cases, manifest symptoms
develop only later in life, sometimes very late, and often in connection
with some special tax on the vitality, such as measles, whooping cough,
pneumonia, pregnancy, etc. Moreover, if an infection acquired during
the first weeks or months after birth may remain latent for years, it
is hard to- see any valid reason why an infection acquired during the
last weeks or months before birth should not behave in the same way.
Certainly, the struggle for existence is not more difficult or exhausting
for the child in utero than for the child not thus protected.

2. The mechanism of latency may be placed on a more solid theoretic
basis than that provided by Baumgarten's theory that the bacillus
preserves and prolongs its individual life by "slumbering" over long
periods. It has been abundantly proved that immune bodies present in
the mothers' blood may find their way into the fetal circulation (Coca,^
Krause '=• and others). Since these are isoimmune bodies, and therefore
practically identical with any immune bodies which might be produced
by the fetus itself, they would tend to be only very slowly eliminated,
and would no doubt protect the child as completely, and over as long
a period as an active immunity of equal titre. A partial immunity
would tend to prevent active progress of the disease, without, however,
lieing able to destroy the bacillus completely, thus creating a situation
analogous to that which we assume to exist, in order to account for
the progressively narrowing field of activity ordinarily observed in
the later stages of syphilis. Sitzenfrey noticed that the extent of the
disease in fetal tuberculosis is roughly directly proportional to the
extent and activity in the mother. This is probably usually true, though
the cases of Warthin and Cowie, Grulee and Harms,^" and our case
prove that it is not necessarily true. The same rule stated differently
may l)e useful here. Let us put it in this way: the extent and severity
of the disease in the fetus is, roughly, inversely proportional to the
mother's immunity.

16. Von Jaksch: ZcnlralM. f. Inn. Med. 39:543, 1<)1H.

17. Griffith: New York M. J. 109:485, 1919.

19. Gnilce and Harms: Am. J. Dis. Child. 9:.122 r.^prin 1915.

11 HITMAX-GREEXE— FETAL TUBERCULOSIS 269

3. Numerous researches of recent years prove conclusively that
what we term latency is often not latency at all, but defective diagnosis ;
or more often, no diagnosis at all. We shall mention only a few of
the many works in this field. Landouzy -" reports in detail seven cases
of infantile tuberculosis, with necropsy findings, in children varying in
age from 6 weeks to 12 months. These occurred in a total of twenty-
three deaths, or 30.4 per cent., in a hospital population of 127 children,
in a period of 4 months. In another series -^ he reports twenty deaths,
fifteen with necropsy, three without necropsy, and two doubtful, in
children varying in age from 3 to 24 months. Total deaths for the
period were sixty-nine, including eleven incidental to a severe epidemic
of diphtheria. The fifteen deaths with necropsy constitute 20.7 per
cent of the total, and the largest single item in the list of causes of
death. Landouzy, who is a firm believer in the Baumgarten theory of
latency, contends very earnestly that tuberculosis, instead of being rare
in childhood, "as commonly supposed," is the most important cause
of infantile mortality. Klare,^^ in a clinical lecture on tuberculosis
in childhood, maintains that practically every one becomes infected by
the fourteenth year of life. Griffith '' collected data showing that, when
the diagnosis is based on sensitization tests, tuberculosis is rare in the
first three months of life, but increases rapidly up to as high as 83
per cent, of the children examined, at 14 years of age. Under the
more rigid criterion of necropsy, the percentage is naturally somewhat
Je%ver, viz., 40 per cent, in 848 necropsies at Vienna, 35 per cent, in 332
necropsies at Philadelphia, 38 per cent, in 1,675 necropsies at Paris,
distributed as follows :

Pur Cent.

Up to age of 3 monlhs 1.82

From 3 to 6 months ; 18.0

From 6 to 12 months 26.0

From 1 to 2 years 40.0

From 2 to 5 years 60.0

From 5 to 10 years 67.0

From 10 to IS years 71.0

It is to be noted that the failure to find visible tuberculosis in early
infancy does not at all prove that the infection had not already taken
place, as is shown by the large number of cases in the table given
earlier, of infection without histologic changes. Honl,^-' who gave
this condition the name, status bacillaris, suggested that it might be
due to retardation of growth of the bacillus, through lack of free

20. Landouzy: Rev. de med. 7:383, 1887.

21. Landouzy: Rev. de med. 11:721, 1891.

22. Klare: Aertzl. Rundschau 28:67. 7.1 1918.

23. Honl: Bull, intern, de YAcad. de Sc. de rEmpereur Francis Joseph I,
1894 (cited from Warthin and Cowie).

270 ARCHIVES OF IXTERXAL MEDIC IS E

oxygen. Spolverini ~* studied a series of 900 children, varying in age
from 3 to 12 months. Marantic and undergrown children, and chil-
dren ill with, or recently recovered from, other diseases were not
included in the series. The children were first tested by the percu-
taneous tuberculin method and if this was positive, were examined
radioscopically and by the other recognized methods of physical diag-
nosis. Some children negative on the first test became positive when
tested again several months later. In almost every case, when one child
of a family gave positive results, other children in the family were also
found to be positive. Of the 900 children examined, sixty-three, or
seven per cent., were found to be tuberculous. Of the sixty-three
positive cases, eight children were from 3 to 4 months old, twenty-two
from 4 to 6 months, and thirty-three from 6 to 12 months. He thinks
that his results are too low because several of the children were under-
sized, and probably failed to react because of general ill health.

The term, latency, is usually taken as covering all that period in
the progress of the disease from its actual inception to the onset of
manifest disturbance of health. The above figures would indicate
that this period would be materially shortened, and perhaps reduced to
something like a rational period of incubation, if every child, regardless
of the apparent state of its health, could enjoy the benefit of an
exhaustive search for the disease, by the aid of the most refined diag-
nostic methods. Latency as a problem might then disappear.

4. It seems to us also that not enough account is taken of possible
variations in the virulence of the bacillus. In one of Schrumpf 's ^=
cases the bacillus recovered from the placenta and fetal tissues was so
avirulent that it failed to kill a guinea-pig. The case of Repaci,^"
of a child dying on the twenty-ninth day of life, is doubted by Weber
simply because the child hved so long. For the same reason Weber
doubts the authenticity of Hamburger's -" case, in which the child
lived nearly eight weeks. But everyone who has employed animal
tests for the diagnosis of tuberculosis in adults knows that the animals
sometimes live an exasperatingly (when one is waiting to make the
diagnosis) long time. Certainly slow progress of the disease is no
reason for doubts as to its nature.

Whatever the facts may uhimately prove to be. tiiere is already
sound reason for believing that prenatal infection is an ini])ortant, if
not the most important, nictlKid of |)r(i])agaling the disease. Nor is

24. Spolverini: Riv. de clin. Pcdiat. 17:169. 1919; also Tu1)erciilosi 10:239,
1918.

25. Schrumpf: Bcitr. z. path. .Anat. ». z. allg. Path. 42:225. 1907.

26. Kepaci: Osp. d. Bamb. d. Milan 1:147, 191.3 (alistr. Brit. I. Child. Dis.
10:547. 1913; cited from Weber).

27. Hamburger: Beitr. z. Klin. d. Tubcrk. 5:197, 1900 (cited from Weber).

UHITMAX-GREESE— FETAL TUBERCLLOSIS 271

the mother the only possible source of danger. In the case of the
father, animal experimentation and for man statistical studies consti-
tute almost our only weapons for attacking the problem. Friedmann's -*
e.xperiments are classical. He found that when buck rabbits are
injected in one or both vasa deferentia with either the human or bovine
type of bacilli and mated a few weeks later with healthy females,
tubercle bacilli could usually be demonstrated in the seven day embryos.
If the injection was made into the testis instead of the vas, the bacilli
could likewise be found in the seven day embryos, provided that not
too long a time elapsed between the injection and mating. When the
interval was too long (four weeks or more) conception did not take
place. If the injection was made intravenously, the bacilli could some-
times be found in the six day old embryos. When the interval between
injection and mating was four weeks or more, the bacilli were found
less readily but could still be demonstrated in a few instances. When
the difficulties of the search, a veritable hunting for a needle in a
haystack, are taken into account, these positive results become very
significant.

Friedmann also analyzed 983 cases of pulmonary tuDerculosis with
hereditary taint from the records of the second medical clinic of the
University of Berlin. Among the 983 cases were 503, or 51.2 per cent.,
with a history of tuberculosis on the father's side, 323 cases, or 32.8
per cent., with a similar history on the mother's side, and 157 cases,
or 15.9 per cent., with such a history on both sides. He quotes the
paper of Klebs -" who found that in a family with a history extending
back to the middle of the eighteenth century, many of whose members
he had himself known, consumptive males married to healthy females
often got consumptive children. Several times he observed that one
and the same woman bore consumptive children when married to a
consumptive husband, and later, with a healthy husband, bore healthy
children. Forty per cent, of the children of consumptive mothers
and only 4 ])er cent, of those of consumptive fathers were healthy, so
that tuberculosis in the father is ten times as dangerous for the
child as tuberculosis in the mother. Such data are difficult to
harmonize with the view that tuberculosis is always or generally
contracted after birth, tiiat it recurs in families simply because of
greater exposure, or that on the other hand, it is due to any sort
of predisposition.

Finally, Friedmann cites from the literature two cases of "proved"
congenital tuberculosis derived from the father. Sarwey ■"' reported

28. Friedmann: Vircliows Arch. f. patli. .-Xnat. 181:150. 1905.

29. Klebs: Munchen. med. Wchnsclir. 48:129. 1901 (cited from Friedmann).
,30. Sarwey: .-Xrch. f. Gynak. «:162. 1892; al)5tr. .Schmidl's Jalirb. 240:

174. 1893.

272 ARCHIVES OF IXTERXAL MEDICI \E

a case (included in Warthin and Cowie's list as "very doubtful")
from Baumgarten's laboratory, of a still-born hemicephalic monster,
with missed labor, carried over eleven months, in which there was a
tuberculous abscess of the cervical vertebrae. The father had suffered
for a long time with cough and a tough expectoration, and the paternal
grandfather had died of chronic pulmonary tuberculosis. There was
no demonstrable tuberculosis in the mother. Landouzy described the
case of a child who died when 24 hours old of tuberculosis. The
father was a consumptive and the mother quite healthy. Landouzy ^'
reports the following two family histories, which could, he thinks,
be duplicated from the experience of any active practitioner. An
officer fell sick in 1878, in the midst of apparent sound health, with
tuberculosis, which progressed with various complications to his death
in 1888. He was married in 1876 to a "superb young girl," 21 years
old. He had five children ; the first, a boy, born at term in 1876,
developed normally and died at eight months of cholera infantum.
The second, a girl, born before term (seventh or eighth month) in 1878,
died twenty-four hours later in convulsions. This is the case, so far as
we can make out, cited by Friedmann as a proved case of paternal
transmission. Comment seems superfluous. The third child, born at
term in March, 1881, was brought up like the first under the best
possible conditions. It died when 6 months old with classical symp-
toms of tuberculous meningitis. The fourth, a girl, born at term in
February, 1882, also died when 6 months old with symptoms of tuber-
culous meningitis. The fifth, a boy, born at term in 1883, was breast
fed and brought up far from the father, in the open country. At 5
months, he fell sick with a purulent otitis media, which ultimately
e.xtended to the meninges and was diagnosed by a physician, the uncle
of the child, as tuberculous. The mother remains perfectly well fifteen
years later, in spite of five pregnancies in seven years and the sorrows
and trials of her married life. It will be agreed, he says, that if these
children died of tuberculosis, they received the disease from the father,
not from the mother. In another family, four children, all brought
up carefully and fed e.xclusivcly on breast milk, partly from the healthy
mother, partly from a healthy nur.se, died of tuberculosis at ages ranging
from 3 to 12 months. The father, who had no genito-urinary tuber-
culosis, had a pulmonary process in the first or second stage. He
coughed but raised no spuluin. Morcuvt-r three of tiic children lived
removed from the father from liirth. Landouzy nieniions an experi-
ment of his own, similar to one of Friedniann's, in which six out of
sixteen male guinea-pigs, inoculated with tubercle bacilli, begot tulier-
culous offspring.

31. Landouzy: Rev. <le med. 11:411. 1891.

UHITMAX-GREESE-FETAL TUBERCULOSIS 273

The importance of this matter lies in this, thai to the extent that
the conclusions outlined are sound, all our present day methods of
attacking tuberculosis are merely palliative, designed to prolong the
life of the consumptive, but having not the slightest effect on the
source of contagion. A year ago one of us (R. C. W.) heard Dr.
V. C. Vaughan say, in the course of an impromptu after-dinner talk
to a gathering of physicians that thirty years ago he had been wont
to prophesy, with the enthusiasm of unbounded faith, "No tuberculosis
by 1920." Some contemporaries, less sanguine than himself, thought
1950 a safer estimate. "Now." said Dr. Vaughan, "the years have
brought disillusionment and I feel that we shall be lucky if we stamp
out tuberculosis in five hundred years." We would not be too pessi-
mistic, but we believe that five times five hundred years may well find
us just about where we are today, with a falling death rate, due to
an increasing proportion of cures and greater prolongation of life of
the consumptive, but with no very great reduction in the morbidity
rate, unless indeed we change our methods. It will seem harsh and
heartless to many to argue that although it may well be doubted whether
if leprosy had been allowed to spread unchecked, it would ever have
wrought anything like the havoc caused by tuberculosis,^- yet general
fear and the tradition of centuries sanction a degree of harshness in
the control of leprosy, which we would not dream of employing
toward the consumptive. It is worth considering whether our methods
of control in tuberculosis are not in large measure determined by
what we like to call sentiments of compassion, but which really deserve
a much harsher name. It could be argued with a good show of reason
that methods of control based on education, rather than segregation,
constitute a cowardly repudiation of the rights of unborn, uncounted
millions.

CONCLUSIONS

1. I'renaial tuberculosis has ceased to be a mere curio.sity of the
laborator\-, and has become a pressing problem of the sanitarian.

2. The facts at hand are significant enough to command the active
employment of every agency by which further facts may be elicited.

3. To the extent that the spread of the disease is due to prenatal,
rather than postnatal, infection, present methods of control must be
revised and amended, even at the expense of those sentiments of com-
])assion and tolerant forliearance by which our present eft'orts are so
notably handicapped.

i2. II Kings. V. 1 et scq. "Xow Xaamaii. captain of the host of the king
of Syria, was a great man with his master, antl lionorable, because by him tlic
Lord had given deliverance unto Syria: he was also a mighty man in valor;
but he was a leper." etc. Certainly, the picture of this warrior and national
hero, living en famille with wife and servants, hardly suggests that leprosy
inspired the horror three thousand years ago that it ilm-; i(,rl;iy.

A CONVENIENT ELECTRODE FOR EXPERIMENTAL
ELECTROCARDIOGRAPHIC WORK*

CARL S. WILLIAMSON, M.D.

ROCHESTER. MINN.

In undertaking a series of experimental electrocardiographic studies,
we found a large number of animals that were not suitable for our
work. But before the adaptability of an animal could be determined
with the usual experimental electrocardiographic electrode, it was
necessary to anesthetize it and make an incision in the skin in order
properly to insert the electrodes. This procedure requires considerable
time and sometimes results in an infected wound if the animal is not
used at once.

(o) Attachment for lead wire.

(b) Long screw for adjusting diameter of electrode.

(c) Nut for screw b. It is not necessary to have
the adjustinent leads.

To obviate this difficulty we have adopted a very convenient type
of electrode, consisting of an adjustable copper plate with an attach-
ment for the lead wires. The width of the cuff is 1% inches, and the
minimum diameter, after it has been bent roughly to conform to the
shape of the leg, is 1 inch. The copper is flexible, and is readily
adjusted to a large or small animal by means of a thumbscrew.

The coimection of the lead wire to the copper plate is secured by
soldering a copper rivet to the outside of the electrode. Care is taken
to place the solder only around the edge of the rivet; otherwise it would
interfere with conduction. The rivet is then threaded and provided
with lugs to hold the lead wire.

* From the Division of I'".
Foundation.

cntal Surgery and Pathology, Mayo

U'lLLIAMSOX— ELECTROCARDIOGRAPHIC ELECTRODE 275

In using this electrode the animal's leg is shaved and the copper
cuff snugly adjusted by means of the thumbscrew. No especial atten-
tion is given to the cleansing of the skin, although excess dirt should
always be removed.

We have found the instrument to be very adaptable in making
studies on unanesthetized animals. In a large series of experiments
we have not had more than 1,000 ohms resistance, and in all respects
this electrode has been an improvement over the type generally used
in experimental electrocardiographic work.

ARCHIVES OF IXTERXAL MEDICINE

BOOK REVIEW

THE HEART AND THE AORTA: STUDIES IX CLIXICAL PATH-
OLOGY. By H. Vaquez. Profeseur agrege a la Faculte de Medecine de
Paris, Medecin de I'Hospital San-Antoine ; and E. Bordet, Chef de lab-
oratoire Adjoint a la Faculte de Medecine de Paris. Translated from the
Second French Edition by James A. Hoxeij, M.D., and John Macv. M.A.
Pp. 256. 181 illustrations. Yale University Press, New Haven, Conn.

In this book the authors begin with a description and comparison of the
various radiographic methods of examining the heart and the aorta, includ-
ing their personal technic. They consider that there are three reliable
methods, teleradiography, orthodiography and leleradioscopy. The latter two
furnish identical information and are designated as radioscopy of percision.
They point out that teleradiography and radioscopy of percision each have
its advantages and that the association of the two methods gives nearly
perfect results. If. however, only one is to be used the orthodiographic exam-
ination gives more precise information.

In the following chapter the authors describe the normal cardiac shadow
in the frontal and oblique positions and consider variations in the physiological
form of the heart. In subsequent chapters changes produced in the cardiac
shadow by various pathological conditions as chronic valvular diseases, con-
genital defects, affections of the pericardium, aortitis and aneurisms of the
thoracic aorta are illustrated. The last chapter discusses the localization of
war projectiles in the heart and pericardium. The points of the authors are
well illustrated by diagrams.

This volume with its numerous illustrations demonstrates very clearly the
importance of radiographic examination of the heart and aorta. The authors
seem conservative in their claims and have apparently accomplished their
purpose in compiling the work. It is fortunate that Dr. Honeij has translated
it in English. It will further the empkiyment of this valuable method of
examination and should serve as a u.setui reference to the radiologist and
the physician.

Archives of Internal Medicine

CLINICAL STUDIES ON THE RESPIRATION

IX. THE EFFECT OF EXERCISE ON THE METABOLISM, HEART RATE,

AND PULMONARY VENTILATION OF NORMAL SUBJECTS

AND PATIENTS WITH HEART DISEASE *

FRANCIS W. PEABODY, M.D. and CYRUS C. STURGIS, M.D.

WITH THE ASSISTANCE OF

BERTHA I. BARKER and MARGARET N. READ

BOSTON

The investigations described in this paper were undertaken with a
view to obtaining further information concerning dyspnea in patients
with heart disease. Attention has been directed particularly to the
moderate degrees of dyspnea which are incidental to. the life of most
persons in whom cardiac disease is a limiting, but not an incapacitating
lesion, and certain aspects of the respiration and circulation have been
studied while the subjects were performing exercises which did not
exceed in amount or diflfer in kind from what they were accustomed to
in normal life. The problem has been to determine, so far as possible,
the differences in reaction to exercise which account for the fact that
patients with heart disease become short of breath as the result of an
amount of exercise which does not affect normal persons. It seemed to
be important to avoid the complicating factor of muscular fatigue by
selecting a type of test exercise which did not involve muscles which
the subject was unaccustomed to use, and, since patients with heart
disease frequently state that they get out of breath when walking
upstairs, it was considered that stair climbing was the form of exercise
best adapted to the purposes in view. On account of the complex
nature of the observations undertaken, it was necessary for the subject
to remain as nearly as possible in one position so that a stair climbing
treadmill was made use of. Two groups of subjects, of ai)proximately
the same age were studied, one group consisting of eleven normal young
men, and the other of eleven young men with valvular heart disease.
Unfortunately, such fundamental factors as the circulation rate and
minute-volume of cardiac output cannot easily be approached by
accurate experimental methods and our observations were, therefore,
limited to determinations of the rate, depth, and minute-volume of the
respiration, oxygen consumption, carbon dioxid production, and heart
rate, before, during, and after a standard amount of exercise.

From the Medical Clinic of the Peter Bent Brigham Hospital.

278 ARCHIVES OF IXTER.XAL MEDICIXE

EXPERIMENTAL PROCEDURE

The experimental procedure was essentially the same in all cases.
The subject came to the laboratory at about 9 a. m. after having had
his usual breakfast. Electrodes connecting with the electrocardiograph,
which was used for counting the heart rate, were attached to his body,
and he then sat on a chair, resting, for at least one-half hour. After
this a half-mask, covering the nose and mouth, was put in place, care-
fully tested for leaks, and the subject stepped up on the treadmill.
The mask was connected with two rubber tubes of wide bore, one con-
ducting air for inspiration from outdoors, and the other carrying
expired air to large Tissot spirometers. Inspired and expired air were
separated by rubber flap valves of the type used in gas masks and
described by Boothby and Sandiford.^ A pneumograph, which recorded
the respiratory rate on a smoked drum, was adjusted round the chest,
and the electrodes were connected to the wires leading to the electro-
cardiograph. The subject then stood quietly at rest for at least ten
minutes more, after which the actual experiment was begun. Each
experiment consisted of a series of periods in which the subject was
either standing at rest or performing a given amount of exercise. The
first period served as a base line for subsequent observations, and in
the analysis of the results obtained the percentage variations from
the standard resting conditions of this period are reported. In the first
period the subject stood quietly at rest for five minutes, during which
the expired air was collected, the respiration rate was determined with
the pneumograph, and the heart rate was recorded by means of the
electrocardiograph. From the data thus obtained the minute-volume of
air breathed and the average volume per respiration could be calculated,
while analyses of the expired air, with the Haldane portable apparatus,
gave the data for calculation of the oxygen consumption and carbon
dioxid production. Similar observations were made in each of the
subsequent periods. The second period followed the first after about
two minutes, during which the necessary technical adjustments were
made. In this period the treadmill was started and the subject walked
sixty steps up.stairs in one minute. The height of each step of the
treadmill was 18 cm. At the end of the second period the expired air
was immediately shunted into another spirometer and observations were
continued without any interval through the third period, during which
the subject stood at rest for five minutes, and again through a fourth
period of five minutes at rest. During the fourth period all the factors
under .study were usually essentially the same as they had been before
the walk. During the second, third and fourth periods additional
observations were made on the volume of air expired for each fifteen

1. Boothby, W. M., and Sandiford, I.: Laboratory Manual of the Tcchnic
of Basal Metabolic Rate Determinations, Philadelphia, 1920.

PEABODY-STURGIS—STUDIES O.V RESPIRATION 279

seconds and this proved to be of particular interest in the first minute
immediately after the walk. Preliminary experiments showed that,
while the amount of exercise in Period II was sufficient to produce
dyspnea in the cardiac patients, it did not cause noticeable shortness of
breath in the normal subjects. It was considered desirable, however, to
make observations on normal persons when they were slightly short
(if breath and in a condition analogous to that of the cardiac patients
at the end of Period II, and this was accomplished without altering
the length of walk or number of steps taken, by having them repeat the
walk of sixty steps upstairs in one minute while carrying a load consist-

p

I

II

II

IV

E

1

2

3

4

5

1

2

5.

4

5

21

,

20

/

\

15>

/

\

t6

, \

17

/

\ .

16

/

\

\

15

/

\

\

14

\

\

'5

\

,

1

12

\

\

t1

\

\

►^.

10

c,

V

N

/

k-<

9

)

\

6

\

>

7

N

r^

>-^

^

>-(

>

6

ir

The average ventilation (in liters) per minute of eleven normal individuals
and eleven cardiac patients. The figures at the left represent liters. P, period;
I standing at rest five minutes, II walking upstairs on a treadmill for one
minute, period III standing at rest for five minutes, period IV a second rest
period of five minutes. In periods III and IV the readings are recorded for
each minute of the period. N, curve of eleven normal individuals ; C, curve of
eleven cardiac patients ; E, point at which exercise began. In the curve of the
eleven cardiac patients it was impossible to average the minute-volumes in the
fifth minute of the third period, as in five instances the capacity of the
spirometer did not permit the expired air to be collected during the fifth minute.

ing of a knapsack weighing 50 pounds on the back. This produced
in the normals a degree of dyspnea fairly comparable to that of the
cardiac patients in Period II. In the normal subjects, therefore, Period
I\^ was followed by Period V, in which a 50 pound load was carried
sixty steps upstairs in one minute, and then by Periods VI and VII,
each five minutes long, in which the subjects stood quietly at rest.

280

ARCHirES OF IXTERXAL MEDICI. \E

In reporting results all volumes of air have been reduced to 0 C.
and 760 mm. barometric pressure.

The Subjects — The eleven normal subjects were healthy medical
students and doctors. None of them was in particularly good physical
training at the time the observations were made, but they represent aver-
age specimens of young manhood. The eleven subjects with heart disease

Vital C.\PACiTy of Normal Subjects and Cardiac P.\tients

Vital

Capac-

Surface

Vital

ity.t

No.

Normal

Age

Height,

Weight,

Area,*

Capacity,

per

Diagnosis

Subjects

Cm.

Kg.

Sq.M.

C.c.

Cent.
Normal

1

J. D. T.

31

174.2

86.3

2.01

5,000

100

P. F-S.

25

179.2

69.9

1.88

4.200

90

3

L. L.

179.3

70.4

1.88

4,475

95

H. B.

176.6

71.1

1.87

4,900

106

5

P. W. P.

39

175.2

71.7

1.86

4,400

95

6

P. B. S.

25

172.2

71.3

1.84

4.700

102

7

W. E.

2B

168.8

69.9

1.80

4.600

102

8

I. C. S.

27

170.0

67.7

. 1.78

4.075

92

9

.7. T. L.

24

176.2

61.8

1.76

4,850

110

10

H. R. M.

25

170.0

65.8

1.75

4,500

102

11

J. W.

24

179.8

54.1

1.69

5,100

121

Cardiac

~

Patients

'

M. McG.

26

182.2

77.1

1.98

3,550

72

Aortic insufficiency; mit-
ral stenosis and Insuffi-
ciency: chronic bron-

2

F. D.

36

172.2

75.7

1.89

3,900

83

3

E.G.

24

180.2

68.5

1.88

4,500

96

Aortic insufficiency: mit-
tral stenosis

4

L. I.

25

179.4

64.0

1.82

4,100

90

Aortic and mitral insuffl-

5

P. C.

17

173.4

66.7

1.79

3,600

81

Mitral insufficiency

6

M. B

25

174.3

57.2

1.68

3,650

87

Pulmonary stenosis

7

C. S.

20

164.7

53.1

1.57

84

Aortic insufficiency; mit-
ral stenosis and insuffl-

8

BO

25

154.0

58.1

1.56

3,400

87

Aortic^insufflcicncy; mit-
ral stenosis and insuffi-
ciency

Aortic insufficiency and

9

A. V.

24

156.8

55.3

1.54

3,650

95

(?) stenosis; mitral In-

sufficiency
Aortic and mitral insuffi-

10

E. W.

33

165.9

49.9

1.53

2.375

62

ciency; (?) mitral steno-

sis

11

F.K.

25

160.6

48.5

1.49

3,100

83

Aortic Insufficiency; mit-
ral stenosis and insuffi-
ciency

all had valvular lesions but the degree of disability varied greatly from
ca.se to case. In some of them dyspnea was scarcely a more prominent
symptom than it is in many normal persons, while in otliers it caused
marked limitation of physical exertion. In conformity with previous
observations ^ it was found that the degree of tendency to dyspnea was
indicated with considerable accuracy by the vital capacity of the lungs,

2. Pcal.i
1917.

.dy, F. W.

We

rth. J. A.: .'\rch. Int. Med. 20:443 (Oct.)

PEABODY-STURGISSTUDIES OX RESPIRATIOX 281

and that the subjects in whom the vital capacity was lowest became
most short of breath while walking on the treadmill. With three
exceptions the vital capacity of the cardiac patients was below that of
all the normals when calculated according to the standards based on
body surface area suggested by W'est.^ All but one ot the patients with
heart disease were discharged soldiers and sailors who were ambulatory
patients at a Public Health Service hospital, and they were in good
general condition, except for their cardiac lesions. They represent
very nearly the same age group as the normals, but were on the whole
slightly smaller men. The average surface area * of the group of
normal subjects was 1.83 square meters, while that of the cardiac
patients was 1.70 square meters.

Oxygen Consumption. — As an index of the total metabolism of the
body the oxygen consumption per square meter of body surface area
was determined, and the results are shown in Table 2. In addition to
the actual oxygen consumption in cubic centimeters there is also given
the percentage variation from the figure in Period I, during which the
subject stood at complete rest. This is taken as a base line and con-
sidered as 100. In e.xamining the figures in Period I for the two groups
it is noticeable that the oxygen consumption of the cardiac patients
was slightly more than that of the normals, the average values for each
of the two groups being 190 c.c. and 170 c.c, respectively, per square
meter. In eight of the cardiac patients the oxygen consumption was
over 185 c.c. per square meter, while it was above this figure in only
two normals. This difTerence in the average oxygen consumption of
the two groups, while standing at rest, amounts to about 12 per cent. It
is quite possible that this small difference has no ^significance and that
it depends on the fact that so limited a number of individuals has been
studied, but since this is not certain it is of interest to consider the
possible explanations of an increased metabolism in the cardiac patients.
Previous observations = have shown that the basal metabolism, at
complete rest and fasting, of patients with heart disea.se may be elevated
considerably if they are severely decompensated, but that it is within
normal limits in persons whose circulation is as well compensated as it
was in the subjects of these investigations. Unfortunately, it was not
practicable to obtain observations on the metabolism of these individuals
when at complete rest and in a fasting state. It seemed possible that
in some of the cardiac patients the slightly high metabolism in Period I
might be due to nervousness, but in others there was absolutely no
reason for considering such a factor, and it is conceivable that the

3. West, H. F.: Arch. Int. Med. 25:306 (March) 1920.

4. DuBois, D., and DuBois, E. F.: Arch. Int. Med. 15:868 (July) 1915.

5. Peabody, F. W. ; Wentworth, J. A., and Barker, B. I.: Arch. Int. Med.
20:468 (Oct.) 1917.

lie

SSgS££SS£g8

gSSSiSSSSSS.t g

|Sg|S igsggs I

liSis SiiiSjl

iLlI

.5 £11

l.«l

I
iliiigiisii I

S§i§iiiSlii

§SSScgSSSgg

>i°-|'SSg|2£SiisS

II

^il

iSSiSSgSSSSSgS

i§igiiiiiii

iisiiigiisg

88888888888

|g£gSgsa8gs

ssssagssss;

siiiisi§Ssi

SSSSSKSS88S

isiisisisfei

111

gsgligSeii

iiiiiiisii^

sg8|ge|ggs|

§1"
111

PEABODY-STURGIS— STUDIES OX RESPIRATION 283

exertion of standing produces a relatively greater increase of metabolism
in patients with heart disease than it does in normal persons. It is
also possible that the somewhat greater activity of the respiratory
muscles, which will be discussed later, accounts for the slightly larger
oxygen consumption of the patients with cardiac disease.

The actual oxygen consumption during the walk upstairs (Period
11) was also greater in the group of cardiac patients than in the normals,
but the average percentage increase in oxygen consumption over that
in Period I was somewhat less in the case of patients with heart disease.
This difference may be partly due to error inherent in using so short
a period, but it was greatly accentuated by subject F. F-S., who had
a very low oxygen consumption at rest, and an extraordinarily large
percentage increase with exertion. The averages are somewhat mis-
leading, for ten normals had a percentage increase of between 218 and
307, while the cardiac patients had increases between 218 and 294.
The difference between the two groups is thus not great. In several
instances it was not possible to obtain strictly comparable results for
Period III because the minute-volumes of some of the patients with
heart disease were so large that the spirometer became filled and the
period was only four minutes, or, in one instance, three and one-half
minutes long, but. in general, the return to normal on the part of the
metabolism, at least as far as can be determined from periods of this
length, seemed to proceed at about the same rate in the two groups of
subjects. In Period IV the oxygen consumption had returned to
approximately what it was before exercise in both groups. It is clear,
therefore, that when normal subjects and cardiac patients undertake
similar amounts of physical exercise, there is little or no difference in
the effect of the exercise on the metabolism in the two groups.

The increase of metabolism among the normals was much greater
in Period V, when they walked upstairs with a load on the back, but
the return to resting condition was quite rapid and was complete in
Period VII.

It is, of course, quite obvious that the periods of exercise, which
lasted only one minute, were too short to give accurate information
as to the actual oxygen consumption resulting from the work done, and
a considerable proportion of the increased consumption fell in Period
III, but the results are comparable in the two groups and they do not
show any striking difference in behavior.

Carbon Dioxid Production.— In Table 3 the figures for carbon
dioxid production in each period are reported. They follow closely
the changes already mentioned for oxygen consumption, but it is
worthy of note that the return to resting conditions after exercise was
less complete in the case of the carbon dioxid production.

Il

^lll

sgssis^iiii

s

co£o

siggssigaii

§

ifll

siisg ;Ssiia

i

isiii'iisii

i

1
If

^

iSiiiiiiaii

§

:

llSllsSsilS

1

>
il

lip

SSSi§iS8|88

1

2|SSSSS||g|

1
2

ill

P2gSg55!8gSS3

s

i§llSl2§sli

S

1

rtP

ligSSISSigS

S

issi^sssiii

1

igigiiisgii

i

lllsil^sgli

«

Ss

il

siiiiiiiisg

i

iiiagiggggs

s

HP

§SI§§iia«g§

i

g55§iiii§ii

s

|i

il

8§8|8|8S888

§

§||8|88SSii

i

11

SiSSgSgSSSS

3

g|liiSS|sss

%

2

I
1

iiJj

4p^'adP3-<K*

1
•<

i
<

III

III

la?

=-,

1

1

•So

sisilsiiiisi iiiii---

f*n

Is

1

sssssasssgg

o

_

2v.

sssas isssss

p

Is

3^

(^

1

SSSSS :gS5S3SS

o

J

II

SSSSSSS^SS?:?

o*

i

If

1

siiiiiiisii

1^

s

n

gg3S5SgS2S5

fesissassssl

t

8i

mM»««t.5M-»i.m=c

i'

I

SSSSS8ss,tr:ss

gsisssssassi

o

11

SSSSSSSffSoSSS

iissssasgfg

s„

|5

-°l

s

u

SSSSSSSS^gg

i|i53Ss8S3l?

5

il

sss2:s;;23sgs

ssssssgsssa

1

liiSiiissii

gsggigS25see

o

Ctj

s«

S3gSSS52SSS

mMmm

-

oS

^n^

a '

S

ggljSSSSUSSS

?8SSg53SS2S2

o

: ; ; ; i

1

IMM;

«

{

S2^

;»;»
«&«''

"^«&

««i6i^^

1

►^'a-V

a'fa&iE:

_;>:;Mr;

a»:a$>ji^

adaj-iiw

•:

286

ARCHirES OF IXTERXAL MEDICINE

Table 4 gives the percentage of oxygen absorbed from the inspired
air and the percentage of carbon dioxid excreted in the expired air.
It will be seen that the figures for both are lower in the group of
cardiac patients than in the normal subjects. The average percentage
of oxygen absorbed by the normals while at rest was 4.36 and during
exercise was 5.57, while the corresponding figures for the patients with
heart disease were 3.60 and 4.51. These results are of special interest
in connection with the observations on the minute-volume of pulmonary
ventilation.

TABLE 5.-

-Respiratory Quotients.

Xonnal
Subjects

Period I
Rest

Period 11
Exercise

Period III
Rest

Period IV
Rest

Period V
Exercise Per

with I

Load

odVI

Jest

Period VII
Rest

J.D.T

F F-S

0.69
0.84
0.83
0.79
0.90
0.68
0.83
0.87
0.84
0.86
0.78

0.85

0.74

O.fiS

0.67

0.76 .

0.69

0.66

o!74
0.68
0.78

0^93
0.88
0.60
0.94
0.65
0.87
0.91

0!88

0.78

o:s7

0.89
0.87
0.95
0.82
0.65
0.92
0.88
0.87
0.80

0.83
0.77
0.91
0.78
0.79
0.74
0.71
0.85
0.77
0.73
0.74

.10
).92
.96
).95

;"*

).93
).98
.00
.04
).90

0.96

i.i :..::::...:.

F. W. P

F.R.S

0.91 ■

I c s

094

J. T.L. .;::::::::

H.R.M

J. W.

0.89
0.92
0 86

Average

0.S3

0.73

0.87

0.88

0.78

).98

0.90

Cardiac PatieDts
F.D. ..'.['.'.'.'.'.'.'.'.

0.85
0.91
0.98
0.81
0.81
0.S3
0.79
0.82
0.84
0.88
0.82

0.80
0.79
0.86
0.74
0.72
0.77
0.79
0.78
0.74
0.97
0.67

0.96t
1.031
l.lOt
0.90
0.82
0.93t
0.91
0.86
0.95

i.m

0.96

0.90
1.05
0.99!
0.84
0.84
0.88
0.84
0.84
0.87
0.91
l.OOTI

...

M B

A V

E W

Average

0.85

0.78

«

0.90

....

* Gas analysis unsatisfactory in this period.
+ Four minute period.

tFour minute and 30 second period.
S Three minute and SO second period.
fi Four minute and 45 second period.

# .Average could not be calculated

in length o( periods.

Respiratory Quotients. — The respiratory quotients (Table 5) are
approximately similar for the two gfoups of subjects in the first rest
period. The figures for F. R. S. are low and probably should not be
taken into consideration as there would seem to be some technical error
involved. In Period II the exercise was associated with a lowering of
the quotients in nine of the eleven subjects in each group. This is
contrary to the results obtained by Krogh and Lindhard * who found a
sudden rise in the respiratory quotient at the onset of hard work. Krogh
and Lindhard called attention to the fact that this does not represent an
altered metabolism but merely a disturbance of the balance between

6. Krogh, A., and Lindhard, J.: J. Physiol. 47:112. 1913.

PEABODY-STURGIS— STUDIES OX RESPIRATION 287

ventilation and blood flow, with a proportionally greater increase of
pulmonary ventilation. The difference in findings may be due to the
very mild character of the exercise performed by the subjects of this
investigation. The low quotients during the exercise period indicate
an inadequate excretion of carbon dioxid, and this was compensated
for by the washing out of carbon dioxid after the cessation of exercise,
as is shown by the general tendency to a rise in respiratory quotient
in Period III. The same lowering of quotient during exercise and rise
of quotient immediately after exercise was seen in the normal subjects
in Periods V and VL

Minute-Volume of Pulmonary Ventihtion. — The actual volumes of
pulmonary ventilation are given in Table 6, and it is seen that the
figures for the first period of standing at rest (Period I) are consider-
ably higher in the group of cardiac patients than in the group of
normal subjects. On account of the fact that the cardiac patients were,
in general, smaller men than the normal subjects, this difference becomes
even more evident when the minute-volume is calculated per square
meter of body surface area (Table 7). Nine of the patients with heart
disease had a minute-volume of between 4.61 and 6.20 liters per square
meter surface area, while eight of the normals had a minute-volume of
between 3.63 and 4.33 liters per square meter surface area. The average
minute-volume of the cardiac patients was 5.37 liters per square meter
surface area, or 37 per cent, higher than the average minute-volume of
the normals, which was only 3.90 liters per square meter surface area.
The difference between the minute-volumes of the two groups, while
standing at rest, was thus much greater than the difference in oxygen
consumption which amounted to only 12 per cent. This fact, which is
of considerable importance, will be discussed later.

As the result of walking upstairs (Period II) the minute-volume
increased in both groups of subjects. It reached a considerably higher
level in the patients with heart disease than it did in the normals, but
it is striking that the percentage rise over the resting minute-volume
(Period I) was approximately the same in the two groups.

In the study of the minute-volume of pulmonary ventilation the
most important points are to be gathered from a careful analysis of
Period III, the first rest period after the walk upstairs. Practically
every subject in 'both groups stated that he was more conscious of his
breathing, or felt more dyspneic, immediately after he stopped walking
than he did during the actual exercise. In the light of this statement
it is exceedingly interesting to discover that the minute-volume was
almost invariably higher in the first minute of rest than it was during
the exercise period. This is shown in Tables 6 and 7, in which Period
Ila represents the first minute of rest, and also in the chart, in which the
minute-volume is plotted for each separate minute in Periods II, III

II

II

Alii

mi

gssssssssss

gg||||Sg2SS

?sgggsgssg

SissSiiiiiS

gssgsssgggg

§sgSgS35SS3

gesgsggssss

gS2S88gSx?gg

iisisgsii?

8888S8S8S8S

H;6."4Bii:s;e:K:-;K-:

fegfssgssass

SSSSSSSSSSBS

jiissiggsii

ilSsiiiiSss

sssfsjiiigguss

§Sgg§Sills2

SSgS88SgS8ag

88888888888

lip

•r'S'O g

Sis';

■E-

o|il

° a'°S
Eg|E

sg4

^iii

5SS§

SSS|S§§

1

Si55!§5sB§

i

<

<

>
II

il

gg|||S|g|S|

i

s

5

Slii^isisli

I

1

1

il

llSSSIiiiSS

1

!

1

5

iiiiiiiP

2i

11

M
Si

§SiiiglSSii

§

1

i

1

||SSgg|||.||

S

t^

>

l§SS3S^S2IS

2SSHgS38Sg§|2

<

s

5

Ss!'S!3Ss3l3

s

silSiSsSill

>

i'

l|l

liSSSI3ss§3

S

ggsSSSBggii

><
<

Hi

i

s

liiiiiSisis

2

iipspaili

r

5

i«i

isiiSss§i§i

i

aSgigSissSs

s

a

1

5

liiPiiii

3

%

Eg

ii

igisgg5r!g§5

i

isiig'Sssigs

1

S

2

iis^iiiSilS

i

mmmn

1

3

¥

^iii

li§|SS8||||

1

li§li§8§|8|

1

<

1

^iiSiiSsiii

8

ssiaiissiii

§

:

i

<

JdOSa

I
1

'

!si

iS5|
ogll

Sa|a

290 ARCHIVES OF IXTERXAL MEDICIXE

and IV. Determinations of minute-volume for shorter periods of time
are, of course, less significant, but it is noteworthy that the pulmonary
ventilation was greater for the first fifteen seconds of rest than for the
last fifteen seconds of exercise in six normals and in five cardiac
patients. The greatest subjective dyspnea thus corresponds closely
with the highest minute-volume of pulmonary ventilation.

Immediately after the first minute of rest, in which the pulmonary
ventilation reached its highest peak, there was a rapid recovery with a
drop in the minute-volume from this maximum toward the level at
which it was before the exercise was begun. There is a good deal of
variation in the rate at which the minute-volume falls in different
individuals, but in the normal subjects it generally reached to within
about a liter of the original minute- volume during the fourth minute
of rest. In the case of the patients with heart disease the decrease in
the minute-volume after exercise was usually less rapid. This is best
shown in the chart. It is impossible to determine just when the minute-
volume of this group tended to return to what it was before the exercise,
because in five instances the collection of air in Period III was not made
in the fifth minute, but during the fourth minute the average minute-
volume was two liters above the resting level. The cardiac patients
thus took about one minute longer than the normals to return to
approximately the same pulmonary ventilation that they had before the
beginning of exercise. Even then, however, the return was less complete
in the group of patients with heart disease than it was in the group of
normals, for during Period IV seven of the former still had a minute-
volume which was relatively higher than that of any of the normal
subjects.

When walking upstairs with a pack (Period \') the normals began
to experience a degree of dyspnea which was comparable to that
observed in the cardiac patients in Period II. It is extremely interest-
ing in this connection that the actual pulmonary ventilation per square
meter surface area averaged almost exactly the same for the normals
in Period V as for the patients with heart disease in Period II, although
the increase over the ventilation at rest was proportionately much
greater. In this second series of experiments the subjects again stated
that dyspnea became more marked after they stopped their exercise.
Period Va in Tables 6 and 7 gives the figures for the minute-volume
during the first minute after the exercise. Just as in the case of the
first walk (Period II), it was found that the minute-volume of air
breathed was almost invariably greater in the first minute after exercise
than it was during the exercise itself, and the most marked subjective
dyspnea corresponded in time with the highest minute-voluume of
pulmonary ventilation.

PEABODY-STURGISSTUDIES OX RESPIRATIOX 291

Rate of Respiration. — The observations on the rate of respiration in
Table 8 show that it was distinctly higher in the patients with heart
disease than in the normal subjects. This point was noticed only after
the experiments on the normals had been completed, and there was no
opportunity to obtain comparative observations on the rate of respira-
tion at complete rest, but in eight cardiac patients pneumographic
records w-ere made after they had been lying on the bed absolutely
quietly for at least one-half hour, and the rate was found to vary
between 14 and 21 per minute. In all subjects a record of the rate
of breathing was taken while they were sitting on a chair after having
been at rest for at least one-half hour and before the mask was put on.
The rate of respiration of the normals was between 12 and 20 per
minute, and in eight subjects it was between 12 and 17, while in all
the patients with heart disease it was between 17 and 26 per minute.
The same difference holds with regard to the respiration rates while the
subjects were standing at rest on the treadmill (Period I). In the
group of normals the rate varied between 7.80 and 17.60 per minute,
being below 15 in eight instances, while in the cardiac patients the rate
was between 13 and 27.80 per minute, and was below 19 in only four
instances. The fact that the rate of breathing was usually slower
while standing in Period I than when sitting at rest resulted from
wearing the face mask. During the walk upstairs (Period II) the
rate of breathing increased in both groups, and the actual rate became
considerably higher in the cardiac patients, but it is interesting that
the average percentage increase over the average rate before exercise
was almost exactly the same in the two groups. The fall in the rate
of breathing, after the cessation of exercise, also appears to proceed
in a similar manner in both sets of observations, and it was essentially
complete in Period I\'.

During the second exercise period (Period V), when the pack was
carried, the rate of respiration of the normal subjects rose still further,
reaching an average of 21.20 per minute, but this was followed by a
rapid drop so that in Period \TI the rate had returned to nearly what
it was before the exercise.

Volume per Respiration. — Table 9 gives the average volume per
respiration for each period in the two groups of subjects. This was
obtained by dividing the total volume of air collected during the period
by the rate of respiration. It is immediately apparent that the patients
with heart disease breathed less deeply than the normal subjects when
standing at rest (Period I). The average volume per respiration for
the normals was 584 c.c. and for the cardiac patients it was 467 c.c,
a difference of about 25 per cent. The variation was considerable in
both groups, running from 412 to 870 c.c. among the normals, and from
320 to 683 c.c. among the patients with heart disease, but in eight of

» .

nil.

ggSSggggESg

-'

>„

g>

11

. s

Oe>f«lMO©«!S!OtO

«^i

sssssgs^sisa

"

^S.i: =

gS2g??SSSSgS

c!

:::::::::::

>

-i>-

II

.,2

£

1^1

>5

rfll

|g|g|gg||g|

1

n"-2

«> £

r-l

fc

^.|

i:?i5SSi22gSS

S

fi

2;

Si '

$SSSS8S?8SS?S

s

^issssagsgss

S

t>

P

!>*"

o £

^

^

5k.5
S

"

SSSSSSSSSSS

«

H

1 1

<

^^i?

SSSgi:SS?S§2g

s

SS8S5SSgS?SSS

\

c

^g^=

"

00

f

o 2

„.o...».»oo

KKS«.^?r=,«*^^

^s.°

12

16

22
19
11

16
13

15

sssssssassg

•

<

H

o

1

i|?«R

SSSg3?S§5SS

S

SSSSSSSS5SS

ss

S-S=:s

IJ

sasasssssss

sgs&sssssss

s

v"-rS

S88SS8SS88S

S

8888888S88S

s

■?1S

-.^•^

^«

«

&^

SSSSgSft"-S2

S

S?lSSS?iS8S2S

8

II

<0-»«a0O00O(N«j-.c<l

-•»-*ioow»«aoiNiO

«

ssssasgssss

S

ssasssssssg

s

s • •

i|

; ; •• ;

?;„

!: : ^

i

^1

t^to

:<^«

r^^jS :

.5cJ :

s

oi^«&KWo^;«6i >

lldcJMoeSoid^tiM >

►ifcH

n^>

^

^'iW'i

^Sl^

a^j^

aci

B<

tL^ph

Si

sll

IP

Has
III

ooSS

illi

•ISBB

li

ii

gss=sgsse8«

«

1^

sSiiliSgili

i

ii

iiii

i

>

^iis

S||S|Sg|82g

S

%>"

Id

1

'

>
1
1

^iJi

ii§lis§IBsS

S

j^

IIPIIIIH!-

I

n

II
11

^1

S|S||RSg8£i

s

■<

.1=

>

s.ii8|Siiiig:

i

1

«

m

ISSS8S8S8SS

i

sagsffisssiss

i

1

1"

gliSI'lliiiS

1

§iii5iisiil

g

i
>

ii

.Si,

£=SSgSSS2S=

2

sSgSi5S|as3s

1
1

l3

|iiii§siisi

g

liiiilssiiE

t

5

1

il

iESslg^SSiS

s

igg=Hiig||

1

1^

>

lig;iiji|i.i.i

p|igisgg|§

i

1

Ifil

S^ggSSfSSSS

,

aSBIIgsSSis

3'

>

mi%%nm

U

iiM§§iiiiiii

i

is

IS.

Ifil

|||§|8§|8||

s

|§|8§|||8||

1

fc

iisiiSiSiig

g

Iii-ii5i§gii

S

.

5 =

if

:

i

«5

i

<

£gg^

ill

Iiii

294 ARCHIVES OF IXTERXAL MEDICIXE

the normal subjects the volume per respiration was between 501 and
870 c.c, while in nine of the cardiac patients it was between 320 and
529 c.c. The difference in volume per respiration between the two
groups depends, in part, on the fact that the normals were the larger
men, but this does not completely explain it, for, when the volume per
respiration was calculated with relation to the size of the subject, it
was found that the average for the normal group was 317 c.c. per square
meter of body surface area, and for the cardiac patients 274 c.c. per
square meter. In eight of the normals the volume per respiration was
between 278 and 483 c.c. per square meter of body surface area, and in
ten of the cardiac patients it was between 215 and 283 c.c. per square
meter.

In the period of exercise (Period II) the depth of respiration
became greater in both groups of subjects. The actual volumes were
considerably larger among the normals than among the cardiac patients,
but, as was the case with the minute-volume and rate of respiration,
the percentage increase of the volume per respiration over that before
exercise was approximately the same in the two series of experiments.
Immediately after the cessation of exercise, when the pulmonary venti-
lation was greatest (Period Ila), the depth of breathing increased in
all the subjects, but the increase was proportionally more in the normal
subjects than in the patients with heart disease. The actual tendency
to more shallow breathing on the part of the cardiac patients was
shown by the fact that at this period the volume per respiration was
between 540 and 850 c.c. in seven of the eleven subjects, while in ten
of the eleven normals it was more than 900 c.c

During the rest period after exercise (Period III), there was a
rapid fall in the volume per respiration which proceeded at about the
same rate in the two groups, and in the second rest period (Period IV)
the depth of respiration in both groups was very nearly the same as it
was before the exercise began.

In Period V, in which the normal subjects walked with the load, the
volume per respiration was almost constantly a little greater than in
the first exercise period, and it increased considerably in the first minute
of rest after exercise (Period Va), but decreased again rapidly during
the rest after exercise and was practically normal dunng Period \TI.

An analysis of the factors of the pulmonary ventilation which have
been studied shows that there is a fairly definite relationship between
the rate of respiration and the minute-volume of respiration, in that
there is a general tendency, to which there are individual exceptions,
for the higher minute-volume to be associated with the more rapid rates
of breathing. In a similar manner, the higher minute-volumes are
usually accompanied by more shallow respirations. This general asso-
ciation between high minute-volume and rapid, shallow respiration is

PEABODV-STURGIS—STUDIES OX RESPIRATIOX 295

probably explained by the fact that with shallow breathing the "dead
space" of the upper respiratory tract becomes a more important factor
since a relatively larger portion of each inspiration remains in it and
takes no part in gaseous exchange. In order to maintain the necessary
volume of alveolar ventilation it is, therefore, essential that the rate,
and consequently the total minute-volume of respiration, shall be
increased. There was no definite relationship between the volume per
respiration and the vital capacity of the lungs in the patients with heart
disease studied in these experiments. This is not at all surprising since
the vital capacities were on the whole only slightly less than normal and
the exercise was not severe enough to require deeper respiration than
could easily be produced.

Heart Rate. — In an investigation of dyspnea in heart disease it
would, of course, have been highly desirable to obtain information
regarding the output of the heart, but owing to the lack of any suitable
method it was necessary to limit the observations to determinations of
the heart rate. In how far the heart rate and the minute-volume of
cardiac output run parallel in such cases as are under consideration is
a problem which cannot be settled definitely on the basis of our present
knowledge, but the relation between the two is probably close enough
to warrant a careful study of the cardiac rate. Since it is almost
impossible to count the heart rate accurately by ordinary methods when
subjects are walking on the treadmill, recourse was had to the electro-
cardiograph, which has the great advantage of enabling one to obtain
continuous graphic records. By the use of two electrodes applied firmly
to the sides of the body just below the axillae and a third electrode,
which was grounded, placed just above the pubis, it was possible to
obtain satisfactory records in almost every instance. Short electro-
cardiographic records were taken after the subject had been sitting
quietly for one-half hour and again when he was standing on the tread-
mill (Period I). Twenty seconds before the exercise was begun a con-
tinuous electrocardiographic record was started which ran through the
exercise period (Period II) and throughout the first minute after
exercise. Frequent counts of the heart rate were made after this by
watching the string of the electrocardiograph, and these counts were
averaged to get the rate in each of the following minutes of the two
rest periods (Periods III and I\'). The same method was followed in
Periods V, VI and VII. The results are shown in Table 10.

The heart rate was, in general, somewhat more rapid in the patients
with heart disease than in the normal subjects. While sitting at rest
the rates of the cardiac patients were between 86 and 102, while the
rates of the normals were between 69 and 95, and in ten out of eleven
subjects they were between 69 and R8 per minute. The same difference
holds true for the heart rate while standing at rest on the treadmill

Jiesasssssss;

m

si

il

;g3SS :S :SSS

i^SaSS^SoSSSS

JgaSSKg :8Sg

= S8g$!S{: :se

i5^SSc8SSi -S -53

iSSSx^SSRSSS

3SSSSS :PS28

sgssss .-KSg;

gss^^^ i^IHa

SS8S§3SSgS|g

3S2S8?8BSSS

SSSSSSgSS

5S|SS||S|SS
§SSSb2S8SSS

SS??$!38SSS5g

|2|SSg|2ggS

sgassss

assgsssssss

:SaiSS8SS38S2

SoSBSSSSSSSSS

i!

u as

PEABODV-STURGIS—STUDIES OS RESPIRATIOX 297

(Period I). Eight out of ten of the cardiac patients had rates between
86 and 118, and eight out of eleven normals had rates between 74 and
88 per minute. The average heart rate of the cardiac patients while
standing at rest was 95 and that of the normal subjects was 86 per
minute. The difference is thus about 1 1 per cent.

The column of figures under the heading '"preliminary" in Table
10 gives the heart rate during the twenty seconds preceding the exercise
period. The subject knew that he was about to start to walk and the
changes in heart rate represent what is probably largely a nervous
reaction. Among the normals there was an increase in rate of from
five to nine beats per minute in four instances, a decrease of from one
to three beats in three, and no change in two instances. In the group
of cardiac patients seven showed a rise of from one to twenty beats,
with an average increase of five, and three showed a fall of from one
to ten beats, the average being four per minute. There is thus no
constancy in the behavior of the subjects in this preliminary period
and no distinct difference between the two groups.

At the onset of the exercise (Period II) there was an immediate
rise in heart rate in both groups of subjects. In the normals the rate
during the first fifteen seconds of exercise averaged 16 beats per
minute higher than the rate while standing at rest, with variations of
from nine to twenty-two among ten individuals. In the cardiac group
the corresponding average increase was 17 beats, with variations of
from three to thirty-seven in ten cases. During the last fifteen seconds
of exercise, when the heart rate was usually greatest, the average
increase of the normals was thirty-one, with extreme variations between
twenty-three and forty-six, while the average increase in the group of
cardiac patients was thirty-four, with extreme variations between
twenty-one and forty-three per minute. The mean increase in heart
rate was a little greater in the cardiac group than in the normals, as
in seven out of ten cases of the former there was an increase of
between thirty-two and forty-three, while in seven out of ten of the
latter it was between twenty-one and thirty-two per minute. The
differences are not striking, however, and the high heart rates reached
by the patients with heart disease seem to depend largely on the fact
that they started with a higher resting pulse rate.

Immediately after the cessation of exercise there was almost inva-
riably a sudden drop in pulse rate in the normal subjects. This was
such that even in the first fifteen seconds of the rest period (Period
III) there was a fall of from two to twelve beats per minute from the
rate at the end of exercise, with an average decrease of seven beats
per minute in seven out of nine of the normal subjects. In one
instance there was an increase of one beat per minute and in one instance
there was no change in rate. The drop in pulse rate did not begin

298 ARCHU'ES OF IXTERXAL MEDICISE

SO promptly in the cardiac patients, for in six out of eleven cases it
rose slightly (between three and seven beats), in one it remained
unchanged, and in only four was there a fall of between one and four-
teen beats per minute in the first fifteen seconds of rest. The decrease
in heart rate continued to be rapid in the normals so that between thirty
and forty-five seconds after the exercise was stopped only four of
the ten subjects had a pulse rate more than five beats per minute higher
than it was before exercise was begun, and between forty-five
and sixty seconds after cessation of exercise none had a heart rate
more than five beats per minute above the rate standing at rest, and
in six instances it was below this rate. This tendency to bradycardia
continued so that during the second minute of rest the pulse was lower
than it was before exercise in eight subjects. In the patients with heart
disease the fall in heart rate went on more slowly. Between fifteen
and thirty seconds after exercise was stopped the heart rate remained
above what it was during exercise in two cases. Between thirty and
forty-five seconds after exercise the heart rate was more than five beats
per minute above that before exercise in all but one instance, and
between forty-five and sixty seconds after exercise five of the ten
cases still had a rate of more than five per minute above the rate while
standing before exercise. Three out of nine patients with cardiac
disease continued to have a heart rate considerably above that preceding
exercise even during the second rest period (Period IV).

An analysis of the heart rate in these two groups of subjects does
not allow of any satisfactory quantitative relation being established,
but it is quite evident that the fall in heart rate in the normal subjects
usually began immediately after exercise stopped and was constantly
such that within one minute it had reached the level that it was before
the exercise began, while in the group of patients with heart disease
there was often a slight rise in rate after exercise stopped, followed by
a slower fall to the level at which it was before exercise. During the
rest period there was a tendency, in both groups of subjects, for the
heart rate to pass through a phase in which it was even more slow than
it was before the exercise.

In the period in which the normal subjects walked upstairs carrying
a load on their backs there was again a rapid rise in pulse rate varying
from ten to thirty- four beats per minute, with an average rise of
twenty-one above the heart rate at rest in the first fifteen seconds of
exercise. In the fourth fifteen seconds of exercise the increase of
heart rate was from thirty-four to fifty-four, with an average rise of
forty-two beats per minute. \\'hcn the exercise was stopped there was
an immediate slight fall in rate, averaging three beats per minute in the
first fifteen seconds of rest in seven subjects, a rise of seven beats
in one, and in one subject no change of rate. Between thirty and

PEABODV-STURGIS—STUDIES OX KESPIRATIOX 299

forty-five seconds after the cessation of exercise all but one of the
subjects continued to have a rate of more than five beats above the
resting values, and between forty-five and sixty seconds after exercise
the rate was more than five beats per minute above the rate before
exercise in all but four subjects. In the second minute of rest the rate
had come down at least to what it was before exercise in six out of
nine instances, and in the third minute of rest this was the case in nine
out of eleven subjects. There was a general tendency here also for
the heart to assume temporarily a slower rate than before exercise.
In general, therefore, this more severe exertion was followed by a
rather gradual fall in heart rate in the normal subjects which resem-
bled that observed in the patients with cardiac disease after moderate
exercise.

DISCUSSION

In the previous sections a detailed analysis has been given of the
changes produced by slight amounts of exercise in the metabolism,
pulmonary ventilation, and heart rate in a group of normal young men
and in a group of patients of approximately the same age with valvular
heart disease. All of the latter were ambulatory patients, but the effect
of the cardiac lesion varied in different cases so that while some of
them suffered from dyspnea on moderate exertion, others were scarcely
at all limited in their physical activities. In general, those subjects
with the greatest tendency to dyspnea were those in whom the vital
capacity of the lungs was lowest. Certain fundamental differences
between the two groups were observed while they were standing at
rest before the exercise began. In the patients with heart disease, the
oxygen consumption, the heart rate, and the minute-volume of pul-
monary ventilation were greater than in the normals, and the respiration
was more rapid and more shallow than in the normal subjects. The
metabolism of the group of cardiac patients, as measured by the oxygen
consumption, averaged 12 per cent, higher than that of the group of
normals, and, in the light of the fact that basal metabolism and heart
rate have frequently been found to bear a close relation to one another,
it is interesting to find that the average heart rate for the group of
cardiac patients was about 11 per cent, above that of the group of
normals. It seems probable that the more rapid heart action depended
largely on the higher rate of metabolism. The minute-volume of
pulmonary ventilation, on the other hand, averaged 37 per cent, greater
in the group of cardiac patients than in the normals — an increase which
is much too large to be accounted for entirely by the difference in
metabolism. This high minute-volume of the respiration was asso-
ciated with breathing which was more rapid and more shallow, and in
which the percentage of oxygen taken out of the inspired air was less
than in the case of the normal subjects. The higher minute-volume

300 ARCHIVES OF IXTERXAL MEDICI XE

breathed by the cardiac patients was probably the result of shallow
breathing, with a consequent increase of the effect of the "dead space"
of the upper respiratory tract. Under such circumstances the ventila-
tion of the alveoli which is essential for proper gaseous exchange can
only be obtained if the total amount of air breathed is increased, and
this must be accomplished by raising the rate of respiration. Another
possible explanation, however, deserves attention. It has been shown
by Barr and Peters ' thaf not only the total pulmonary ventilation,
but also the effective minute-volume, is increased above normal in
patients with cardiac decompensation. According to their observations
this is due to the low percentage of carbon dioxid in the alveolar air
of patients with cardiac decompensation and the consequent necessity
of an increased effective ventilation in order to bring about the requisite
elimination of carbon dioxid. Further studies on the blood by Peters
and Barr * indicate that this low alveolar carbon dioxid is "largely
brought about by an impairment of the efficiency of the pulmonary
mechanism for the exchange of gases between the blood and the outside
air" which "necessitates the maintenance of a greater difference in
carbon dio.xid pressure between the blood in the pulmonary circulation
and the alveolar air to effect the normal carbon dioxid output." The
cardiac patients who were the subjects of the present investigations,
however, were all ambulatory, and one can hardly assume that they had
a low alveolar carbon dioxid tension since Barr and Peters found this
to be characteristic only of patients who were actually decompensated.

According to the calculations of Barr and Peters, the effective
ventilation of a group of decompensated patients studied by Peabody,
Wentworth and Barker ^ was about 30 per cent, greater than that of a
similar group of compensated patients. Calculated on the same basis,
with the assumption of a "dead space" for the subjects of 130 c.c. and
an instrumental "dead space" of 50 c.c. additional, it is found that the
average eft'ective alveolar ventilation of the present group of normal
subjects, while standing at rest (Period I), was 5,210 c.c, while that
of the patients with heart disease, under the same circumstances, was
5,800 c.c. The difference, which is scarcely more than 10 per cent.,
can well be accounted for by the fact that the oxygen consumption of
the group of cardiac patients was about 12 per cent, greater than that
of the normals. Beyond this, therefore, the evidence indicates that
in the two groups the effective, or alveolar ventilation, is of essentially
the same magnitude and thus the findings of Barr and Peters do not
account for the high minute-volumes observed in the compensated cases
studied in this investigation.

7. Barr, D. P.. and Peter.s, J. P.. Jr.: .^m. J. Physiol. 54:345. 1920.

8. Peters, J. P., Jr., and Barr, D. P.: J. Biol. Chem. 45:537, 1921.

PEABODV-STURGIS— STUDIES OX KESPIRATIOX 301

The fact that cardiac patients have a greater total pulmonary
ventilation while obtaining practically the same effective ventilation as
normal subjects has an important bearing on the production of dyspnea,
because it has been found that the sensation of dyspnea is closely asso-
ciated with the breathing of a high minute-volume of air, and the
explanation of this phenomenon should be sought. The question arises
as to whether an essentially mechanical cause may underlie the tendency
to a rapid and shallow respiration observed m patients with heart
disease. In the group of patients under consideration the vital capacity
was so Httle below normal that it is rather difficult to conceive of there
being any restriction to the depth of breathing, especially while they
were standing quietly at rest and using only a small portion of their
vital capacity at each respiration. The possibility of some limitation
to the depth of breathing is, nevertheless, worthy of consideration.
Drinker, Peabody and Blumgart " have shown experimentally that
engorgement of the pulmonary vessels may interfere with the entrance
of air into the lungs, and their results indicate that the decrease in the
vital capacity of the lungs in early cases of heart disease, in which there
is no pulmonary edema or effusion into the pleural cavities, is due to
congestion of the pulmonary circulation. Cardiac lesions potentially
capable of causing engorgement of the pulmonary vessels were present
in all the patients in the present study and in all but three the vital
capacity was below the usual normal limits. The decrease was not
great in any of the cases but it does not seem unreasonable to suppose
that whereas the violent muscular effort exerted in measuring the
vital capacity of the lungs would result in an essentially normal volume
being taken into them, nevertheless, it might be easier for the same
subject to take shallow respirations. If such were the case, they would
instinctively breathe less deeply, even when they were at rest, than
normal persons. Quite possibly this is a matter which concerns the
Herring-Breuer reflex and there is much evidence to indicate the
extreme sensitiveness of this mechanism. Even if such a conception
is accepted it is not surprising to fail to find complete harmony in
individual cases between the volume of the tidal air and the decrease
of the vital capacity of the lungs. E. J. W., for instance, whose vital
capacity was 62 per cent, of normal, breathed slowly and deeply while
at rest and even during exercise. This may have been due to more or
less conscious effort. Individual exceptions will always be found in the
study of any mechanism which, like the respiration, is under the control
of so many factors, chemical and nervous, and at present one can only
suggest that the general tendency to shallow breathing seen in patients
with heart disease is due to engorgement of the pulmonary circulation.

9. Drinker, C. K.; Peabody. F. W., and Blumgart, H. : J. Exper. M. (in
press).

302 ARCHU'ES OF IXTERXAL MEDICIXE

During the period'of exercise, which consisted in walking sixty steps
upstairs in one minute, there was an increase in- oxygen consumption,
heart rate, minute-vohime, and rate and depth of respiration in the
two groups of subjects, and the eflfect produced was essentially the same
in both. A given amount of exercise apparently causes a rise in
metabolism, heart rate and pulmonary ventilation in patients with cardiac
disease that is proportionally the same as that observed in normal
subjects. There was no difiference in the type of reaction to exercise
between the two groups.

The patients with heart disease became more short of breath as
the result of the exercise than the normal subjects did, but in both
groups the dyspnea was almost constantly described as being greater
immediately after the exercise stopped than it was during the actual
exercise itself. For this reason the first minute of rest after exercise
was studied with particular care. It was found that the minute-volume
of the respiration rose so that it was considerably higher during the
first minute of rest than during the minute of actual exercise in both
groups of subjects. The highest minute-volume of pulmonary ventila-
tion was thus coincident with the greatest subjective dyspnea. The
actual volumes breathed were greater in the case of the cardiac patients
than in the case of the normals. The former, therefore, approached
nearer to their maximum pulmonary ventilation and encroached farther
on their pulmonary reserve.

Why is the sensation of dyspnea more noticeable after exercise than
during exercise? It is reasonable to seek the answer to this question
in the explanation of the high pulmonary ventilation after exercise,
for the degree of dyspnea varies with the intensity of the stimulus to
respiration, and the response of the respiratory center is indicated by the
extent of the pulmonary ventilation. One of the most important
stimuli to the respiratory center is carbon dioxid, and the respiratory
quotients, which were generally low during exercise and high during
the period following exercise, suggest that there was a lag in its
excretion. This would result in a prolongation of its stimulation of
the respiratory center, and cause a high pulmonary ventilation even
after the exericse was stopped. The fact that the minute-volume of
re.spiration was sometimes actually greater immediately after exercise
than during exercise may depend in part on mechanical interference
with the respiratory movements at a time when other bodily movements
are being actively carried on.

Following the rise in the minute-volume of the resj)iration, which
occurred immediately after exercise, there was a gradual decrease in
minute-volume in both groups of subjects, but the return to the level
which existed before exercise was distinctly slower in the patients with

PEABODY-STURGIS—STUDIES OX RESPIRATIOX 303

heart disease than in the normal subjects. Hunt and Dufton" have
recently called attention to the same phenomenon. It is undoubtedly
due to an abnormal delay in the excretion of carbon dioxid or other
stimuli of the respiratory center. The fact that the minute-volume
increased during and after exercise in the same proportion in the cardiac
patients as in the normals indicates that this lag is not due to the
pulmonary ventilation being less efficient in the patients with heart
disease. Further evidence of this is gained if the effective alveolar
ventilation is calculated for the exercise period. Assuming again a
total "dead space" of 180 c.c. the average effective ventilation of the
normals was 11.6 liters and of the cardiac patients 13.2 liters per
minute. The difference amounts to approximately 13 per cent, and is
just about the same as the difference found during the rest period
before exercise. There is, thus, no reason for considering that the pro-
longed increase of pulmonary ventilation after exercise in the patients
with heart disease studied in this investigation was the result of inef-
ficiency of the lungs in removing carbon dio.xid from the system. The
cause must be sought elsewhere and it is natural to consider the
circulation.

The evidence to be derived from these investigations is incomplete
because there were no direct determinations of the minute-volume of
the circulation before and during exercise, but the observations on the
heart rate bear on the question and are suggestive. In the group of
normal subjects the heart rate dropped suddenly and rapidly imme-
diately after the cessation of exercise, reaching the level at which it
was befof exercise in less than one minute. In contrast to this the
decrease in heart rate after exercise in the patients with heart disease
began later, was slower, and was sometimes preceded by a slight rise
in rate just after the cessation of exercise. Martin and Gruber "
believe that the sudden rise of heart rate at the onset of exercise is
due to associated innervation and the passage of impulses from the
cortex which inhibit vagys activity. The sudden fall in heart rate after
exercise might be explained by the cessation of such depressing impulses
but it is difficult to understand, on this basis alone, why the heart often
remains rapid in the patients with cardiac disease. It may, perhaps,
be accounted for by assuming the existence of some metabolic stimulus
to cardiac acceleration which had failed to be eliminated with normal
rapidity in these patients with impaired hearts. The rapid heart action
would, then, represent an attempt to raise the minute-volume of the
circulation, and this abnormal acceleration would constitute an indica-
tion of the inefficiency of the heart to meet the demands put upon
it by the strain of exercise. It is, of course, in this sense that the same

10. Hunt. G. H.. and Dufton. D.: Quart. J. Med. 13:165, 1919.

11. Martin. E. G.. and Gruber. C. M. : Am. J. Physiol. 32:315. 191.3.

304 ARCHIJ-ES OF IXTERXAL MEDICIXE

phenomenon is commonly used as a clinical test. With an inadequate
circulation the slow elimination of carbon dioxid is easily explained
and its retention would result in a stimulation of the respiratory center
and account for the prolonged increase of pulmonary ventilation.

According to the conceptions which have been outlined, therefore,
the fact that the patients with heart disease had a greater subjective
sensation of dyspnea after slight amounts of exercise than normal
persons depends on two factors. They had a tendency to breathe less
deeply than normal individuals, with the result that the total pulmonary
ventilation was greater in order to obtain the same effective alveolar
ventilation, and this large pulmonary ventilation was maintained for a
longer period of time because the circulation became inadequate during
the strain imposed by exercise and the elimination of carbon dioxid was
delayed. In cases of heart disease which are more severe than those
reported on here, and in whom the vital capacity of the lungs is
markedly decreased, the mechanical interference with pulmonary
ventilation, owing to inability to breathe deeply, plays a considerably
greater role in the production of dyspnea.

The nomial subjects did not become definitely short of breath as
a result of walking upstairs and they were put through a second test,
in which they took the same number of steps carrying a load on their
backs in order that the development of dyspnea in them comparable
in degree to that which occurred in the cardiac patients, might be
studied. Greater changes were observed in the metabolism, heart
rate and pulmonary ventilation, but the general type of change was
the same as was present in the patients with heart disease who under-
took less exercise. Even the slower fall in heart rate and minute-
volume of respiration after cessation of exercise was seen. The most
striking difference noted was that the tendency of the normals to
breathe more deeply gave them more effective ventilation of the
alveoli of the lungs and thus enabled them to meet a greater demand
for oxygen with a lower total pulmonary v.entilation. Thus, when
walking upstairs and carrying a pack, the normal subjects had a con-
siderably higher oxygen consumption than either they or the cardiac
patients had when walking without the pack, but the average minute-
volume of pulmonary ventilation was almost exactly the same for the
normals when walking with the pack as for the cardiac patients when
walking without the pack.

coxci-rsioNS

In a group of eleven ambulatory patients with heart disease it was
found that the oxygen consumption and heart rate were slightly greater
than in a similar group of normal subjects while standing at rest.
Under the same conditions the minute-volume of the respiration was
much greater in the cardiac patients and the breathing was mere rapid
.and more shallow.

PE.-lBODy-STURGIS—STUDIES OX RESPIRATION 305

The slight amount of exercise involved in walking up sixty steps
produced the same relative changes in oxygen consumption, pulmonary
ventilation and heart rate in the two groups, but it caused more sub-
jective dyspnea in the patients with heart disease. Exercise which
was severe enough to cause a corresponding amount of dyspnea in
normal subjects caused the same type of changes in oxygen consump-
tion, pulmonary ventilation, and heart rate, but they were greater in
degree. The response to exercise of patients with heart disease is
qualitatively the same as that in normals, and their greater liability
to dyspnea depends on a quantitative limitation.

Shortness of breath was most noticeable immediately after exercise
was stopped, and at this time the pulmonary ventilation was largest.
The return to normal of the minute-volume of the respiration and of
the heart rate was delayed in the cardiac patients.

It is suggested that the two factors which account for the greater
dyspnea in the cardiac patients are the inadequate circulation, which
results in a delayed elimination of carbon dioxid, and the tendency to
shallow breathing, which necessitates a relatively large pulmonary
ventilation.

This investigation was begun in association with Dr. Howard F. West and
Miss Edna H. Tompkins, and a large amount of preliminary work was done
while they were in the laboratory. We regret that they were forced to leave
Boston before the present series of experiments was begun, and wish to express
our great appreciation of the assistance which they contributed.

STUDIES IN DIABETES INSIPIDUS, WATER BAL-
ANCE, AND WATER INTOXICATION

STUDY I
JAMES F. WEIR, M.D.; E. ERIC LARSON, M.D.,

AND

LEONARD G. ROWXTREE, M.D.

ROCHESTER, MINN.

Two years ago we developed a special interest in diabetes insipidus,"-
and as a result have been carrying out investigations on patients suffer-
ing from this disease. During the course of clinical studies certain
observations were made which led to experiments in animals relative
to water balance and to water intoxication.

In this study we shall report (I) the results of our clinical studies
in the fifteen cases of diabetes insipidus observed during the last two
years in the Mayo Clinic and in the University Hospital, Minneapolis;

(2) the effects on urinarj' secretion of the subcutaneous injections of
the extract of the posterior lobe of the pituitary gland in normal dogs
and in dogs which have undergone resection of the renal nerves; and

(3) the production of water intoxication by the administration of
water by mouth subsequent to the subcutaneous administration of
pituitary extract.

STUDIES IN DIABETES INSIPIDUS "

Etiology. — \^'e have made an attempt to determine the etiology of
the diabetes insipidus in our cases. One patient had brain tumor, not
definitely localized but known to be supratentorial ; another had a hypo-
physeal tumor, and four had syphilis ; in nine cases no causative factor
could be determined. The incidence of syphilis in our series is high.
Its etiologic importance has been emphasized by Fournier.' Futcher,*
and Oppenheim."' Whether or not syphilis was responsible for the
diabetes insipidus in our cases could not be determined, but vigorous
treatment for syphilis failed to exercise any significant effect on the
course of the diabetes insipidus.

1. Rowntree. L. G. : Diabetes Insipidus. In: Oxford Medicine, London.
Oxford Univ. Press 4:179. 192L

2. The details of the clinical studies will appear in a separate publication
by Dr. E E. Larson.

,3. Fournicr: Quoted by Futcher.

4. Futcher, T. B. : Diabetes Insipidus. In: Osier. W.. and McCrae, T.; Mod-
ern Medicine, Its Theory and Practice, Philadelphia, Lea & Febiger, 1914 2:721.

5. Oppenheim. H. : Die syphilitischen Frkrankungen des Gehirns. Specielle
Pathologie und Therapie. Vienna, Nothnagel 9:196. 1896.

irElR-LARSOX-KOirXTREE— DIABETES IXSIPIDUS 307

The Nature of the Disturbance of Water Balance. — The cardinal
symptoms of diabetes insipidus are related to disturbance of water
metabolism ; water is ingested and excreted in amounts far in. excess
of normal. In Trousseau's " case, for example, the most extreme on
record, the fluid intake amounted to 40 liters and the urinary output to
43 liters a day. Thirst, the most annoying symptom, occasions extreme
discomfort at times, while polyuria and polydypsia are the source of
great inconvenience. Constipation and absence of sweating minimize the
loss of fluid from the bowels and skin. The exact amount of fluid lost
by way of the lungs and skin has not been determined accurately in
diabetes insipidus so far as we know. Such studies are most desirable,
but they necessitate the employment of special apparatus. Loss of
fluid by way of the lungs plays a relatively smaller part than in the
normal person. Studies of fluid balance in our cases show a striking
correlation between the fluid intake and the urinary output. The litera-
ture abounds with similar results ; but some writers have recorded
urinary excretion somewhat in e.xcess of the fluid ingested.

The fundamental question of whether the thirst or the polyuria is
primary, v\-e find difficult to answer. In this connection we have con-
sidered factors as follows: (l)the sequence of onset of symptoms,
that is, whether thirst or polyuria constittites the initial symptom ;
(2) the influences of various procedures in controlling thirst and
urinary output; and (3) the presence or absence of organic or func-
tional changes in the urinary tract.

Sequence of Symptoms. — By careful questioning with regard to the
onset of the disease, it was possible, in three of our cases, to elicit a
definite history of sudden onset of thirst, and to establish that in two
thirst was primary, that is, it preceded the polyuria. In the first case
the circumstances surrounding the onset of the disease were such as to
make the evidence almost conclusive. The patient, a farmer's wife,
was left alone in her home throughout the entire evening. She became
extremely nervous and was frantic with fright. She lay in bed most
of this time too frightened to move. During this period she developed
a terrific thirst. She suffered no discomfort from bladder distention,
nor did she void. She states that on the return of the family she
rushed to the spigot and drank copiously of water, taking about two
quarts. She insists that she did not void at this time, for in order to
do so she would have been compelled to pass through several dark
vacant rooms, which in her nervous state she would not have attempted.
Polyuria developed on the following morning. The second patient, an
intelligent man. insists that thirst was the primary symptom, and that
it developed during the night. He got up and drank copiously of water.

6. Trousseau. A.: Lectures on Ch'nical Medicine. London. New Sydenham
Soc. 8:528, 1870.

308 .-iRCHirES OF IXTERXAL MEDICIXE

voided but little, and did not suffer from any discomfort from bladder
distention. In the third case both thirst and polyuria came on suddenly,
but the patient was unable to recall the order of their appearance. In
the remaining twelve cases the onset of the disease was insidious and
consequently no conclusions could be reached concerning the priority
of the symptoms.'

Tlic Influence of Varioits Procedures in the Control of Thirst and
Urinary Output. — Cannon * explains nomial thirst on the basis of a
local sensation resulting from local dryness of the buccal mucous mem-
brane due to decreased secretion from the salivary glands, which in
turn is dependent on the diminished supply of fluid furnished these
glands by the body because of its depletion in water. Cannon further
states that the osmotic pressure of the blood remains unchanged despite
deprivation of fluids in the tissues. Dehydration of all tissues, includ-
ing salivary glands, leads to diminished secretion of water by these
glands. Normally the salivary glands furnish the index to the body
need for water.

Thirst and a dry mouth were complained of by all our patients.
Two volunteered the information that they experienced a peculiar
sensation "as though the mouth were full of cotton." One demon-
strated that "tenacious strings of saliva stretched from the roof of the
mouth to the tongue, when the mouth was opened wide." There is
little doubt that the viscidity of the saliva is increased.

We attempted to determine the role of the local sensation in the
mouth with regard to water intake and urinary output in diabetes
insipidus. We believe that specific local nervous influences have been
excluded as a cause of the thirst since cocainization of the mucous
membrane of the mouth and of the nasopharynx to the point of anes-
thesia failed to control either the polydypsia or the polyuria. In fact,
in one case cocainization was pushed to the point of mild constitutional
toxicity with no efifect on the diabetes insipidus. In other cases- pilo-
carpin was injected to procure salivary secretion, but without results.
In four of these cases there was no evidence of decreased urinary
output nor of diminution of thirst. .As a result of these studies it
appears that the thirst in diabetes insipidus is more than the mere
expression of dryness of the oral mucous membrane.

The Presence or Absence of Organic and Functional Disturbances
in the Urinarx Tract. — Careful studies of renal function in diabetes

7. Since this paper was written Bailey and Bremer liavc repented tlieir
studies in experimentally induced diabetes insipidus before tbc Society of
Endocrinology, Boston, June, 1921. They have demonstrated conclusively that
thirst and polydipsia may precede polyuria.

• 8. Cannon, W. B. : The Phvsiob.Rical Basis of Thirst, Proc. Rov. Soc,
London, s. B. 90:283. 1918.

WEIR-LARSOX-ROWSTREE— DIABETES ISSIPIDVS 309

insipidus failed to incriminate any part of the urinary tract as an
etiologic factor, and revealed no deviations from normal function,
except in relation to water excretion and perhaps to salt excretion.

In all our cases the urinary output was markedly increased, from
3 to 14 liters a day. The specific gravity was markedly decreased,
ranging, as a rule, between 1.001 and 1.004. Albumin appeared inter-
mittently in small amounts in three cases, while demonstrable glycosuria
was persistently absent. The excretion of phenolsulphonephthalein
and the values for blood urea, creatinin, and uric acid were uniformly
normal. Cystoscopic examination in two cases showed normal bladders.
Ureteral catheterization revealed the fact that the increased secretion
of urine was bilateral, and approximately proportional for both kid-
neys, and that the appearance time for phenolsulphonephthalein was
normal from each side. Administration of pituitary extract with the
ureteral catheters in place, reaching to the pelvis of the kidney, resulted
in control of the polyuria and in a change in the concentration and
color of the urine, within eight minutes in one case, and within ten
minutes in another. These results conclusively exclude constriction of
the ureter as a factor in the control exercised by pituitary extract on
the polyuria and gross organic changes in the kidney and urinary tract.
Hoppe-Seyler " presents the situation aptly when he says of renal
function in diabetes insipidus, "it fails only by a little pituitrin of being
normal."

The Results of Treatment. — The discovery by Schafer ^^ and his
colleagues of the influence of the extract of the posterior lobe of the
pituitary on the urinary output has created a new and growing interest
in the subject of diabetes insipidus. Unfortunately, Schafer's observa-
tions were made on anesthetized animals and led to the belief that the
active substance of the posterior lobe possesses dmretic properties.
Physiologists and clinicians were alike deceived with the result that
diabetes insipidus was looked on by many as the result of overactivity
of the posterior lobe. However, in 1913, von der Velden,'' in Germany,
and Farini and Ceccaroni,'- in Italy, working independently, adminis-
tered extract of the posterior lobe to patients suffering from diabetes
insipidus and obtained an effect almost specific; thirst, polydypsia, and
polyuria were all promptly and efifectively controlled, at least for a
temporary period.

9. Hoppe-Seyler. G. : Ueber die Beziehuiig des Diabetes insipidus zur
Hypophysc und seine Bebandlung mit Hvpophysenextrakt, Miinchcn. med.
Wchnschr. 62:16.3.^. 1915.

10. Schafer. E. .\.. and Herring. P. T. : The action of Pituitary Extracts on
the Kidney. Proc. Roy. Soc. London, s. B. 77:571, 1905.

11. Von der Velden. R. : Die Niercnwirkung von Hypophysenextrakten beim
Menschen. Berl. klin. Wchnschr. 1:208.3. 191.3.

12. Farini. A., and Ceccaroni, B. : Influenza dcgli estratli ipofisari suU'
eliminazione dell' acido ippurico. Gazz. d. nsp. 34:879. 191.3.

310

ARCHIVES OF IXTERXAL MEDIC IS E

DATA COXCERNING THE FIFTEEN CASES IN THE SERIES

The Effect of Pituitary Extract. — The effect of pituitary extract
was studied in all our cases. Marked temporary results were obtained
in all. The effects were immediate, and lasted for periods ranging
from a few hours to four or five days (Figs. 1, 2 and 3). In mild
cases the water balance was reduced to its normal level, while in more
severe cases the level was reduced strikingly. Pituitary extract in gum
acacia exercised as good an effect and resulted in a somewhat more
prolonged control, although the local reaction was somewhat more
severe. Effects of short duration sometimes follow the administration
of pituitary extract by rectum.

Fig. 1 (.Case .A 293,184).— Fluid intake,
in the three-hour periods while the patient

lie vulunie and specific gravity
receiving varying treatment.

With the decrease in the amount of urine the specific gravity rose
to 1.010 or 1.025, and the color increased proportiotiately. With the
readjustment of urine secretion the saliva flowed more freely ; the skin
became moist, and actual perspiration occurred in a large number of
the cases. With the diminished desire for water the taste for it was
frequently perverted, the patients referring to water at these times as
being "stale" or "flat." A sensation of satiety and a feeling of fulness
resulted from the intake of small quantities of water, for example,
half a glass. This is in striking contrast to their usual experience.
Prior to the administration of pituitary extract, a quart of water often
failed to relieve the consuming thirst.

ll-EIR-L.4RS0.\-R0n'XTREE—DL-}BEriLS INSIPIDUS 311

Fig. 2 (Case A ,510.741 ). — Fluid intake, uriiic vuluine and sjiecilic gravity
in three-hour periods during marked antidiuretic effect of pituitary extract and
water intoxication.

HliE»tiiLlfa.^u»J^j|ajiaM|t^i

r:K 1 Ma~c 1 4i i.o-i.: k- -l'rnirinKc<i atnuluiretic effect of i)Uuitar_v extract
when patient exercised voluntary restriction on his thirst.

312

ARCHIVES OF IXTERXAL MEDICIXE

The volume of saliva secreted after the administration of pituitary
extract increased definitely. This is illustrated in Figure 4 which con-
trasts the action of pituitary extract on the volume of urine and
saliva of normal persons and of a patient with diabetes insipidus.

Effects of Histamin. — Histamin was administered subcutaneously in
three cases. Constitutionally and locally its effect differed entirely
from that of pituitary extract. In no instance was either the thirst or

Case Al 40642

I III "

1 ath i1 p : lUok i riiia

4"i r, - 1 I ■'Mk 1

:::" : _:::: :: -

: : ::: ::x:: :

30 c.e.

:::::::":::: :: :

i ]

::::_$ " ' -- -■

::::: i\ — ■ -

20 c.c. ;«3»'^

iiji£-3iu.,_..i=.±--

--± ^,;---=^.

■""==- -^^v

15 c.c.:: : ::: ::"-:"4.

"::: ::::"5:i;: :

:. _-_-.- :±

__ - - . _ =<i

10 c c L " - - -

.

\

1

:::: : y.\ i ::S::

'"=-=-S5^i:iiiiJ5u

._.-_,-=....: = ==. ^,

:::'!!:'':±":::"::3

Fig. 4 (Case .A 140,642).

V c.e. pituitary extract
".ffect of pituitary extract on salivary ami urinary

the polyuria decreased, excc|)t on one occasion when slight control was
noted for three hours. The usual type of result may be seen in
Figure 1.

Results of Spinal Puncture. — A spinal puncture with the with-
drawal of from 5 to 10 c.c. cerebrospinal fluid was performed in six
cases, without demonstrable effect on any of the cardinal symptoms.
Because of the favorable reports in the literature we believed that the

irElR-L.-iRSOX-KOliXTKEE—DI.-IBETES JXSIl'IDiS 313

imount of fluid withdrawn might be of significance. Consequently
spinal drainage was done in three other cases, but in not a single
instance was a favorable effect obtained.

Results of Treatment for Syphilis. — In the four cases with evidence
of syphilis, treatment was carried out, but it has proved unavailing to
date. Spinal puncture in three of these cases also failed to give relief.
The Effect of Restriction of Fluids. — Voluntary restriction of fluids
was attempted in several of our cases, but it met with minor success
in two only. One patient, a woman, succeeded in decreasing the intake
from 5 to 6 liters a day to 1.5 liters, through her determined efifort to
refrain from water. Her case was unusual, however, the diabetes in all
probability being hysterical in nature. In the second case early syphilis
of the central nervous system complicated the diabetes insipidus. By
the aid of 1 c.c. pituitary extract restriction to approximately normal
amounts was obtained for a period of five days (Fig. 3). Without
pituitary extract attempts at restriction accomplished but little, while
pituitar}' extract alone failed to exercise control for periods longer
than two days.

Before prescribing restriction of fluids the intense distress accom-
panying water fasting should be duly considered. One of our patients
who had undergone a "restriction cure" at the hands of her local
physician, had tears streaming down her face as she told us her story.
In true diabetes insipidus little or nothing can be accomplished through
restriction of the fluid intake, and the suffering inflicted is out of all
proportion to the slight benefit obtained.

Metabolism." — In a consideration of metabolism in diabetes insipi-
dus the changes induced in normal persons by copious water drinking
must be kept in mind, as has been pointed out by several authors.
Rosenbloom and Price '* give an excellent review of the literature on
the metabolism in this disease. Gibson and Martin " have recently
published results of metabolic studies before and after administration
of pituitary extract.

Metabolism studies have been carried out in two cases of diabetes
insipidus (Cases 140,642 and .130,908), the study in one being repeated
during the patient's second visit to the clinic. Studies of blood chem-
istry were made in two other cases (Cases 317,545 and 323,015) before
and after the administration of pituitary extract. The results in one
case (Case A 330,908) are shown in Tables 1, 2 and 3. The water intake

13. The details of these metabolic studies will appear in a separate pub-
lication by Dr. J. F. Weir.

14. Rosenbloom. J., and Price. H. T.: A Metabolism Study of a Case of
Diabetes Insipidus. Am. J. Dis. Child. 12:53, 1916.

15. Gibson. R. B., and Martin, F. T. : The .Administration of Pituitary
Fxtracf and Histamin in a Case of Diabetes Insipidus. .Arcli. Int. Med. 27:
.■?51, (March) 1921.

ARCHIFES OF IXTERXAL MEDICJXE

TABLE 1.— Urine on Metabolic Experiment (A 330908)

1.009 343 4.30 8.02 .... 0.26(i 0.511 0.048

During treatment

1.025 268 2.59 S.134 5.497 0.466* 0.493 0.071 67.5

1.024 356 3.40 8.20 7.092 0.209 0.537 0.120 86.5

I.OIS 318 7.19

Alter treatment

voided, alkaline at one period on tliis day.

TABLE 2.— Balance of Water, Nitrogen and Chlorin (A 330908)

Intake

Uri

ne Outpi:

„

Balance

' ^

-

' ^

_-

_•

' ^

=■

_-

>.

"o

m

-°S

y

ts

HO

^^

-o

£|

P

z

J

5,189

8.42

5.94

4,525

8.26

3.45

+ 664

+0.16

+2.49

162.6

5,740

8.22

6.38

4,960

7.96

4.30

+ 760

+0.26

+2.08

162.5

3

8.14

2.59

+2,278

—0.36

+2.85

163.0

4

2,380

7.92

6.27

1,200

8.20

3.40

+1.180

-0.28

+2.87

5

4,732

8.61

5.62

9.27

7.19

+ 287

6

6,382

7.57

5.80

5,960

8.02

5.72

+4,422

-0.45

+0 08

162.0

TABLE 3.— Blood on Metabolic Experiment (A 330908)

Whole Blood Mg. per 100 C.(

11

67

68

0.515
0.530

1,171
1,142

648

Before treatment

29.8 14.9 2.68
26.5 14.9

1.90 128
106

Normal
Normal

69

72

0.542
0.531

l^

640
623

During Treatment
26.6 16.2
24.8 14.8

14*
106

Pituitary extract
Pituitary extract

71
69

0.522
0.546

1,099
1,129

683
648

After treatment

28.1 14.8 ....
27 0 13.3 ....

108
112

Alter period
After period

irEIR-LARSOX-KOUXTKEE— DIABETES JXSIPIDUS 315

showed a marked fall in the three hourly and daily quantities after
the administration of pituitary extract. The variations in the volume
and specific gravity of the urine were marked and obvious. The posi-
tive water balance in the control period was small, harmonizing with
the clinical observation of dry skin and mouth. There is marked reten-
tion of water after the first dose of pituitary extract, corresponding
with the decreased urinarv- output. The patients generally report an
increased excretion of urine in the period after treatment. In this
case, however, the water intake kept pace with the increased excretion,
but this fact, together with the increase in weight during treatment and
the return to normal in the after period, indicates retention of water
in the blood or tissues. The excretion of nitrogen varied but slightly

TABLE 4.— Effect of Pitu:t.\rv Extract in Normal Person and in P.^tient
wrrH Diabetes Insipidus Contrasted in Blood Picture

Plasma Hg. per
I. 100 C.c. Whole Blood Mg. per 100 C.c.

.1

1

1

H

<I

s

O

11

1

=

1?

~

56

0.534

1,008

605

60

977

610

61

0.534

942

618

61

0.522

1.068

629

0.549

1.163

67

0.513

970

597

Diabetes iosipidus (A31V543)

. . . Control

146 Control

152 Pituitiirj

149 Pituitary

Control
Control

Pituitary e.xtraet
Pituitary extract

throughout the experiment and the nitrogen balance showed no appre-
ciable change. The chlorin excretion was decreased under treatment
and chlorin retention was evident.

In the second case (Case 140,642), similar changes in the urine
were noted, and more marked change in the water balance ; there was a
large negative balance as the effects of the drug wore olT. There was
a definite decrease in the excretion of nitrogen and an increased posi-
tive balance in the period of treatment. Chlorin also showed rather
marked retention.

In the analysis of the blood in diabetes insipidus before and after
administration of pituitary extract certain definite changes appeared.
Typical results in one case (Case 317,545) are contrasted with those
of a normal person (Table 4). The relative plasma volume is increased,
and the molecular concentration, the total nitrogen, and the sodium
chlorid of the blood are somewhat decreased, findings which might
indicate a dilution of the blood. However, in the findings in the

316 ARCHIVES OF IXTERXAL MEDICINE

experiments on metabolism there is evidence of a slight retention of
nitrogen and chlorin, which does not appear in studies on the blood.
Evidently water retention is the most prominent feature following
administration of pituitray extract, and in all probability the effects on
nitrogen and chlorin are secondary to water retention.

In all our experiments on metabolism and in all our cases of diabetes
insipidus in which pituitary extract has been administered, there has
been no diarrhea,^'' and an analysis of the feces in one case showed a
normal quantity of water. There is no evidence that pituitary extract
prevents the absorption of water from the gastro-intestinal tract of
man, as has been claimed for it for rabbits by Rees.'' Only small
quantities of water have been found in the intestines and feces of our
animals at necropsy.

Blood Volume. — Blood volume determinations by the vital red
method were carried out in four of our patients, and in three of them
before and after the administration of pituitary extract, that is, in the
periods of high and low urinary excretion. The results obtained in the
two earlier cases led us to believe that the blood volume, particularly
the plasma volume, was increased in diabetes insipidus following the
administration of pituitary extract, that the pituitary extract had spe-
cifically raised the renal threshold for water, and that the kidney failed
to excrete water even in the presence of hydremic plethora in the blood.
However, in one case no increase was found in either total blood volume
or plasma volume, despite the completeness of control exercised by the
pituitary extract on the polyuria. Prior to the administration of the
drug the plasma and total blood volume were 3,460 and 4,80.^ c.c,
respectively, and after the drug, 3,495 and 4,855 c.c, respectively.
From these results it is evident that pituitary extract can cantrol the
polyuria of diabetes insipidus and prevent polyuria following excessive
ingestion of water without the development of any significant increase
in blood or plasma volume. These findings are in keeping with the
experimental results which we obtained in animals.

THE EFFECTS OF PITUIT.\RV EXTR.ACT ON THE W.ATER I!AL.\.\CE

Behaznor and Role of Water in Metabolism. — In order better to
understand the eft'ect of the pituitary extract on the water balatice, it is
desirable to know more concerning the nature of water exchange in the
body, the mechanism involved in its regulation, and the part played by
the pituitary gland under normal and abnormal conditions. Water
plays a fundamental part in metabolism, and is essential to the life

16. Diarrhea was present in animals suffering from water into.xication, but
the amount lost by way of the bowels was relatively very small.

17. Rees, M. H. : The Influence of Pituitary Extracts on the .Absorption of
Water from the Small Intestine. Am. J. Physiol. 53:43, 1920.

]V EI R-LARSOX-ROW XT REE—DIABETES IXSIPIDUS 317

and function of every living cell. It is transported to and from the
cells by the blood, provision for its exchange being made locally by
the presence of lymph spaces and lymph channels. It constitutes 80
per cent, of the blood. Oxygen and food are also transported to the
cells by the blood. Nature has arranged to supply the blood with water,
oxygen, and food, according to the constancy of the need of the cells
for each and in accordance with the complexity or simplicity of the
mechanism involved in their metabolism. Water holds a position
intermediate between oxygen and food, the continuation of its supply
being less vital than that of oxygen and more than that of food. In

T.^BLE 5.-

— SUPPL\

AND Output of Water

Amount

Author

Remarks

Supply

rom

Average

Data collected in Munich: fluid in-

and

3.500

2.300

2.2S0

2,440

3.-00

2.2-25

Benedict

ject in repose
Over a period of 66 days with sub-

480

240

oxidation

4,000 to 2.000 calories, respectively

Output

Urine..

»

3.000

SOO

Emerson

So many factors effect the urinary
output that only a general aver-

age Is given here*

Skin...

550

vonNoorden These are average figure- lor patient

Lungs

400

indoors on a medium diet of 1,800
to 2,000 calories. Total fluid lost
by evaporation may reach 1,600
c.c. on diet ol 3.500 calories and
3.230 c.c. on diet of .i.OOO calories
with patient doing moderate or
heavy work. As much as 7..5 liters
may be lost during excessive work

Fecest.

60
to
300

300

GO

On vegetarian diet with large stools
the water content may reach 300
c.c. In diarrhea greater losses
may occur

' In Ave cases followed by us during summer iiionti
c.c. The flgurog for maximum and minimum simply
Obviously this refers to normal loss in the stools.

iiealth blood varies in volume and consistency within narrow limits
only, possibly in order that the supply of these essentials to the cells
may be kept almost constant. Fluctuations in physiologic needs,
depending on variations in function are met largely by changes in the
rate and volume of the blood flow to various organs.'*

The source of the water of the body is found in the fluids taken by
mouth in the form of food and drink and in the water which results
from oxygenation of food within the organism. Water is excreted
by the skin, lungs, kidneys, and bowels. The intake and output vary
tremendously with the life, habitat, and habits of the individual. The

18. The intermediary water metabolism and the effects exercised by water
ingestion are purposely omitted from discussion.

318

ARCHIVES OF IXTERXAL MEDICIXE

relative amount excreted through the various channels also varies
extremely, reciprocal relationships playing a striking part at times,
particularly with regard to sweat and urine. Although the variations
are extreme, some idea concerning the normal relations in water bal-
ance, that is, in the intake of fluids and in the output of water by the
various channels under differing conditions of diet and activity are
presented in Table 5.

Effect of Extract of Posterior Lobe of the Pituitary Gland on the
Urinary Excretion in Normal Men and Animals. — In order to ascertain
the effects of pituitary extract on the urinary excretion of normal
persons, five men were placed under careful observation. Urine was
collected in three-hourly periods throughout the day, and as a single

X i-

4 sif^ 1 ^'\ jL ^

-^^1 ^A £ ,^ V" .^ .

2o2- , ^ V-^ ^£g ,r

.t ^-^ A- S-E^^"

tV ^ ul t

I ^"^^^ A f j

i £ \^ ^J -n

V . v^^ y' 0

^^ ^ t

Oay 1 Day 2 Day 3 Day 4

Speoifie gravity - - - -. volume

Fig. 5.— .Average three-hour volume and specific gravity of the urine of
five normal men before and after pituitary extract.

specimen at night. Diet and climatic conditions remained fairly con-
stant during the period of observation, August, 1920. A two-day
period of control preceded the administration of the drug, which was
given at 8 a. m. on each day of the experiment. A composite curve
in the urinary output and the specific gravity is shown in Figure 5.
From these studies it is evident that the subcutaneous injection of 1 c.c.
pituitary extract produced no appreciable effect on the urinary excre-
tion, on the daily or three-hourly amounts, or on the specific gravity.

Other experiments were carried out on dogs ; the water was given
with ])ituitary extract and in some instances repeated at varying periods
after the administration of the drug, in order to determine if diuresis

WEIR-LARSOX-ROn XT REE— DIABETES ISSIPIDUS

319

could be induced. Characteristic curves are reproduced in Figure 6.
It is evident that diuresis can be readily induced by forcing the water
intake after a single dose of 1 c.c. pituitary extract.

In Figure 7(a) is shown a composite curve of the results of a series
of experiments in dogs, in which the urine was collected in half-hourly
periods during the course of three hours. Each animal was studied
under three conditions: (1) normal conditions, (2) after administra-
tion of water by mouth in amounts corresponding to 50 c.c. per kilo-

.0.

~.

I

f

T

Ji

i

^ . ^ . /

t- 1— XX- X-J-

-^f ^iJ't

i "^C '

t xt

10 X At

1 ^^

3^-,^^=.^

^Z^"^

Fig. 6. — The antidiuretic effect of I c.c. pitui
overcome by repeated administration of water.

ary extract in dogs can Ije

gram of body weight, and (3) after the simultaneous administration
of pituitary extract, subcutaneously, and water by mouth in the amounts
already indicated. These experiments demonstrate that the subcu-
taneous administration of pituitary extract definitely checks the produc-
tion of water diuresis, although if the water is administered, urinary
excretion is on a somewhat higher level than normal.

The next step was an attempt to localize the site and mode of action
of the pituitary extract, that is, to ascertain the relation of pituitary

320 ARCHIVES OF IXTERXAL MEDICINE

extract to the nervous control of renal function. The dogs of the first
series were operated on and the splanchnic nerves to both kidneys
sectioned. ''' A second series of experiments identical with those
described was carried out (Fig. 7 [b]). A comparison nf the total

a

h

~\

\

\

\

^ ^

1

i' \

V

/

T

' ^

\

I A-

/ ,

\

/

X '

1

/

J _L

\

T

1

/

,

,

\ )',

'n^

T'

•

,-•■■

■-..

n

7t ~^

>■■'

••-..; -v

/ ■■■: ,^,

/

— ^i

'S>

/_

s

^

■^-=^ "

^^ -.

±i"r^::

Pituitary extract

Fig. 7. — Average half-hour rate ot secretion of urine in dogs under normal
conditions, after water, and after water and pituitary extract (a) before sec-
tion of the renal nerves (six dogs), and (b) after section of the renal nerves
(four dogs).

results under these two conditions is shown in Figures 7 and 8. From
these two series of experiments it is possible to deduce that (1) kid-
neys with nerves sectioned respond normally with a diuretic curve to

19. We wish to thank Dr. F. C. Mann and his associates for their kindii
in performing the operative work en the animals used in these experiments.

irElR-LARSOX-ROlVXTREF.—DlABETES IXSIPIDUS

321

water""; (2) this curve is largely abolished by pituitary extract, as in
normal kidneys (the diuretic response after pituitary extract admin-
istration is on a slightly higher plane than under normal conditions),
and (3) practically no change in the curve of the average half-hourly
output under the three conditions of the experiment result from section
of the renal nerves. It is evident, therefore, that the influence of
pituitary extract in the prevention of diuresis is independent of the
nerve supply of the kidney.

The Mechanism Involved in the Control of Diuresis by Pituitary
Extract. — It is obvious from our clinical studies in diabetes insipidus
and our experimental observations on the efifect of pituitary extract in

10

g

1

i

■f V

< o.

/

s

?n

/

^V

k

A

\

10

/

\

'C'^

y

Before operation-
After operation •

Fig. 8.— Curves showing average of secretion of urine in a series of dogs
under normal conditions, after water, and after water and pituitary extract,
before and after section of renal nerves.

the prevention of polyuria and our investigations (jf water intoxica-
tion that pituitary extract prevents polyuria. The mechanism of its
action is difficult to explain.

Earlier in our work we were inclined to believe that pituitary
extract raised the renal threshold for water and that a hydremic
plethora resulted. However, in one case of diabetes insipidus and in
several animal experiments, blood volume determinations failed to

20. Our results with the denervated kidney correspond to those already
described by Marshall and Kolls.

Marshall. E. K.. Jr., and Kolls. A. C: Studies on the Nervous Control
of the Kidney in Relation to Diuresis and Urinary Secretion, .'Vm. I. Physiol.
49:302, 1919.

i22 ARCHirES OF IXTERXAL MEDICJXE

indicate any increase of total blood or plasma volume. As a result of
our failure to demonstrate hydremic plethora and increased renal
threshold for water we are considering other possibilities.

In view of the results of our studies of blood chemistry and blood
volume we are at present attempting to determine if it is possible
that pituitary extract disturbs the relative ease with which the water
of the blood leaves the blood stream to pass through the kidney to
form urine, or to the tissues as body fluid. At present we are forced
to admit our inability to explain the antidiuretic efTect of pituitary
extract. A\'e are investigating the possibility of an abnormal escape
of fluid from the circulating blood into the tissues.

INTO.XICATIOX FOLLOWIXG THE .^DMI XlSTRATHlN OF W.\TER BY

MOUTH SUBSEQUENT TO GIVING. EXTE.\CT OF THE

POSTERIOR LOBE OF THE PITUITARY

SUBCUTANEOUSLY

One of our patients at our solicitation continued, after the adminis-
tration of pituitary extract, to take water in amounts to which he
had become accusomed " and during a period of eight hours he
ingested 5.25 liters, and excreted 800 c.c. of urine. In the course
of three or four hours he became very ill, was nauseated and developed
severe headache, and was forced to go to bed. Physical examination
revealed only puffiness of the lower eyelids and slight edema of the
ankles. Repetition of this experiment on the same and on another
patient gave similar results, that is, nausea, vomiting, and headache,
which forced the patient to go to bed. In one instance definite ataxia
appeared.

On account of the severity of the reaction no further experiments
were tried on patients, but similar studies were carried out on dogs.
Water in large quantities was given by mouth subsequent to the
administration of pituitary extract. This resulted in marked tremor,
salivation, and at times vomiting. The symptoms were independent of
the temperature of the water and were elicited with warm water as
rapidly as with cold. On forcing the experiment convulsions and
coma developed, and in one instance death supervened. As a result
of these observations we undertook an investigation to determine the
responsible factors, and studied separately the toxicity of water and
of pituitary extract, and later the toxicity of water following the
administration of pituitary extract.

Toxicity of Water. — In the classical experiments of Cohnheim and
Lichtheim''^ in which they demonstrated the absence of anasarca.

21. The edema and toxicity resulting in this ca.se were reported by one of
us before the Society of Qinical Investigations, Atlantic City, 1920.

22. Cohnheiin. I., and Lichtlieiin. L. : Ueber Hydr.imie und hvdrainisches
Oedcni. Arcli. t. path, .\iiat. u. Physic!. 69:106. 1877.

HEIR-LARSOX-ROUXTREE— DIABETES IXSIPIDUS 325

especially of edema of subcutaneous tissues when animals were given
physiologic salt solution intravenously in large amounts (up to 90
per cent, of body weight) cramps were reported in a few instances.
In an elaboration of this work by Magnus and Schafer -^ no mention
is made of these toxic manifestations.

Miller and Williams -* recently pyblished their results following
the administration of large quantities of water to patients with chronic
nephritis. They observed headache, dizziness, restlessness, chills, ful-
ness of the abdomen, vomiting, dyspnea, and cramps in the legs, which
were associated with marked increase in weight and definite eleva-
tion of blood pressure. In our experiments large amounts of water
were given to dogs without the development of untoward signs or
symptoms. One animal retained 2,500 c.c. during the course of seven
hours and passed 1,500 c.c. urine without manifesting symptoms of
intoxication. In another the administration of 3,500 c.c. resulted in
salivation, vomiting, asthenia, and ataxia.

One of our colleagues. Dr. Amberg, has kindly furnished us with
the following note, which is of interest in this connection. For the
purpose of studying the influence of water ingestion on the elasticity
of the skin as determined by the elastometer of Schade, two normal
adults drank from 2 to 3 liters of water (not cold) within about
fifteen or twenty minutes. For some time following the ingestion of
the water no elastometric determinations could be made, on account
of muscular twitchings. In the light of the experiments here reported
it appears not impossible that this was the result of a slight water
intoxication. There were no particularly uncomfortable subjective
sensations.

Toxicity of Pituitary lixtnict. — Schafer and \'incent -" state that
subcutaneous injections of pituitary extract to small mammals results
in paralytic symptoms similar to those observed after suprarenal
extracts. Cushny ^"^ states that large quantities can be injected without
producing symptoms other than somnolence and muscular weakness.

23. Magnus. R. : Ueber die Enstehung der Hautodeme bei experimenteller
hydramischer Plethora, Arch. f. exper. Path. u. Pharmakol. 42:250, 1899.
Magnus, R., and Schafer, E. A. : The Action of Pituitary Extracts on the
Kidney, J. Physiol. 27: Proc, ix, 1901.

24. Miller, J. L., and Williams, J. L.: The Effect on Blood-Pressure and
the Nonprotein Nitrogen in the Blood of Excessive Fluid Intake, Am. J. M. Sc.
161:. 327, 1921.

25. Schiifcr, E. A., and Vincent. S. : The Physiological Effects of Extracts
of the Pituitary Body, J. Physiol. 25:87, 1899.

26. Cushny. A. R.: A Textbook of Pharmacology and Therapeutics. Phila-
delphia, Lea & Fcbiger, 1910. p. 340.

324 ARCHirES OF IXTEKXAL MEDICI XE

According to Sollmann,-' ataxia and motor symptoms, changes in
carbohydrate metabolism, and emaciation develop after the excessive
and long continued use of pituitary extract.

In order to determine the etiologic relationship of the extract of
the posterior lobe of the pituitary to the symptoms observed in our
studies some experiments were carried out ; illustrative results appear
in the following protocols :

PROTOCOLS OF EXPERIMENTS
Protocol 1.— Dog 1, weight 9.8 kg. Pituitrin (Parke, Davis & Co.) sub-
cutaneously.

Feb. 11, 1921.— Pulse 100, regular.
10:00 a. m. : 1 c.c. pituitary extract.
10 : 30 a. m. : 1 c.c. pituitary extract.
11:00 a. m.: 1 c.c. pituitary extract; diarrhea.
12:00 m.: 1 c.c. pituitary extract; diarrhea.

1 : 00 p. m. : 1 c.c. pituitary extract.
1:30 p. m.: 1 c.c. pituitary extract.

1 : 45 p. m. : Pulse 48, sinus arrhythmia, extrasystoles ; bright and inter-
ested in other animals; no salivation; respiration 30.
2:25 p. m.: 1 c.c. pituitary extract.

3:15 p. m.: 1 c.c. pituitary extract. Pulse, 60; respiration, 40; appeared
sick; lost "pep"; no interest in other animals; sleepy; no
diarrhea or bladder frequency; no ataxia.
5:00 p. m.: 1 c.c. pituitary extract.
5 : 30 p. m. : 1 c.c. pituitary extract.
5:45 p. m. Animal more sleepy; no ataxia.

February 12: 10:00 a. m., pulse. 140, regular; animal active; appears nor-
mal. Intake from 6:00 p. m. to 10:00 a. m. 65 c.c; output. 300 c.c.

Protocol 2. — Dog 2. weight, 5.4 kg. Pituitary extract (Lilly Company)
intravenously.

Feb. 16, 1921 : Animal active; pulse. 80. regular.

2:25 p. m.: 3 c.c. pituitary extract; two minutes later inactive.

3:00p.m.: "Sick": at times seems unable to use hind limbs; lies flat

on side.
3:05 p. m. : Heart rate, 44; sinus arrhythmia; restless.
3 : 07 p. m. : Defecation.
3: 13 p. m. : 3 c.c. pituitary extract; lacks "pep"; diarrhea; pulse. 60; sinus

arrhythmia.
3:29 p. m. : 3 c.c. pituitary extract; no resistance: slight tremor; appears

weak.
3:31 p. m.: \omitcd 15 c.c. yellow fluid, frothy.
3:40 p. m. : Pulse 62; reflexes active; vomiting again.
3:45 p. m.: 3 c.c. pituitary extract; several attacks of vomiting and bowel

movements.
4:(K) p. m.: 3 c.c. pituitary extract; feces and vomitus bile stained; mucus.

present.
4: 15 p. m. : 3 c.c. pituitary extract; inactive; does not resist; pulse, 80.
4:40 p. m.: Greater activity.
February 17, 10:00 a. m. : Normal activity.

27. Sollmann, T. : .'\ Manual of Pharmacology and Its .^pplicati
Therapeutics and Toxicology, Philadelphia. \V. B. Saunders Company
p. 344.

11 EIR-LAI<:iO.\-l<Oll.\Th'EE— DIABETES IXSIPIDUS 325

Toxicity of IVatcr After the Administration of Pituitary
Extract.-^ — Subsequently animals were given pituitary extract sub-
cutaneously and large amounts of water through a stomach tube.
Blood volume determinations and studies of the Wood chemistry were
made before the administration of pituitary extract, and again after
the onset of the symptoms. The summary of the results is given in
Table 6.

Early Manifestations of Toxicity. — The first symptoius noted may
possibly be due to pituitary extract alone, namely asthenia, restlessness,
frequent attempts at urination, diarrhea, and vomiting. Other symp-

TABLE 6. — Blood Volume ix \\'.\ter Intoxication

H

£«

t:t

<k

1=^

¥

5=i _

>S

u

OO

i§

M c! u

°i i =

= o. 5..

\£

it

2o s

4 l5

_ = Z 5,'

1

1

=1

.><yiiiptoms

1

1

-a ^

t -p

fj^

1

5

1

oE
HO

V

Im

10.4

V 6,400
(4+3) (4,000+

(i:r+ '^

613

1,251

59

120

0.763

,»

Drowsiness, salivation,
restlessness. Frequent

2,400)

1,300) 51

«01

1,178

58

113

0.743

430

attempts at urina-
tion, vomiting,
ataxia, convulsions.

Tim

-A

3

l,S0O

500 46

370

794

50

107

Drowsiness, salivation.

1 46

389

840

53

114

tremor, ataxia, vom-
iting, convulsions,
coma, involuntary
evacuations of bowels
and bladder

E200

9.0

5

SfiOO

2,000 53

618

1.157

69

128

0.490

460

Preoperative: Marked

67

526

923

58

102

0.595

390

frequent attempts at
urination; diarriiea,
drowsiness, rcstless-

twltching's, ' saliva-

t i o n, convulsions.

E290 8.5

3

3.150

2,325 1 65

SVO

1,086

87

122

Postoperative

!S7

650

1,140

76

134

.....

Same symptoms

toriis, tremor and salivation, next appear, possibly due to water alone.
These develop usually after approximately 300 c.c. of water has been
given per kilogram of body weight as a minimum. This stage con-
tinues for a variable period when another set of symptoms develops.
The animal becomes very drow.sy, and later shows muscular twitchings
and ataxia on standing or walking. This may be designated as the
preconvulsive stage. In spite of mental dullness the animal evidences
hypersensitiveness to external stimuli, .such as result from picking up
the stomach tube or passing it. .Such procedures may precipitate the
convulsive stage.

28. Water inlrorluced intravcnouisly as physioldgic solution of sodium chlorid
results in an entirely different set of phenomena, the dogs dying with edema of
the lungs.

326 ARCHIVES OF IXTERXAL MEDICIXE

Later Manifestations of Toxicity; Convulsions, and Coma. — Con-
vulsions were characteristically epileptiform in type ; first a tonic stage,
followed immediately by a clonic stage. During the tonic phase the
head retracted, the jaws set, respiration ceased, and the animal became
cyanosed. Then sudden and violent clonic spasms developed, accom-
panied by frothing at the mouth and involuntary bowel and bladder
evacuations. During the convulsive stage the pupils were markedly
dilated, contraction occurring immediately after the cessation of the
clonic spasms. During or following the convulsive stage the dog
usually showed running, swimming, or snapping movements and at
times barked and growled, as though participating in a fight. Coma
or somnolence persisted for a longer or shorter period, but the animal
usually recovered fully during the course of the next few hours.
Passage of the stomach tube and the administration of more water
brought on further convulsions, which, if continued, ended in death.
Vomiting and salivation occurred usually after the administration of
from 200 to 300 c.c. water per kilogram of body weight. Occasionally
an animal failed to manifest salivation. Some animals made frequent
attempts to urinate while others apparently exhibited no evidence of
bladder irritability.

Several animals in which double renal nerve section had been per-
formed were given pituitary extract and water. The same toxic
symptoms developed but no increase or decrease in the facility with
which they could be produced was noted. This further indicates that
the antidiuretic action of pituitary extract is independent of the nerve
supply of the kidney. A protocol of one of these experiments is
presented below :

Protocol 3. — Dog E249, weight, 11.4 kg."^ Water by mouth sttbsequent to
pituitary extract subcutaneously.

Feb. 7, 1921.-11:50 a. m. : 1 c.c. pituitary extract and 500 c.c. water.

1 : 45 p. m. : 1 c.c. pituitary extract and 500 c.c. water.

2:45 p. m.: 1 c.c. pituitary extract and 500 c.c. water; some frequency ot

micturition.
3:15 p. m.: Has vomited 1,000 c.c.
3:25 p. m. : Animal found in cage, unconscious, frothing at mouth, head

retracted; jerking of jaw and leg muscles. On being taken

from cage, was unable to walk, lay on side and went through

running movements, jaws still snapping. Prostration marked.

movements easily restricted; semiconscious.
3:40p.m.: Involuntary bowel evacuation, liquid. Running movements

stopped ; attempts to walk unsuccessful.
3 ; 45 p. ni. : Conscious, hut lies quiet : unable to attract animal's attention ;

few attempts at walking movements: respiration, 70; muscles

rather flaccid.

29. Other experiments are in progress which involve more careful con-
sideration of the weight of the animal. To dale increases in weight greater
than 1 kg. have not been encountered.

II ElR-LARSOX-KOnXTREE— DIABETES IXSIPIDUS

327

3: 57 p. m.

Pulse, 100; pupils small; twitching muscles; feet lie as if
asleep; ej-es open; listless; few attempts to get up; res-
piration, 32.

Gets up. walks in circular course, lies down.

Diarrhea, hardly able to stand up, does not respond to call,
is not frightened.

\'arious attempts to rise.

Walks, seems blind, bumps into objects.

Walks again, bumps into objects; weight. 11.6 kg.

Walks until exhausted ; defecation.

Improving; vision returning; attention can be attracted slightly.

Pulse 108, deep respiration; frequent urinations, little or no
urine.

500 c.c. water; very restless, running.

Defecation.

Running less^rapidly. slightly ataxic, salivated.

Running less rapidly, walks some, more ataxia, lists to left.

Complete ataxia, unable to stand.

Convulsion, head retracted : tonic and clonic phases, running
movements; vomited; salivated.

Convulsion.

Convulsion.

Convulsive movements.

Has been in continuous convulsions, unconscious, continuous
snapping movements of jaws, pupils widely dilated.

Same condition, more convulsions, pupils contracted after con-
vulsion, dilated during convulsion.

Died in convulsion.

revealed early onset of rigor mortis, normal chest and
lungs, and slight congestion of the liver and kidneys. The stomach
contained about 65 c.c. fluid and the intestine about 25 c.c. No free
fluid was found in the pleural or peritoneal cavities, and there was
no evidence of edema of the subcutaneous tissues of the neck or of
the extremities. The brain, on removal, appeared to be normal in
every respect.

Thus we have definite evidence of toxic symptoms produced by
water, when the water-secreting function of the kidney is diminished
through the subcutaneous administration of 3 c.c. pituitary extract.

While the experiments were in progress an attempt was made to
determine the mechanism of production of the convulsions. Kymo-
graphic records were procured in some of these experiments. A
typical tracing is presented (Fig. 9). Hydremic plethora as a cause
was first considered, but no increase in the total blood or plasma
volume by the vital red method, nor any constant increase in relative
plasma volume by the hematocrit could be demonstrated (Table 6).

The total nitrogen and the sodium chlorid content of the plasma
showed usually a slight decrease after the onset of symptoms. The
urine volume during the following twelve hours was always increased,
usually to about 500 c.c, indicating a rather rapid excretion of the
retained fluids. Clinical evidence of edema was absent in all the
animals studied.

59 p.

m.

02 p.

m.

06 p.

m.

08 p.

m.

15 p.

m.

18 p.

m.

25 p.

m.

27 p.

m.

45 p.

m.

48 p.

m.

50 p.

m.

52 p.

m.

55 p.

m.

56 p.

m.

03 p.

m.

04 p.

m.

12 p.

m.

20 p.

m.

30 p.

m.

45 p

m.

Necrops}

llEIR-LARSOX-ROIiWTREE— DIABETES IXSIPIDUS 329

The Cause and Significance of Convulsions and Coma. — The imme-
diate cause and the mechanism of production of these phenomena have
not been determined. But it has been demonstrated that neither
pituitary extract nor water alone ^" in similar, or even larger amounts,
is capable of producing them. Edema of the brain, suggests itself as
a logical explanation, but we have not succeeded in proving this to
our own satisfaction. Table 6 presents the plasma and blood volume
values as determined before and after water intoxication. From this
it is apparent that increased blood volume has been excluded in at
least some of our experiments. A marked increase in blood pressure
is encountered in water intoxication. The rise is tardv in onset and

TABLE 7.— Blood Pressure Changes in Water I

NTOXIC.\TION

Water,

Weight,

PituItriD,

Blood

Respira-

Time

C.c.

Kg.

C.c.

Pressure

Pulse

tion

13.6

140

n:-20 a. m.

...

13.6

1 "0"
Snbcutaneously

140

n:30 a. m.

700

14.3

140

11M3 a. m.

460

14.6

160

11:50 a. m.

14.6

1 "0"
Subcutaneously

150

72

22

l2:eo m.

15.6

140

l:00 p. m.

15.6

Subcutaneously

140

l:13 p. m.

1,000

16.4

140

2:05 p. Di.

500

16.9

Subcutan<ously

150

70

24

3:05 p. m.

16.8

Subcutaneously

170

3:10 p. m.

300

17.2

180

3:30 p. m.

0

Intravenously

190

4:00 p. m.

1,000

18.1

206

72

IS.l

4:40 p. m.

18.1

Intravenously

200

82

20

t:T« p. iti.

18.1

140

22

•■

gradual in development and may reach a comparatively high level
(Table 7; Fig. 9). However, the changes in blood pressure do not
appear to be striking enough to account for the toxic manifestations.

Nausea, vomiting, muscle twitching, spasms, asthenia, convulsions.
and coma developing in the presence of an obvious derangement of
water excretion strongly suggest uremia, especially when death super-
venes. But in the ab.sence of increased values for blood urea, and in
view of the rapidity of their appearance and disappearance, it does
not seem at all probable. However, it is of more than passing interest
to learn that the entire series of symptoms so frequently found in
uremia can result from a common cause. From the standpoint of
sudden onset and recovery eclampsia is also suggested.

.W. Since this paper was submitted for publication, further experiments liav(
licen carried out and convulsions have been produced with water alone. Tin
results of these further studies will be published later.

330 ARCHIVES OF IXTERXAL MEDICIXE

Convulsions of sudden onset characterized by a tonic and a clonic
phase, associated with involuntary evacuations, and frothing at the
mouth, followed by coma, and ending in complete recovery suggest
epilepsy. On the other hand, nausea, vomiting, asthenia, and ataxia
characterizing the earlier period of development of these phenomena
have no resemblance to epilepsy. The possibility of reflex factors
from the gastro-intestinal tract at least must be considered, since
strangury and retching constitute marked features at times. A condi-
tion resembling status epilepticus and ending in death has also been
observed.

In the absence of definite information relative to the seat, origin,
and mechanism of production of these phenomena speculation is futile;
it would be wiser perhaps to leave all such questions to the future.

CIRCULATORY COMPENSATION FOR DEFICIENT

OXYGEN CARRYING CAPACITY OF THE

BLOOD IN SEVERE ANEMIAS*

GEORGE FAHR and ETHEL ROXZONE

MADISOX. WIS.

It is a well known fact that persons suffering from severe chronic
anemias do not show the signs of anoxemia when at rest even when
the oxygen carrying capacity of each cubic centimeter of blood is
below the normal venous unsaturation, or, in other words, when the
oxygen content of each cubic centimeter of arterial blood is less than
the average amount of oxygen normally abstracted from the blood
by the tissues during its passage through them. We recently studied
a case of pernicious anemia in this hospital in which with a hemoglobin
of 12 per cent.^ and an oxygen carrying capacity of only 2.2 c. c. per
hundred cubic centimeters of blood there was no dyspnea, no acidosis,
no increased pulse rate and a normal basal metabolism or rate of oxygen
consumption. As it has been frequently shown that the normal resting
human organism abstracts an average of 5.5 c.c. oxygen from the blood
in the capillaries, it was necessary to explain how anoxemia was
avoided in this case, with normal basal metabolism or oxygen con-
sumption, and an oxygen content per cubic centimeter of arterial blood
less than half the amount normally abstracted in the capillaries for
the processes of oxidation.

The explanation of this condition in our case contains the solu-
tion of the problem of the compensatory mechanism in severe anemias,
and we feel that it is of sufficient interest to justify publication, espe-
cially as it emphasizes the necessity for bed rest in severe anemias
and introduces a method of minute volume determination on man which
may be of value in cardiovascular investigation.^

* From the Bradley Memorial Hospital, University of Wisconsin.

1. All hemoglobin determinations were done by the method of Palmer.' In
this method a standard blood is saturated with oxygen and its oxygen capacity
determined. An oxygen capacity of 18.5 c.c. per hundred cubic centimeters
blood is given a hemoglobin value of 100 per cent. Having determined the
oxygen capacity and hemoglobin value for the standard blood, this blood is
now colorimctrically compared to the patient's blood by means of the Haldane
carbon dioxid method. In this way the percentage of hemoglobin always has
a definite relation to oxygen carrying capacity and the values of different
observers working in different clinics are based on the same 100 per cent.
standard.

2. J. Biol. Chem. 33:119. 1918.

3. This method was first suggested by Pick, Ges. Werke 3:573.

330 ARCHIVES Of JXTERXAL MEDIC I SE

Convulsions of sudden onset characterized by a tonic and a clonic
phase, associated with involuntary evacuations, and frothing at the
mouth, followed by coma, and ending in complete recovery suggest
epilepsy. On the other hand, nausea, vomiting, asthenia, and ataxia
characterizing the earlier period of development of these phenomena
have no resemblance to epilepsy. The possibility of reflex factors
from the gastro-intestinal tract at least must be considered, since
strangury and retching constitute marked features at times. A condi-
tion resembling status epilepticus and ending in death has also been
observed.

In the absence of definite information relative to the seat, origin,
and mechanism of production of these phenomena speculation is futile ;
it would be wiser perhaps to leave all such questions to the future.

CIRCULATORY COMPENSATION FOR DEFICIENT

OXYGEN CARRYING CAPACITY OF THE

BLOOD IN SEVERE ANEMIAS*

GEORGE FAHR and ETHEL RONZONE

MADISOX. WIS.

It is a well known fact that persons suffering from severe chronic
anemias do not show the signs of anoxemia when at rest even when
the oxygen carrying capacity of each cubic centimeter of blood is
below the normal venous unsaturation, or, in other words, when the
oxygen content of each cubic centimeter of arterial blood is less than
the average amount of oxygen normally abstracted from the blood
by the tissues during its passage through them. We recently studied
a case of pernicious anemia in this hospital in which with a hemoglobin
of 12 per cent.' and an oxygen carrying capacity of only 2.2 c. c. per
hundred cubic centimeters of blood there was no dyspnea, no acidosis,
no increased pulse rate and a normal basal metabolism or rate of oxygen
consumption. As it has been frequently shown that the normal resting
human organism abstracts an average of 5.5 c.c. oxygen from the blood
in the capillaries, it was necessary to explain how anoxemia was
avoided in this case, with normal basal metabolism or oxygen con-
sumption, and an oxygen content per cubic centimeter of arterial blood
less than half the amount normally abstracted in the capillaries for
the processes of oxidation.

The explanation of this condition in our case contains the solu-
tion of the problem of the compensatory mechanism in severe anemias,
and we feel that it is of sufficient interest to justify publication, espe-
cially as it emphasizes the necessity for bed rest in severe anemias
and introduces a method of minute volume determination on man which
may be of value in cardiovascular investigation. ■■

* From the Bradley Memorial Hospital. University of Wisconsin.

1. .MI hemoglobin determinations were done by the method of Palmer.' In
this method a standard blood is saturated with oxygen and its oxygen capacity
determined. An oxygen capacity of 18.S c.c. per hundred cubic centimeters
blood is given a hemoglobin value of 100 per cent. Having determined the
oxygen capacity and hemoglobin value for the standard blood, this blood is
now colorimctrically compared to the patient's blood by means of the Haldane
carbon dioxid method. In this way the percentage of hemoglobin always has
a definite relation to oxygen carrying capacity and the values of diflferent
observers working in different clinics are based on the '^ame 100 per cent.
standard.

2. J. Biol. Chem. 33:119, 1918.

3. This method was first suggested by Pick, Gcs. Werke 3:573.

332 ARCHIVES OF IXTERXAL MEDICIXE

REPORT OF CASE

History.— A male, aged 49. entered the hospital complaining of pain in epi-
gastrium and weakness, progressively increasing during past five years. He
had a sore mouth two years ago. The skin had been yellow for years. He
now complains of paresthesia in the legs.

Examination. — Lemon colored skin, with large pigment patches over neck
and lower arms; no emaciation; petechiae on various parts of the body; mucous
membranes pale ; tongue smooth. Lungs normal. Heart apparently not
enlarged to percussion. Systolic murmur present, loudest just inside apex.
Systolic murmur at base is transmitted into carotids. Venous hum over jugulars.
Spleen just at rib margin, moves 2 cm. below on deep inspiration. Tendon
reflexes exaggerated, vibratory sense lost in legs. Sense of position question-
able in legs.

Blood Findings on Entrance— RtA Cells, 1.600,000; hemoglobin, 42; index,
L3; white cells, 3,600. Smears show many poikylocytes, anisocytosis, poly-
chromatophilia and stippling, many megalocytes ; no normoblasts, polymorpho-
nuclears, 62 per cent. ; lymphocytes, 33 per cent. No platelet count was made
on entrance, later on when the hemoglobin had fallen to 12 per cent, the
platelet count was 22,000; 4 per cent, reticulates. Normoblasts and megalo-
blasts were seen frequently in later smears.

Stomach Content After Boas Meal. — No "free acid" and a total acidity of
1 per cent.

Fragility test showed beginning hemolysis at 0.5 per cent, and complete at
0.35 per cent. The daily average of output of urobilin was 25.000 units.

Despite transfusions and regulation treatment, the patient's con-
dition gradually became worse, and eight months after entrance the
patient died with a hemoglobin content of only 8 per cent. Throughout
a period of two months the hemoglobin content was 12 per cent, and
the red count was about 550,000. It was during this period that the
laboratory data were obtained. The necropsy confirmed the diagnosis
of pernicious anemia.

Necropsy revealed a dilated heart with moderate increase in the
thickness of ventricular musculature. Microscopic sections of heart
muscle showed both hypertrophic and atrophic fiber. In certain areas
there was vacuolization of this fiber. Tiiere was every evidence that
death was due to circulatory failure. There was about one liter of
transudate in each thoracic cavity. There was passive congestion of
liver and spleen. The pathologist said that death was due to failure
of the circulation.

April 3, 1921, two basal metabolism determinations were done
with the Benedict unit apparatus and were found to be 1,454 calories,
or 4 per cent., higher than the average normal value as predicted
according to DuBois * and 15 per cent, higher than the basal metabolism
as predicted according to Benedict.' Certainly the basal metabolism
was not slowed up in this case. A few days later basal metabolism
determinations were repeated and gave 1,451 calories. The oxygen

4. DuBois, D., and DuBois, E. F. : .■\rch. Int. Med. 17:863 (July) 1916.

5. Benedict and Harris: Biometric Study of Basal Metabolism in Man. 1919.

FAHK-ROXZOXE— BLOOD IS AXEMIAS o3i

consumption as determined by the apparatus was 213.5 c.c. per minute.
If now we can find the oxygen abstracted from each cubic centimeter
of blood leaving the lungs and passing through the peripheral circu-
lation back into the right heart we can very easily calculate the number
of cubic centimeters of blood passing through the lungs per minute or
the number of cubic centimeters of blood leaving each chamber of the
heart per minute. It is thus necessary to know the oxygen content
of the arterial blood and the oxygen content of the venous blood. The
oxygen content of the arterial blood is given by the hemoglobin deter-
mination according to the Palmer method, it is true the Palmer method
gives the maximum oxygen content of the blood at atmospheric pres-
sure of oxygen but tlTis value has been shown by experiment to be
only 5 per cent, above the value for arterial blood as actually deter-
mined. The Palmer determination on this blood, April 2, gave hemo-
globin 12.2 per cent. The total combined and dissolved oxygen was
2.4 c. c. per hundred c. c. of blood. The determination of the mean
oxygen content of the venous blood contains a slight error. We took
the blood from the arm vein and determined its oxygen content accord-
ing to the Van Slyke method.'' \N'e found this blood to contain 0.6 c. c.
oxygen per hundred cubic centimeters blood.

The oxygen content of the venous blood varies slightly according
to the part of the body from which it comes. During absolute rest
in bed and before breakfast the blood coming from the heart will have
more oxygen abstracted from it than blood which has passed through
the capillaries of resting organs, like the arms, legs and trunk. The
venous blood coming from the brain will also contain less oxygen than
blood coming from an active organ like the heart. Metabolism in
the portal area is not at its height during morning rest. It is safe
to assume that although an active organ like the heart might possibly
under the conditions of a very low oxygen content of the arterial
blood abstract nearly all the oxygen, yet it is not very probable that
this would occur, for a glance at the dissociation curve of oxyhemoglo-
bin " shows that when the oxygen content has dropped to 10 per cent.
of its normal maximum content * then the diffusion pressure is only
10 mm." or about one tenth of the mean diffusion pressure in the
capillaries and one fifth the average diffusion pressure in normal venous
blood. The speed of diffusion is lowered to a point where oxygen must
leave the blood in the capillaries very slowly and it is doubtful if
under the conditions of increased velocity of flow in this case, which
we shall prove in this paper, the last vestiges of oxygen can be

6. Van Slyke: J. Biol. Chem. 33:127. 1918.

7. Bohr. Hasselbach and Krogli : Scand. Arch. f. Physiol. 16:402. 1904.

8. In this case 0.4 c.c. per hundred centimeters blood.

9. Assuming that the partial pressure of oxygen in the tissues is zero.

336 ARCHIVES OF IXTERXAL MEDICINE

to Rowntree's '" method gave 5.2 per cent of body weight for plasma
volume or a little more than the normal average for plasma volume.
Because of the small corpuscular volume the total blood volume was
only 5.4 per cent, of the body weight, or considerably less than the
average normal. Therefore, the total blood content of the tubing was
less than the normal, and it would seem hardly probable on first thought
that the eilfective cross section of the tubing could be larger. Besides
we know that the filling of the heart must be considerably greater in
order to get the large output we have calculated. Moreover, frequent
examinations of the finger capillaries with the method of Lombard ''
showed both a very marked contraction of the capillaries and fewer
open capillaries. It is impossible to check up the lumen of all the
vessels but the above observations do not mitigate against the assump-
tion that the effective cross section of the vessels was greater, thus
causing increased velocity, and at the same time the total volume content
of the whole vascular tubing was diminished to correspond to the
decrease in blood volume and the increased filling of the heart itself.
For Poisseulle's formula for the velocity of blood flowing through
tubing is V — j'-P'^t ^^[-,£^6 v is velocity, k is the reciprocal of viscosity,
p is blood pressure, r is the radius of the section of tubing imder
consideration, t is the time and 1 the length of the tubing. As the
radius is in the numerator in the fourth power and 1 is in the denomi-
nator in the first power it is easy to see how by increasing the diameter
of the short capillary area and at the same time diminishing the lumen
of the long arterial and arteriolar area both the effective resistance and
the total volume of the tubing may be reduced at the same time.

Our capillary observations showed that some of the skin capillaries
are remarkably reduced in diameter. We believe that this is one of the
compensatory factors for securing increased flow of blood in more
important organs where metabolism is greater. As observed by the
microscope the flow in the skin capillaries was very slow but this is
easily explained for the diameter of these capillaries was less than
half of that of other patients and normals. Toward the end it was
approximately one third the diameter of our own capillaries. On the
other hand, the vital organs at necropsy impressed the pathologist as
being more than normally filled with blood.

It is interesting to calculate the work done by the heart per minute
and its oxygen consumption. In calculating the work of the heart we
shall make use of Evans''" formula ^^'— g- Q R -|--^, where Q is the
minule volume, R is the combined aortic and pulmonic blood pressures

16. Rowntree. Geraghty and Keith: Arch. Int. Med. 16:547 (Oct.') 1915.

17. Lombard: Am. J. Physiol. 29:.US. 1912.

18. Evans: J. Physiol. 52:6. 1918.

FAliR-ROSZOSE— BLOOD IS ASEMIAS 337

in terms of a water column/^ V is the linear velocity calculated from
the aortic and pulmonic cross section, the minute output, and the actual
time during which blood is flowing out of the ventricles, and G is the
constant of gravity acceleration. We find that the heart performs 18.5
kg.m. of work per minute necessitating an oxygen consumption of
30.5 c. c. per minute.-" or one seventh the total oxygen consumption of
the body. If we assume that the venous blood leaving the coronary
veins has an oxygen content of only 0.2 c.c. per hundred cubic centi-
meters, then we can reckon that the coronary circulation is 1.48 liters
per minute, an enormous flow. Even if we assume that the oxygen
content of the whole venous blood of our patient was 0 we are
compelled to calculate a minute volume of 9.5 liters for the circulation
and an oxygen consumption in the heart of 26 c. c. each minute. Even
under these conditions the coronary flow would be 1,190 c. c per minute.
It would be necessary to have a coronary circulation of 1.2 liters per
miiuite to supply even this oxygen need. Evans has shown that the
heart during the performance of very severe work must have a coro-
nary circulation of about 850 c.c.-' With blood of the viscosity of our
patient's such a flow would become 1.87 liters, so that it is not at
all impossible as at first it might seem. Of course, it is possible that
the efficiency of the heart is greater than 30 per cent., ^ut from
Evans' -• work we would assume that 30 per cent, is the highest mechan-
ical efficiency of the human heart.

Our calculation shows that the heart may very easily suffer from
lack of o.xygen. especially if the patient is not at absolute rest. There
is very good support in this paper for the pathologist's contention
that the heart muscle changes of pernicious anemia are due to lack
of oxygen. ^^

CONCJ.USIONS

1. In severe anemias increased minute volume is the outstanding
compensatory mechanism for loss of oxygen carrying power of the
blood.

2. In a case of severe pernicious anemia the minute volume was
increased about 250 per cent, and the systolic output in the same degree.

19. Evans uses 1.7 for R, in persons of normal blood pressure = 120. As
the blood pressure in our patient was 105 we have used 1.5 for R.

20. Assuming that the mechanical efficiency of the lieart is 30 per cent, and
respiratory quotient 0.8.

21. Evans: J. Physiol. 47:407, 1914.

22. Evans: J. Physiol. 52:6. 1918.

23. Laboratory data of interest : CO.- tension in alveolar air 37 mm. =5.2 per
cent. Hydroxybutyric acid in blood 1.2 mg. per hundred cubic centimeters. Other
acetone bodies, 0.53 mg. per hundred cubic centimeters. COj combining power
of blood, 62.6 c. c. per hundred cubic centimetres plasma. CO: content of
blood, 51.3 c.c. per hundred c.c. plasma.

338 ARCHIVES OF IXTERXAL MEDICIXE

3. The increased blood velocity was very largely due to lowered
blood viscosity, this being lowered to 45 per cent, of its normal value.

4. Another factor was increased effective cross section of the vas-
cular tubing.

5. Microscopic examination of the skin capillaries showed that
they were contracted down to half the normal diameter or less, thus
determining a lessened blood flow through the skin and a larger flow
through other organs. The lessened quantity of blood in the skin
is certainly one factor in the degree of paling of the skin.

6. The coronary circulation is at the upper limit of the possible
being about as large as that found in very severe work. There is a
very great possibility that when a patient with severe anemia tries to
work that anoxemia of the heart muscle is produced. The pathologic
changes in the heart muscle in pernicious anemia may well be due to
lack of oxygen.

A HITHERTO UNDESCRIBED TUMOR OF THE
BASE OF THE AORTA*

GEORGE R. HERRMANN, M.D.

AND

MONTROSE T. BURROWS. M.D.

ST. LOUIS

Tumors of the heart are not commonly met witli and their clinical
diagnosis up to the present time has been difficult, if at all possible,
to make. Tumors of the aorta are much less frequent than those of
the heart. Newer clinical diagnostic methods are promising much,
however, in the early recognition of these cases. The tumor of the
aorta attached to the heart which we found in our case has no counter-
part in the literature. It became of interest, therefore, to describe
this case not only from its clinical but also its anatomical aspects.

REPORT OF CA.SE

N. R., a white male, married, aged 54. was admitted to Barnes Hospital on
the service of Dr. George Dock for the first time, June 26. 1918. The chief
complaint at that time was shortness of breath occurring in paroxysmal attacks
off and on for over fifteen years.

Family History. — This was unimportant, except that one brother, who had
died of pneumonia, was supposed to have had "heart trouble."

Personal History. — The marital history was insignificant. The patient had
been married eight years, he had one child, a boy of 6 years. living and well.
His wife was living and well and had had one miscarriage.

Past History. — He had not had chorea, diphtheria, scarlet fever, pneumonia
or tonsillitis. He had had measles, pertussis and typhoid fever in childhood
but had recovered from these without compHcation. He had had gonorrhea
at 28, and a "hard chancre" of the urinary meatus, with bubo and inflammation
of the testicles, at 29. He had been treated for syphilis by local applications
and by mouth. No secondary lesions developed. "Smothering attacks" with
shortness of breath, which he termed "asthma" had troubled him for fifteen
years. These attacks would come on after exertion and were accompanied
by palpitation and the appearance of a dusky gray color. He had taken three
or four glasses of whisky, smoked two to three pipefuls of tobacco, and drank
two to three cups of coffee daily for years.

Present Illness. — The trouble that brought him to the hospital was more or
less dyspnea which gradually became exaggerated following an attack of
"rheumatism" in February, 1918. He began to notice difficulty in breathing and
wheezing, particularly at night, when lying in bed ; no jialpitation or pain was
noticed and the symptoms were relieved by his getting on his feet. The attacks
of dyspnea on exertion which he had had for years had grown more severe.
His feet had been swollen for a week and there had been blueness of the lips
and face during several of the attacks. The patient fainted once. He had
also had pain in the epigastrium which radiated down into the abdomen.

Physical Examination. — "Xhe patient had a robust frame with moderately
heavy musculature and thick panniculus, weighed 185 pounds and measured
70 inches in height. He walked into the ward but was very dyspneic. There

♦From the Departments of Internal Medicine and Surgery. Washington
University School of Medicine.

340 ARCHirES OF JXTERXAL MEDIC IX E

was a dusky ashen pallor about his face while the rest of the body was
cyanotic. On reclining the face became very cyanotic. Orthopnea was noted
and the respiration was Cheyne-Stokes in type. The patient would drop off
to what appeared to be sleep during the apneic periods. The veins of the
neck were slightly engorged but there were no abnormal pulsations. The lungs
were negative, except for many crackling rales at both bases posteriorly.

Heart : The heart was defintely enlarged. The apex impulse was not well
localized but was felt best about IS cm. to the left of the midsternal line in the
fifth intercostal space. There was a general precordial heave without a strong
impulse on palpation. No thrills were felt. The cardiac outline to percussion
was 2.5 cm. to the right and 2.5 cm. to the left of the midsternal line in the
first intercostal space ; 3.5 cm. to the right and 4.5 cm. to the left in the second
intercostal space; 4 cm. to the right and 9 cm. to the left in the third inter-
costal space. What was considered to be liver dulness was elicited on the
right side below the third intercostal space, while to the left the figures of the
outer cardiac dulness were : 14 cm. in fourth intercostal space ; 16 cm. in fifth
intercostal space, and 17 cm. in sixth intercostal space.

The contractions were regular and rapid. A systolic murmur was heard at
the apex. This was transmitted toward the axilla, but not toward the sternum.
Both sounds were accentuated at the apex. The pulmonary second sound was
accentuated and louder than the aortic second sound. The blood pressure was
from 110 to 120 mm. Hg. systolic, and from 100 to 85 mm. Hg. diastolic.

The pulse was regular and rather small. The radial and brachial artery
walls were very much thickened, but not calcareous. The feet and legs were
oedematous.

Abdomen : The abdomen was aliove the costal margin. There was disten-
tion of the superficial veins in the left hypochondrium. Tenderness was
elicited in the epigastrium. The liver edge was found to be 8 cm. below the
costal margin in the midclavicular line. Tenderness was noted all over the
enlarged liver.

There was a linear 2 cm. scar of the old bubo in the left groin. The right
lip of the urinary meatus was large and verrucous. The patient said that the
"hard chancre" had been at this site. A small varicocele was found on the
left side. The prostate was of moderate size, thickened throughout and adherent.

No abnormal reflexes or neurologic signs were elicited.

The patient was seen by Dr. G. Canby Robinson who found a general pre-
cordial heave in the region of the apex with no local heave but a definite tap.
The heart borders extended 5.5 cm. to the right and 16.5 cm. to the left of
the midsternal line. The heart sounds were blurred and distant. \ faint
nontransmitted systolic murmur was heard at the apex. The pulse was palpable
with difficulty and counted 116 per minute. The pulmonary second sound was
accentuated.

The electrocardiogram showed right ventricular preponderance and the
chest plate showed right sided hypertrophy. Aside from these there were,
however, no other signs of mitral stenosis.

.\ dose of 10 c.c. tincture of digitalis was administered and the regular
rhythm became absolutely irregular within twenty-four hours. .\ slight pulse
deficit was noted at that time and the difference in the force was more striking
than the arrhythmia, but no pulsus alternans was ever demonstrated. The pulse
later became regular and the edema disappeared. The liver, on the other hand,
at the time of discharge remained 8 cm. below the costal margin in the para-
sternal line. The patient was given a prescription for 10 minims tincture of
digitalis to be taken three times a day and discharged improved, July 20, 1918,
to return to the Outpatient Department for observation.

Clinical Laboratory Fiiirfiiiff.t.— Blood : The red blood cells nuinbercd 4,472,-
000: wliite blood cells, 7,100; hemoglobin, 75 per cent. The differential smear
showed polymorphonuclear neutrophils. 7i per cent ; lymphocytes, 20 per cent.,
and large mononuclears and transitionals, 6 per cent. The blood Wassermann
reaction was negative. The blood nonprotein nitrogen was 78 mg. per hundred c.c.

HERRMAXX-BURROIVS— TUMOR OF AORTA o41

Urine: The specific gravity varied from 1.018 to 1.026. The volumes were
small. Occasionally the night volume was increased over the day volume.
There were also traces of albumin, and numerous hyaline and finely granular
casts. Fifty-three per cent, of plienolsulphonephthalein was e.xcreted in two
hours.

Electrocardiograms :
1774; June 27; auricular flutter; well marked right sided preponderance.
1770; June 28; auricular flutter; well marked right sided preponderance.
1786; July 1; auricular fibrillation (24 hours after administration of 10 c.c.

tincture digitalis).
1797; July 5; auricular flutter in part of the record; impure in another.
1801; July 7; auricular flutter; arrythmia suggesting varying degrees of block;

auricular rate. 250; ventricular rate, 109.
1803; July 10; normal rhythm reestablished; right ventricular preponderance
(suggesting mitral stenosis); S wave in Lead 1 very short again;
negative T waves in all leads: diphasic P in second and negative P
in third leads ; rate. 102 ; P.R. 0.22 sec ; Q R S. 0.08 second.
1815; July 11; one left ventricular extrasystole, otherwise as 1810. Deptli of
S wave in Lead 1 increasing.

There were no changes in the subsequent curves taken while the patient
was in the hospital. The numbers and dates of these were as follows: 1824.
July 13; 1831, July 15; 1837, July 16; 1843, July 17; 1849, July 19; 1855, July 20!

Record of Outpatient Department : The patient was seen at frequent
intervals in the (jutpatient Department where 10 minims tincture digitalis
was ordered to he taken three times a day,

July 30: The cyanosis was marked. The heart action and the pulse were
regular and equal,

August 15: The cyanosis remained tlie same. The heart action and the pulse
were regular and equal.

September 3 : The cyanosis was less, but there was puffiness about the eyes.
The heart was regular at 84 per minute. The urine showed pus cells.

November 4, the patient was seen by Dr. Robinson, who found the outer
cardiac dulness to be 4 cm, to the right and 17.5 cm. to the left of the mid-
sternal line. He also noted an arrhythmia of a peculiar type associated with
periods of regular rhythm. There was also a dropping out of beats without
any evidence of premature contraction.

Electrocardiogram 2027, taken at this time, shows a curious type of arrhyth-
mia suggesting a shift of the i)acemaker in the junctional tissue, delayed auricu-
loventricular conduction and slight left ventricular preponderance.

December 9, Dr. Robinson noted that the patient was somewhat cyanotic.
The point of maximum impulse was 14 cm. to the left in the fifth intercostal
space. The limits of cardiac dulness were 4.5 cm. to the right and 14 cm. to
the left of the midsternal line. A very faint systolic murmur was heard at the
apex. The heart rhythm was irregular. It had a rate of 66 with groupings of
three heats heard at times with regular periods. The liver was not felt and
there was no edema of the ankles.

Jan. 13, 1919, another note by Dr, Robinson states that the ventricular rate
is 74, there is no deficit of the pulse and that the patient complains of pain in
the neck. Electrocardiogram 2163 shows: a tendency to left ventricular pre-
ponderance; delayed A-V conduction; P waves flat in Lead I, diphasic with
deep negative phases in Lead II. negative and notched in Lead III; T waves
upright in Leads I and II, mverted in Lead III; R waves notched in Lead
III. becoming tall with the lowering of the diaphragm; Q R S interval length-
ened ; and periods of striking irregularity marked by changes in the P-R
interval. Changes in the form of P waves are also noted. In two instances
by changes in the form of R waves there is the appearance of a change from
a right to a left ventricular i)rcpondcrance. The P-R interval is 0.24 second
The rate is 78.

342 ARCHirES OF IXTERXAL MEDICIXE

Second Admission.— Uarch 4, 1920.

His complaint at the second admission, in addition to his shortness of
breath, was pain in the lower part of the chest. He had been doing quite well
until six weeks before admission when suddenly he had a "dizzy spell." his
right leg became numb, and he fell to the ground. The attack lasted but a
few minutes, his arm was not involved, there was no loss of consciousness or
aphasia and he was able to walk within ten minutes. His dyspnea had been
gradually getting worse and had been very severe for a week. After working
hard, moving furniture, he had considerable dithculty in breathing. Often at
nights he had to get up out of bed and sit in a chair for a few hours before
he could recline and sleep.

Physical Examination. — This was very similar to that of the previous admis-
sion. The superficial veins, however, were much more prominent. They were
engorged and tortuous over the manubrium below the right nipple and over the
precordium. There were numerous varicosities over the right and left lower
chest. Several large varicose veins were noted over the abdomen about the
umbilicus and in the right hypochondrium. The leg veins were likewise
varicose.

The findings in the lungs, heart and liver were as described previously. The
heart was regular and rapid. The right cardiac border was found to be
from 3.5 to 5.5 cm. to the right of the midsternal line until the seven-foot
roentgenogram showed a large bulge 11 cm. to the right of the midsternal line.
.■\fter this the percussion also routinely revealed a dulness from 10 to 11 cm.
in this direction. According to the fluoroscopic examination made by Dr.
Sherwood Moore and Dr. Frank N. Wilson there was also a clear space
between the heart and the aorta in this region. This they thought ruled out
an aortic aneurysm situated on the descending aorta behind the heart. This
large shadow at the right border of the heart was seen to flicker but the move-
ment was so slight that the observers could not be sure that it was not
imaginary. This wide shadow to the right also gave the impression of a widely
dilated right auricle suggesting the possibility of tricuspid stenosis, but there
were no other signs to bear this out.

Dr. Robinson interpreted the roentgenogram as showing apparently a great
dilatation of the right ventricle. This could not be reconciled, however, clin-
ically with the signs and symptoms present. The latter were also different from
those of aneurism of the heart.

Dr. Dock commented on the conspicuous wavy epigastric pulsation and the
dilated superficial veins and suggested that a tumor could not be ruled out as
other possibilities such as tricuspid stenosis and cardiac and aortic aneurism
had been. An exploratory puncture was not made as had been suggested.

During the period of regularity the rhythm was shown by the electrocardio-
grams to be auricular flutter, with two to one block. Vagus experiments were
done which showed the characteristic marked lability of the His bundle to
vagus stimulation in this condition. Vagus stimulation by direct or ocular
pressure stopped the ventricles for four to five series of four to twelve auricular
beats. There was ventricular standstill for as long as three seconds (Fig. 2).

Dr. Rol)inson suggested a dose of 20 c.c. tincture of digitalis and repeated
vagus tests. The digitalis at first produced irregular blocking with four to
one rhythm. Vagus pressure caused a greater change than before digitalization.
In one run the ventricles stood still for twenty-five auricular beats or about six
seconds. Other series of long runs of 6. 8. 10 and 16 to one rhythm were
obtained. Auricular fibrillation then appeared and persisted. The patient was
discharged clinically improved. March 19. 1920.

Clinical E.vamination and Laboratory Data. — Blood: The red blood cells
numbered 6,470,000; white blood cells, 6,950; hemoglobin, 100 per cent. The
blood Wasserniann was negative. Urine: The specific gravity varied from 1.020
to 1.038. The urinary output was small. N'o albumin or casts were found in
any urine specimen. The plu-nnlsulphriicphlhalein test showed 65 per cent,
excretion in two liours.

HERKMAXX-Bi-RROIIS—TCMOK OF AORTA

343

The blood pressure was from 100 to 120 mm. Hg. systolic, and from 70 to 95
diastolic.

A seven-foot roentgenogram of the heart (.Fig. 1) showed a large sharply
outlined circular shadow projecting to the right of the midclavicular line.

The cardiac dimensions measured on the seven-foot teleroentgenograms were :
Before digitalization. M. R., 11 cm.; M. L., 12.5 cm.; L, 22.5 cm.; A. 7 cm. After
rest in bed and digitalization, M. R.. 10 cm.; M. L.. 12.5 cm.; L. 23 ctn. ; A, 6.5 cm.

Electrocardiograms (Fig. 2) :
3716; March 7. 1920; auricular flutter; auricular rate. 266; ventricular rate, 128.
3725; March 8; auricular flutter; vagus e.\periments (Fig. 2 I.

Fig. 1. — Seven foot roentgenogram of the chest made at the tiiue of the
second admission. March, 1920.

3759; March 9; three hours after digitalization. auricular rate, 256. Mi.xture

2 : 1, 3 : 1 and 4 : 1 block.
3740; March 9; six hours after digitalis; vagus experiments.
3741; March 10; fifteen hours after digitalis.

3750; March 10; auricular fibrillation twenty-eight hours after digitalis.
3769; March 12; auricular fibrillation.
3783; March 16; auricular fibrillation.

Final Xotcs April 3. i920.— After leaving the hospital the patient, contrary to
order, did rather heavy work. He had not been any more dyspneic than usual,
but had some edema about the ankles each evening. He did not take the tincture
of digitalis regularly as it had been prescribed. It was not known whether his
pulse was regular or irregular. He had been complaining for a few days, but

HERRMAXX-BURROIIS— TUMOR OF AORTA 345

refused to return to the hospital. April 2. he carried an iron range (stove) up
three flights of stairs. In the evening he felt more "blue" than usual, but went
to sleep apparently well. During that night he apparently did not struggle, at
least his son sleeping with him was not awakened. On the following morning.
however (April 3). he was found dead in bed. The body was slightly cool,
quite cyanotic, especially the face, but rigor mortis had not set in.

NECROPSY REPORT

Permission for a complete necropsy could not be obtained. .•\n e.xamination of
the heart alone was allowed. This was removed after the body had been
embalmed. We saw it for the first time when it was Itrought to the laboratory.
We are grateful to Dr. Robinson for his effurt in gaining permission for us
to examine this organ.

Heart: The organ is greatly enlarged, and a large tumor mass, the size
of a large orange, is attached to the superior wall of the right auricle, to the
tissue of tile auricular septum, to the sides of the aorta and pulmonary artery.
The attachments to the aorta and pulmonary artery are loose. The firm attach-
ment is to the auricular wall. The fibrous tissue of the pericardium of the
heart is continuous with the tumor. The muscle of the auricular wall beneath
the tumor is atrophic and apparently not continuous with it. The attachment
to the auricle does not appear to be a primary one but secondary to the irritation
of the tumor lying on the auricle. The heart with the tumor weighs 945 grams.

The heart is enlarged, the hypertrophy being more marked in the left than
in the right ventricle. The tricuspid valve measures 14.5 cm.; the mitral, 11
cm.; the aortic, 7.5 cm.; the pulmonary, 10 cm.; the muscular portion of the
left ventricle wall measures 14 mm. : that of the right, 4 mm.

The mitral valve is uniformly thickened throughout and near its base are
several fibrous and fatty placques. a few of which contain gritty material. The
papillary muscles are imicli cnlar.md and there are several areas of fibrous
thickening in the endocardium (if the left ventricle. The aorta is normal in size.
There are a few small fibrous and fatty placques in the intima of the aorta.
The endocardium of the right ventricle and auricle show nothing of interest
except at the attachment of the tumor. Here the endocardium is in close contact
with the hard, gray, opaque wall of the tumor mass, which can be seen through
it. The muscle fibers are largely missing in this region.

The myocardium of other portions of the heart, aside from the hypertrophy,
show nothing of interest.

The epicardium over the whole of the heart contains a considerable amount
of fat. The coronary arteries are straight and aside from the first few centi-
meters of their courses show nothing of interest. In the first centimeter of the
left coronary there are numerous fibrous and fatty placques in the intima. Similar
changes are seen in the first 5 cm. of the right coronary artery.

The orifices of these vessels are slightly constricted and there are several
fibrous and fatty placques in the intima of the aorta immediately about them.

Just to the right of the orifice of the left coronary artery and in the aorta
and just above the sinus of Valsalva is a cone-shaped pouch which extends
inwards and appears to end blindly (Fig. 4a). It is 1 cm. deep and the diameter
of the inner opening is 1 cm. Except for its shape it has the appearance of a
small aneurism. There are several fatty and fil)rous placques in the intima
lining it.

Lifting the tumor away from the aorta reveals the outer wall of this pouch,
which is seen to be continuous with a tough cylindrical shaped cord of tissue
which passes directly to the wall of the tumor to disappear within it (Figs.
3 and 4b).

This slender cylindrical shaped cord is evidently the pedicle of the tumor,
and on close inspection is found to be composed externally of a tough gray tissue
like that of the aorta or one of the large arteries. It has, however, no evident
open lumen, but its center is composed in places of a gelatinous and in other

346

ARCHll-ES OF JXTERXAL MEDICIXE

places of a glistening gray material. It measures 4 cm. long and Vs cm. in
diameter. It looks like an occluded anomalous arterial branch.

The tumor is ovoid in shape and measures 12X9 cm. It has a tough hard
outer wall and is filled with a greasy necrotic material which everywhere
glistens as if it contained crystals. A smear of this necrotic material shows
large numbers of rhomliic plates with notched corners. The outer tough wall
is thickened at the point of insertion of the pedicle. Here there is a large
necrotic fragile nodule which extends in and replaces a part of the soft greasy
material (Fig. 4c ).

Fig. 3. — Scmi(liaj;raniina

ng ot

hca

shinvmg the tv

The remainder of the wall is thin. It measures from 1 to 2 mm. in thickness.
This wall is seen to be divided into several layers. An outer one resembling
muscle in many places, a second fibrous looking layer, and an inner layer which
is in contact with the glistening greasy content. This last layer appears to be
composed of a gelatinous material which in places is streaked and mottled with
gray and yellow opaque lines and splotches.

From the gross description alone it was evident, therefore, that the tumor
was not of auricular but probably of aortic origin. It had arisen from an
anomalous arterial like branch of this vessel.

HERRMAXX-BURROirS— TUMOR OF AORT.I

:-.A7

Microscopic Exaiiiiiiatioii. — The microscopic examination of the heart shows
muscular hypertrophy and nodular thickenings of the endocardium of the left
ventricle. These nodules are fibrous and the centers of the larger ones are
necrotic.

Atheromatous nodules of small size are present in the intima of the aorta.
The media of the vessel appears normal but there are collections of lymphoid
cells about a few of the vessels of the adventitia.

The wall of the aneurysmal-like pouch which marks the point of origin of
the tumor pedicle (Fig. 5) is much like that of the aorta. It has a well formed
intima. media and adventitia. The intima is irregularly thickened and f.brous

Fig. 4. — Similar drawing showing the heart
the aneurysmal pouch are exaggerated. They

(1 lumnr opened. The
e flat and smooth.

in places. The outer portions of the media also show fibrosis and in one place
this layer contains a large calcified placque. The muscle cells and the elastic
tissue of the inner portions stain sharply. At the boundary between the
adventitia and the media there are a few lymphocytes and there are also collec-
tions of these cells about a few of the larger vessels of the adventitia.

The pedicle has the structure of an artery with an excessively thickened
intima which obliterates the whole of the lumen (Fig. 6 and 7). The media is
well formed. It has a sharply defined internal and a poorly defined external
elastic lamella. There is a slight increase in fibrous tissue within it but also

Fig. 5. — Photograph of a cross section of the wall of the aneurysmal pouch.

HEKRMAXX-BURROirS— TUMOR OF AORTA

349

sharply staining muscle cells and elastic fibers. The excessively thickened
intima is composed of loose fibrous tissue with a finely granular intrafibrillar
substance. In places it contains a few polyblastic cells and in other places it
is degenerated and reduced to a homogenous granular mass containing numbers
of open ovoid, circular and rhombic shaped spaces like those of the center of
the tumor.

Sections from all parts of the tumor show the same general structure (Figs.
8 and 9), an outer tough sharply staining thin fibrous wall which grades off grad-
ually or in places sharply into the large granular bluish and pink staining mass.
This mass contains large numbers of open spaces which vary in shape. Many
are spindle shaped (Fig. 10) : others are ovoid, circular or rhomboid in shape.

Fig. 7. — Higher power picture from another section of the pedicle

The outer fibrous wall contains large numbers of poIy1)Iastic, plasma and
small mononuclear cells (Fig. iA). A few of these cells or shadows of them
are seen deeper in the hyaline and necrotic portions (Fig. 8 and 9B). Calcium
deposits arc also not uncommon in the hyaline portions of the wall (Fig. 8 B).

Sections through the auricular wall at the attachment of the tumor show
no direct connection of the tumor with this part of the heart. The tumor wall
is sharply defined and connected with the heart wall by loose fibrous tissue.
In a few places atrophic and degenerating muscle fillers are seen between the
tumor and the endocardium.

350

ARCHIVES OF l.XTERXAL MEDICIXE

The picture as described in this tumor is one of a progressively degenerating
fibrous wall of inflammatory origin. Leaving aside the inflammatory cells it
is the picture of the ordinary atheromatous patch of the aorta. There is no
evidence that it is or has been a neoplasm. It is rather a massive atheromatous
tumor of what appears to be an atypical arterial branch of the aorta. This
branch or pedicle has the appearance of an artery obliterated by an excessive
proliferation of its intima.

Fig. 8. — Photograph of a section
the outer sharply staining area. B

laktn
s the

niKh the wall of the tunio:
■r hyaline and fibrous laye

The pouch at the origin of the tumor pedicle from the aorta has a well
formed wall resembling that of the aorta. It looks like a simple dilatation of
the wall of the aorta at this point rather than an aneurism. The changes in the
wall of this pouch and in the aortic wall are suggestive of a generalized chronic
degenerative and inflammatory disease of the vessels. The nodules as those
described in the left ventricle have been also definitely associated with syphilis
by some. It is possible in this case, however, that these changes in the aorta
mav have been secondarv to the irritation of the tumor.

F'g- S- — I lijiograph of a se^ti.i, ,,i „
tumor. A is the outer sharply staining la}
necrotic layer.

Fig. lU. <Ji,iLi. .,ij.iLC.-, in the .;L-ir.ial part c,t the tumor ina^^

352 ARCHIl-ES OF IXTERXAL MEDICIXE

, CLINICAL SUMMARY

The case, as is evident, is a most unusual one. The cHnical symp-
toms are largely those of a tumor of the heart while pathologically it is
questionable whether the tumor is nothing more than a part of a more
generalized inflammatory disease of this organ.

The patient w-as a chronic alcoholic. The supposed syphilitic infec-
tion at 28 years of age seemed to antedate any of the symptoms. The
"smothering" and "asthmatic" attacks and mild cardiac symptoms
followed this more or less closely. Aside from these facts, as well as
the long duration of the disease, the symptoms might well have arisen
entirely from the presence of a neoplasm. The syn-,ptoms were those of
a gradually progressing myocardial change.

The reason for the insidious exaggeration of symptoms following
"rheumatism" four months before admission is not clear. The dyspnea
and the wheezing on reclining, relieved by arising, might well be
explained as myocardial resulting from tumor pressure.

In the original physical examination, the observation of an ashen
pallor, conspicuous cyanosis of the body and dilatation of the super-
ficial abdominal veins on standing and the spread of the cyanosis to
the face on reclining might have resulted from a tumor pressing on
the venae cavae. Cheyne-Stokes breathing was present as a part of
the cardiac syndrome. The dulness to the right and below the third
rib which was taken to be that of the liver was probably due to the
tumor. The roentgenogram of the chest taken at this time showed a
shadow in this region which appears to be clearly the same enlargement
that produced the striking teleroentgenogram which led to the special
interest in the case and the provisional diagnosis of cardiac tumor.

On the second admission there was the added complaint of pain in
the lower chest which might have been associated with the tumor.
The sharply outlined shadow extending 11 en., to the right of the
midsternal line was considered one of three things : an aneurysm of the
heart, a dilated right auricle from tricuspid stenosis or a tumor. The
latter was not ruled out, while aneurysm and dilated auricle were
considered im]3robable because of the patient's general condition and
the physical findings. Fluoroscopic examination showed no definite
movement or pulsation of this huge shadow.

The electrocardiograms in this case revealed inleresting tindings.
The curves first showed auricular flutter which after digitalization
changed to auricular fibrillation and later to normal mechanism with
inverted P waves and delayed A-\' conduction. On the second admis-
sion the auricular flutter was again jiresent and vagus experiments
were very successful in demonstrating a regular auricular activity with
auriculn-vcntricular blocking (Fig. 2). .Xftcr a massive dose of digitalis
auricular fihrillatinn lifgan and iicrsivted. The-'C .'ibnornial auricular

HERKMAXX-BURROIVS— TIMOR OF AORTA 353

rhytlinis and mechanisms are associated with auricular muscle changes
which in this case resulted from the pressure of the tumor, encroaching
on the sinus area. The inverted P waves signif)' an abnormal course of
the impulse through the auricles. The electrocardiographic findings
point, therefore, definitely to pathology in the auricular muscle which, as
the necropsy showed, was associated directly with a tumor pressing on
the auricle and atrophy of the underlying auricular muscle.

In this case it was evident, therefore, as Dr. Dock pointed out, that
there was but one pos.sible diagnosis to make and that was tumor.
The lack of expansile pulsation or any visible movement at all in the
mass and the absence of a tracheal tug, reduplication of the heart
sounds and pain ruled out in large measure an aneurism of the aorta.
Again, there were no definite symptoms of sufficient gravity to indicate
an aneurism of the right ventricle and the absence of symptoms and
signs of tricuspid stenosis made a diagnosis of dilatation of the right
auricle unjustifiable.

DISCUSSION

Tumors of the heart are not common but do occur in small numbers
among the cases of any large clinic. The tumors which may be classed
together into this group, as is well known from the literature may arise
(T) from the heart itself, (2) from the pericardium, or from the
adjacent portions of the great vessels or neighboring tissues of the
mediastinum.

Link,' in 1909, collected ninety-one cases from the literature which
he concluded were primarily of cardiac origin. Meroz,- 1912, selected
fifty-five cases which he thought had their origin in the heart. Since
that time Norton,'' Gottel ^ and Weltmann '" have each reported a case
of a primary heart tumor.

According to the classification of Meroz the primary tumors of
the heart may be (1) valvular; (2) intramuscular, and (3) intra-
cavitary.

The tumor we observed belongs to none of these latter groups
but must fall in its clinical aspects among those of the pericardium.
It is questionable whether it was attached to the auricle in its earlier
period of development. This attachment with the definite deterioration

1. Link. R.: Klinik der primarcr XeulMldungcm dcs Htrzens. Ztsclir. f. kliii.
Med. 57:272. 1909.

2. Meroz. E. : A Clinical Study of Three Cases of Priniarv Tumor of the
Heart. Internal. Clin. 4:231, 1917.

3. Norton. W. H.: Myoma of the Heart. .■Xm. J. M. .Sc. 158:689, 1919.

4. Gottel. L. : Ein Fall von primaren Hcrztumor, Dcutsch. med. Wchnschr.
45:937, 1919.

5. Wcltmann. O. : Klinschr Bcitrag zur Kasuistik i)rimarcr Herztumoren,
Wien. klin. Wchnschr. 33:537. 1920.

354 ARCHIVES OF IXTERXAL MEDICISE

of the auricular muscle may have occurred in the later stages of the
disease. Our tumor was definitely of aortic origin with secondary
attachment to the heart.

The tumors decribed arising from the pericardium are ( 1 ) fibroma ;
(2) primary sarcoma; (3) myxoma; (4) parasitic cysts, and (5)
dermoids. Fornia " recently made a collection of sixteen such cases.
Secondary tumor metastases are not uncommon. These may arise
from carcinoma, sarcoma, Hodgkins disease or lymphosarcoma situated
in adjacent or distant regions of the body. Inflammatory nodules are
also seen. Recently one of us [Burrows] observed such a case at
necropsy. The nodules in this latter case were multiple. The largest
measured about 1cm. in its longest diameter. The whole of the
visceral pericardium was thickened. The nodules were striking
histologically on account of the large number of foreign body giant
cells which they contained. There were also similar inflammatory
nodules in the lungs, pancreas, suprarenals and kidneys. The etiology
was not determined. Diabetes mellitus was the cause of death.

These various cardiac tumors, as is well known, according to their
position may simulate in their clinical picture any one of a great
variety of grave cardiac diseases." For this reason up to the time of
the development of the roentgen ray their diagnosis was practically
impossible, as has been stated by Oppolzer,' Gottel and others.

At the present time, as a few recent cases and our case illustrate,
this may no longer be true for those tumors presenting on the surface
of the heart. The cardiac cachexia not otherwise explained should,
therefore, be most carefully studied by these means not only for the
importance of improving these diagnostic methods but also from
the standpoint of treatment.

The thorax is no longer a barrier to the surgeon. Tuffier '' recently
reported the successful partial removal of a dermoid the size of two
fists which was adherent to the aorta and encroached onto the wall of
the auricle and ventricle. The clinical symptoms were those of a grave
angina pectoris. The case was diagnosed a cyst by Vaques. The
tumor was partly calcified and was filled with greasy material. The
incision, drainage and disinfection took six months. The patient recov-
ered completely and is now free from the symptoms of angina pectoris
and is in good health.

6. Forni, G. : Primary Tumors of the Pericardium, Tumori 4:523, 1914.

7. Oppolzer : Quoted by L. Krehl ; Diseases of the Myocardium. Nothnagel's
Encyclopedia of Practical Medicine. Diseases of the Heart, translated and
revised by Dr. George Dock.

8. Tuffier: Tumors Primitives du Coeur. T.a Chirurgic du Cncur. (Piece
presentee a la Sooietc dc Chirurgie. Paris) 1920.

HERRMAXX-BURROllS—TCMOk OF AORTA 355

Our case was brought to the attention of Ur. Evarts A. Graham who
"considered it an operative possibility had it been seen before the grave
cardiac arrhythmias, irregularities and signs of heart failure (the
so-called decompensation) had developed.

I'ATIIOLOGV

Pathologically, as stated above, we have been unable to find any
counterpart of this tumor in the literature of heart and aortic tumors.
Tumors of the aorta are not common. The catalogue of the Surgeon
General's Library reports but fifteen papers on this subject. Two of
these were written in the eighteenth century before the time of careful
histologic study of tissues. Three were cases ot secondary carcinoma-
tous metastases. One was a parasitic cyst in a dog, and one other was
a blood cyst. Three were primary sarcomas, and one, reported by
Threadgill, was a cartilaginous mass in the aorta of a negro child. The
paper by Okada was not obtained, and another by Maixner adds nothing
new to this subject.

Joel," in 1890, reported the occurrence of a teratoma of the pulmon-
ary artery. As stated above, our tumor is not a neoplasm but it is
evidently of inflammatory origin. The pedicle has the structure of an
arterial branch of the aorta which has undergone dilatation of its
orifice and secondary atheromatous obliteration. The tumor has all
the characters of an atheroma of a massive size.

We at first thought that similar tumors might have been observed
and confused w-ith dermoids or fibromata of the heart or pericardial
tumors. A careful study of the gross and microscopic descriptions
of such cases described in the literature has failed, however, to reveal
any such similarities. ("Morris,'". Lambert and Knox," Christian '-
and Dangschaat.'^)

Again, there is no evidence that it was primarily an aneurysm. It
is more that of a granuloma undergoing secondary changes.

The etiology is obscure. It is impossible to be certain that it is
related to other evidences of inflammatory disease of the heart and
the aorta in this case and the supposed syphilitic infection. Le\aditi

9. Joel. J.: Ein Teratom auf dcr arteria pulmonis innerliall)s de.s Herz-
beutels. Virchows Arch. f. path. anat. 122:381. 1890.

10. Morris, R. S. : Dermoid Cysts of the Mediastinum. Med. News 87:404,
438. 494 and 538. 1905.

11. Lambert, S. W.. and Knox. L. C. : Intrathoracic Teratoma. Tr. Assn. .'\m.
Phys. 35:17, 1920.

12. Christian. H. A.: Dermoid Cysts and Teratomata of the Anterior Medi-
astinum, J. M. Research, 2:54, 1902.

13. Dangschaat. B. : Beitrage zur Genese, Pathologie und Diagnose der
Dermoid Cysten und Teratom im Mediastinum .Anticum. Beitr. z. klin. Chir.
38:692. 1903.

356 ARCHIVES OF JXTERXAL MEDICIXE

stains were made of the tumor, of the nodules in the endocardium ; of
the left ventricle and of the aorta. The poor fixation of the tissue,
however, made the negative results of no importance.

Its rarity cannot be accounted for. In this location it might be
associated with the rarity of an artery arising from this part of the
aorta. This does not explain, however, its absence of occurrence
in other regions of the body.

It has been of interest, therefore, for us to report this case as a
peculiarly rare clinical entity, the existence of the tumor of which is
possible of diagnosis and which must fall, therefore, within the domain
of the surgical treatment of the present and the future.

BACILLUS ACIDOPHILUS AND ITS THERAPEUTIC
APPLICATION

LEO F. RETTGER. Ph.D.. and HARRY A. CHEPLIN, Ph.D.

NEW H.'WEN, CONN.

Bacillus acidophilus was fir.st observed and described in 1900 by
More' It is a rather large gram-positive bacterium which is quite
pleomorphic and which in many respects resembles Massol's Bacillus
bulgaricus and Tissier's Bacillus bifidus. It was claimed by Moro to be
the chief inhabitant of the intestine of infants that subsist entirely on
mother's milk. This assertion was disputed by Tissier ' who protested
that B. bifidus holds the place of prime importance. B. acidophilus and
B. bifidus are now known to constitute the main flora of the breast-fed
child, the latter being, perhaps, the more prominent of the two. As the
diet changes and becomes more and more complex, there is a corre-
sponding change in the kinds and relative numbers of intestinal bac-
teria, until finally the intestinal population assumes the character of
that of the ordinary adult. B. acidophilus and B. bifidus gradually
disappear to such an extent that their presence can be demonstrated
in the feces with considerable difficulty only. Their place has been
taken by various other organisms, some of which are decidedly fer-
mentative and putrefactive, and, according to Metchnikoff and others,
as.sume a role harmful to the host.

B. acidophilus is practically indistinguishable from B. bulgaricus
and has undoubtedly often been mistaken for the latter. There are,
however, two well-known points of distinction between these two
organisms. B. acidophilus produces relatively little acid in milk (less
than 1 per cent.) even after continued incubation, whereas B. bulgaricus
may produce as much as 3 per cent. Furthermore, B. acidophilus
attacks maltose with acid formation, while B. bulgaricus has no action
on this sugar. The most important mark of diflFerence is that relating
to intestinal implantation. B. bulgaricus, as numerous experiments have
shown, is unable to live and multiply in the intestine of the white rat
and of man, whereas B. acidophilus undergoes rapid development when
administered by mouth, or as the result of milk, lactose or dextrin
feeding.

Rettger and Horton ^ and Hull and Rettger •" had shown th.-it the
feeding of milk or lactose to experiment animals, including the white

1. Moro, E.: Ueber den Bacillus acidophilus, Jahrb. f. Kinderh. 52:38, 1900.

2. Tissier, H. : Recherches sur la flore intcstinale normale et pathologique
du nourisson, These. Paris, 1900.

3. Rettger, L. P., and Horton, G. D. : A Comparative Study of the Intes-
tinal Flora of White Rats Kept on Experimental and on Ordinary Mixed Diets,
Centralbl. f. Bakteriol. 73:362, 1914.

4. Hull, T. H.. and Rettger, L. F. : The Influence of Milk and Carbohydrate
Feeding on the Character of the Intestinal Flora. J. Bacterid. 2:47, 1917.

358 ARCHIIES OF IXTERXAL MEDICI XE

rat, tended to decrease the number of ordinary intestinal bacteria and
to establish an organism of the B. acidophilus type. Hull and Rettger *
demonstrated that an almost complete acidophilus flora may' be obtained
in the albino rat by the daily feeding of 2 gm. lactose. Cheplin and
Rettger ' confirmed these observations on white rats, and showed
further that similar results may be had in man. They found that the
daily ingestion of from 300 to 400 gm. lactose brought about a very
profound change of bacterial types in which B. acidopliihis greatly pre-
dominated, often to the extent of at least 90 per cent, of the viable
organisms present. Torrey " had shown that m the treatment of
typhoid fever patients with lactose from 250 to 300 gm. were necessary
daily in order to establish an acidophilus flora.

It was not until the spring of 1919 that any serious attempts were
made to administer cultures of B. acidophilus for the purpose of bring-
ing about an implantation and proliferation of this organism in the
intestine. Cheplin and Rettger "^ demonstrated that transformation of
the flora may be brought about easily and within the course of four to
six days in apparently normal human subjects by the daily administra-
tion of 300 CO. pure whey-broth cultures of the organism. B. acido-
philus was often present in the feces to the extent of from 85 to 90
per cent, of the cultivable bacteria. Similar results were obtained with
all of these subjects when 150 c.c. of the culture and 150 gm. milk
sugar were given in place of the requisite amount of culture alone.

Implantation of B. acidophilus may be brought about, therefore, by
any one of the three methods just described. In fact, the results, as
measured by the bacterial picture, were practically the same in each
method, as is clearly shown in the curves and tables published at some
length in book form." The whey-broth culture was soon displaced,
however, by the acidophilus milk culture which was devised by us
early in 1920. On account of the extreme importance attached to the
use of Bacillus acidophilus milk the following brief description is given
of this product and of the method of preparalioi..

i;.\CILLUS ACIDOPHILUS MII.K

Fresh skimmed cow's milk is sterilized in one heating at 115-120 C,
the time required being determined by the volume of the milk and the
nature of the container. Quart lots in ordinary glass flasks are heated

5. Cheplin, H. A., and Rettger, L. F. : Implantation of Pacillus acidol)ltilus.
Proc. Xat. .\cad. Sc. 6:423, 704, 1920; Proc. Soc. Expcr. Biol. & Med. 17:192.
1920; 18:30. 1921.

6. Torrv, J. C. : The Fecal Flora of Tvphoid and Its Reaction to Various
Diets, J. Infect. Dis. 16:72, 1915.

7. Rettger, L. F., and Cheplin, H. A.: .A Treatise on the Transformation
of the Intestinal Flora with Special Reference to the Implantation of Bacillus
acidophilus, Yale University Press, 1921.

RETTGER'CHEI'LIS—HACILLIS ACIDOPHILUS 359

22-24 minutes. Properly heated milk should have a dark cream color,
but should not be distinctly browned. After cooling to at least 37 C.
the milk is inoculated with pure strains of B. acidophilus which have
been grown, by repeated transfers, sufficiently long in milk to develop
rapidly and bring about coagulation of the casein within twenty-four
hours at a temperature of from ih to ?>7 C. \'iable milk cultures of
the organism are employed as the inoculum, and at least 10 c.c. of the
inoculum are transferred for each liter of milk treated. After thorough
mixing, the newly inoculated milk is incubated for the period stated.

At the completion of the full incubation period the casein appears
as a soft curd, with a thin layer of clear or almost clear whey over
the surface. On thorough shaking the curd falls to pieces and the
milk acquires a smooth consistency which resembles that of thick
cream. Even during long standing there is no marked separation of
whey from the curd, or collection of the curd into granules or lumps.
The acidity of the acidophilus milk is never high, as compared with
that of B. bulgari-ciis milk, seldom reaching more than 1 per cent.,
even after a week's incubation at ordinary room temperature. The
odor and flavor resemble to some extent those of a high grade butter-
milk and should be agreeable to persons who do not have a dislike for
sour-milk products as such. The odor and taste are, however, decidedly
characteristic, when the milk culture is pure. Contaminating organisms,
if they have developed appreciably in the milk, owing to imperfect
sterilization or subsequent contamination, change the character of the
milk in such a manner as to be recognized readily by the odor and
taste, and often by the physical appearance uf the product. For
example, the ordinary milk souring bacteria cause the usual sour milk
fermentation and coagulation, and the common spore forming organ-
isms of the B. siihtilis type produce offensive products characteristic of
the group. These are the main types of contaminants thus far encoun-
tered in the acidophilus milk by us.

TIIKKAPKLTIC ATPLICATION- OF B.\Cn.LL-S .XCIDOPHILUS

In the first series of experiments with man, seventeen different
subjects were employed. Fifteen were apparently normal and two had
a long history of intestinal disturbances. Most of the subjects served
in more than one experiment, bringing the total number of individuals
up to forty-three. The entire work involved the bacteriologic exam-
ination of 580 stools. Implantation of B. acidophilus, with the predom-
inance of this organism to the extent of at least 80 per cent, of the total
flora of the feces, was brought about almost at will. Particular atten-
tion is called to the fact, however, that no two subjects required the
same minimum amount of lactose, B. acidophilus culture or combina-

360 ARCHJJ-ES OF IXTERXAL MEDICIXE

tion of the two, to yield to the transforming influence within a given
period of time.

In all of the experiments the change in the intestinal flora was
determined by the three following bacteriologic methods : ( 1 ) Plating
of fecal suspensions (in physiologic sodium chlorid solution) in whey
agar and the numerical study of B. acidophilus-\\kt colonies; (2)
preparation of deep whey-agar (Veillon) tubes for the determination
of gas producing organisms of both the aerobic and anaerobic types ;
and (3) staining of slides by the Gram method and a differential
study of bacterial types. These methods proved to be most valuable,
and were employed also in the more recent work on intestinal flora
in abnormal subjects. For a full description of these and all other
methods employed in the earlier investigation, and for a full account
of the results, the reader is referred to our monograph ' on intestinal
flora.

No difificulty was experienced in obtaining the cooperation of
practicing physicians in our investigation of the therai)eutic properties
of B. acidophilus when administered by mouth, either alone as a pure
milk culture or in combination with different amounts of lactose.
It became apparent at the outset that the acidophilus milk is much to
be preferred to the lactose broth or whey broth cultures of the organism
for the following reasons : The milk is tolerated even by those who
are unable to retain the simplest and most wholesome foods for con-
valescents; when properly prepared and preserved the acidophilus
milk remains practically unchanged, and free from bacterial con-
tamination and deterioration : it contains at least 4 per cent,
lactose which in itself serves to stimulate B. acidophilus proliferation
in the intestine ; it is nutritious, and for those who cannot take or do
not tolerate other foods it does much toward the maintenance of
nitrogen balance and the prevention of tissue waste, when taken in
the usual amounts, from 1 pint to 1 quart daily and finally, as a young
culture of viable bacteria it is particularly potent in bringing about
the desired transformation of bacterial types in the intestine.

PATHOLOGIC CASES

It is not our purpose here to present a detailed history of the
individual cases, nor to dwell at any length on the treatment and
the bacteriologic results. A full account of these studies will be
published, however, at an early date. The subjects under observation
may be .subdivided into the following groups ; ( 1 ) Chronic constipation
with the symptoms of so-called autointoxication and other accompanying
pathologic conditions, some of them acute, 20 cases; (2) chronic
diarrhea following an attack of bacillary dysentery. 2 cases ; (3) colitis.

RETTGER-CHEPLIX—BACILLUS ACIDOPHILUS 361

at times bloody, and more or less mucous, 3 cases; (4) sprue. 2 cases;
(5) dermatitis (eczema), 3 cases.

These thirty cases are exclusive of those which have come under
our observation within the past few weeks. Nor do they include
those which are being studied and treated in other institutions through
our cooperation. Particular attention is directed to only a few of
these thirty cases.

The subjects were requested to bring to the laboratory one or two
specimens of stool before taking the acidophilus treatment, and for
a while daily samples, when procurable, after the first administration
of the acidophilus milk or of the milk plus stated amounts of lactose.
Bacteriologic examinations were made of these specimens, and their
results correlated with the clinical findings. It was the aim to obtain
a pronounced transformation of the bacterial flora of the intestine in
the shortest period of time ; hence, so-called "maximum" treatment
was given from the start. As lactose has a marked laxative effect
when taken internally in sufficient amount, persons having a history
of obstinate chronic constipation at first usually received 1 quart of
the acidophilus milk plus 100 gm. lactose daily. The powdered lactose
was added to the acidophilus milk in the flask and the contents
thoroughly shaken. Subjects were instructed to take the daily supply
in three equal portions, one in the forenoon, another in the afternoon,
and the third immediately before- retiring for the night, and in every
instance at least two hours before and after meals. There were no
regulations as to diet, except that the subjects were urged to refrain
from the use of food which by experience or training they knew to
prove injurious.

If, in the course of three or four days constipation was not relieved
except by an enema, which was advised when the condition of the
subject made it absolutely necessary, the daily amount of lactose was
increased by 25 or 50 gm. On the other hand, when stimulation of
peristalsis became too marked and a diarrheal condition resulted from
the taking of the full amount (1 liter of acidophilus milk and 100 gm.
lactose), the quantity of lactose was reduced by 25 or 50 gm. In a
few instances the volume of milk taken daily was reduced to 500 c.c,
with or without a reductiop in the amount of lactose.

In the diarrheal cases (including colitis and sprue) the treatment
consisted in the daily administration of 1000 c.c. acidophilus milk
without any added milk sugar. The milk was easily tolerated by the
patients; in one instance, however, the volume was reduced after the
first few days to 500 c.c. owing to a feeling of fulness and partial
loss of appetite of which the subject complained. Per.sons who could

362 ARCHIVES OF IXTERXAL MEDICINE

not take other food in any form retained the acidophilus milk and
complained of no distressing or otherwise injurous eiTfect.

Chronic Constipation. — The two following cases of chronic con-
stipation are briefly reviewed here, as they happened to be the first to
take the treatment, and were among the most obstinate advanced
cases of alimentary toxemia tliat ha\e as yet come under our
observation.

Case 1. — Subject D. had suiYered from cunstipation lor at least seven or
eight years, and complained of fulness of the abdomen and of gas and pain
in the epigastric region, also of an almost constant headache, visual distur-
liance, general discomfort after meals, loss of energy and initiative, general
malaise, and what was most disturbing to him. melancholia and other indi-
cations of lessened mentality. He regularly required unusually large doses of
la.xative to induce bowel evacuation, and stated that he had been absolutely
dependent on cathartics for at least two years for relief which was even then
only partial and of very short duration.

lilk to diet,

The subject was given 1 liter acidophilus milk daily, and instructed, as
were all others, not to employ a cathartic, and to bring daily specimens (when
possible) of stool. He reported almost daily at the laboratory for observation.
Unfortunately, no sample of stool could be obtained before the beginning of
the treatment. However, in none of the numerous experiments already con-
ducted on man was B. acidophilus observed in the feces before treatment with
B. acidophilus culture or special carbohydrates without much difficulty and then
only rarely and in very small numbers; hence, it is safe to assume that in
cases of constipation at least the per cent, of cultivable B. acidophilus cells is,
without special stimulation, at or near 0.

Within forty-eight hours after the first ingestion of the milk culture the
bacteriologic examination of the feces revealed a marked transformation of
the flora with a preponderance of B. acidophilus-]i\ce organisms. Bv the
end of the fourth day the gas producing organisms had apparently completely
disappeared from the intestine, and the aciduric type reached a percentage level
of at least 90 which it maintained throughout the course of the treatment
(Fig. 1). The administration of the acidophilus milk had to be discontinued
and the experiment ended after a period of thirty days, owing to the seasonal
closing of the laboratory. The subject left very soon after for continued
residence in the Orient.

RETTGER-CHEPLIX— BACILLUS ACIDOPHILUS 363

On the third day after the subject began the treatment he reported a
normal bowel evacuation, and continued to report at least one daily movement.
barring one exception, for the remainder of the experimental period. There
was an apparent change within the first three days in the patient's general con-
dition. By the end of the first week he reported complete relief from gaseous
distension and pain in the epigastic region. His appetite was good, and he was
free from the usual intestinal disturbance following a substantial meal. Little by
little he indulged his appetite more and ate heartily of meats, biscuits and pastries
with apparent impunity, which hitherto he was forced to avoid. His head-
aches and mental disturbances disappeared, and he resumed his full daily
duties with, as he stated repeatedly, new strength and endurance. These claims
were fully borne out by his appearance and actions.

C.\SE 2. — Subject W. was an almost exact duplicate of D. He had. however,
marked dilatation of the descending colon. He received essentially the same
treatment as Subject D., and though he responded more slowly at the begin-
ning, the bacterial transformation was almost complete by the end of the first
week, and the percentage of B. acidophilus remained at 85 to 95 throughout the
course of observation. Gas-forming organisms disappeared early. On the third
day after the first administration of the acidophilus milk he presented a natural
specimen of stool which appeared normal, and thereafter, with very few
e.xceptions, reported daily normal evacuations. His physical and mental con-
dition had been such as to cause him considerable alarm. He had over a
period of five or six years consulted several specialists and had been advised
by two of them to undergo an operation for the colon dilatation.

This subject's general condition began to improve almost immediately. He
reported constant improvement from day to day. which was apparent also in
his facial expression and in his actions. He was under observation for eight
months. After about three months of the treatment he was advised to take
the milk less frequently and to continue its gradual elimination from the diet
for the purpose of determining whether it could not be dispensed with eventu-
ally. Until early in May. and for a period of four weeks, he took only 1 or
2 quarts a week. During the succeeding six weeks he called twice for the
acidophilus milk, stating each time that, while he still continued to be in
very good condition, he thought it desirable to take a pint or a quart because
of slight premonitions of returning trouble. He has not reported in person
since, and in so far as the writers know there has been no recurrence.

Contrary to the rule, both these subjects responded readily to the
use of the acidophilus milk, without any added lactose. Two other
subjects have reacted similarly. On the other hand, most of the
chronic constipation cases required 100 gm. added milk sugar, and
one of them required as much as 200 gm. daily along with the quart of
acidophilus milk. In no instance was there a failure to bring about
relief when the treatment was not interrupted early.

Chronic Diarrhea. — Case 3. — Subject B. A Bohemian, male, 42 years old,
residing in Xew Havea since April, 1921. He was seized in September, 1907.
with an attack of acute diarrhea accompanied with tenesmus. The stools were
bloody and contained mucus. Temporary recovery : no bacterinlogic examina-
tion; recurrence in the spring of 1908; in hospital at Frankfurt a. M. for six
weeks. Bacteriologic examination negative. In sanitarium at Meran in Tyrol.
Limited diet ; temporary recovery. Recurrence in summer of 1908. Diarrheal
condition more or less constant till 1914, when B. dyscnteriac was identified in
the stools. Temporary recovery. Recurrence in January, 1915, while living
in Silesia. In dysentery hospital six months. Shiga-Kruse bacillus again
isolated from feces. Serum treatment. Left hospital as incurable. In 1915 in
hospital in Hamburg ; for six months exclusive diet of oatmeal gruel. Treat-

364 ARCHIVES OF IXTERXAL MEDICIXE

ment resulted in temporary cure. Normal formed stools for the first time
since 1909. Recurrence after four weeks : acute abdominal pain and diarrhea.
Feces examination negative. Condition much improved after four months'
residence in Norway. Partial recurrence in England (December, 1920, to April,
1921). Took ship for America April 2. 1921. Apparently normal on ship and
for a few days after arrival in New Haven.

Within a week after subject B. reached New Haven he was suddenly seized
with an attack of profuse diarrhea and acute pains in the descending colon.
Presented himself for treatment in April. Stools were watery, very dark, and
of pronounced putrefactive odor. By the third day the stools appeared lighter
(yellowish) in color and less watery. After one week's treatment the abdominal
pains had practically disappeared, the feces rapidly approached the normal
and the diarrheal condition had almost completely ceased. During the second
week's treatment he pronounced himself fully recovered. He continued appar-
ently normal for about three weeks when owing to dietary indiscretion there
was a slight return of the disturbance. The night before the partial recur-
rence he attended a banquet and indulged freely in all that was set before
him, including two courses of meat. This setback lasted only two or three days.

Subject B took one quart of acidophilus milk daily from the first without
any intermission. Except for the above mentioned and a second slight recur-
rence early in June brought on apparently by bathing in cold sea water, he
was, according to all appearances and his own claims, in normal condition for
two months and to the time when the experiment was interrupted owing to
the closing of the laboratory (July 1). The subject has on several occasions
since then informed the writers that he has suffered no recurrence.

Another subject whose case in some respects was similar to this
reacted favorably to the treatment. A full history of this case, with
results, will be given in a later paper.

Co/jVw.— The two cases of colitis which came under our observation
were of the acute type. One was apparently uncomplicated, but had
a long history of intestinal disturbance. The other was complicated
with nephritis. The former responded completely to the milk treat-
ment (8(X) c. c. daily) and after about one month returned to his
work. He continued in apparently good health for about two months
when he experienced a partial reversion, due to overconfidence in
his improved condition and indulgence in late shore dinners.

The second patient gave every indication of physical improvement
when, owing to his serious nephritic condition, he required special
hospital treatment and was compelled to discontinue the use of the
acidophilus milk. His severe abdominal pains had almost completely
disappeared, however, and he was suffering apparently very little
from the colitis.

Both of these cases will be presented more fiilly at a later date.

Sprue. — Case 4. — Subject Wr had contracted sprue while in China. Though
the a'ffection was not of an acute type, it was serious and compelled him to
return to the United States where for at least two years he was constantly
under its annoying and debilitating influence. He took one quart of the
acidophilus milk daily for six weeks during which time the character of the
feces changed completely from the clay-colored, soft and extremely offensive
type to the yellow, almost formed and almost odorless. The gas disappeared
from the colon, and the subject stated from day to day that he thought he

RETTGER-CHEPLIX— BACILLUS ACIDOPHILUS 365

was for the time at least in perfectly normal condition. The bacteriological
results of fecal examination are shown in Figure 2.

A second case of sprue was that of a returned missionary who
had contracted the disease during twenty years of service in China.
\\'hen seen by the senior author he had been confined to his bed for
several months with the characteristic symptoms in their most acute
form, including tetany. He was emaciated and subject to abdominal
pains and gaseous distension.

This patient has been under observation for almost six months,
during which period he has taken, with very few brief intermissions,
one quart of acidophilus milk daily (the milk being sent to him by
special messenger). He has shown from almost the beginning gradual
improvement in his condition, though he has from time to time suffered
relapses.

Fig. 2. — Results of addition of B. acidophilus milk to ordinary diet and then
returning to ordinary diet.

Der>natitis. — All of the three cases of dermatitis were those of
eczema. Treatment of two of them was discontinued before any
definite results could be obtained. The third patient (Subject Q)
responded completely to the treatment, and for five months has been
free from the eczema which for at least twelve years had been a
source of constant annoyance and embarrassment. When first seen
practically the entire face was involved, as well as other parts of
the body.

Transformation of the intestinal flora was effected with consid-
erable difficulty, and required fully a month. However, by increasing
the daily amount of added lactose from 50 to 100 gm. a high aciduric
flora was established and maintained throughout the remainder of the
e.\periment. Concomitant with this change there began a clearing of the
skin which continued until after another month almost all traces

366 ARCHIVES OF IXTERSAL MEDICI \E

of the dermatitis had vanished. At the present time there are no
indications of a return of the affection.

GENER.A.L COMMENTS ON THE THER.APEUTIC APPLICATION OF
BACILLUS ACIDOPHILUS

The work of the past two years has shown conclusively that B.
acidophilus can be implanted in the intestine of man almost at will,
and that its colonization there may be such as to displace at least 80
per cent, of the usual mixed flora. This transformation is brought
about by the administration of at least minimum amounts of lactose
or of living acidophilus culture or of a combination of both. The most
practical and effective method is by means of the acidophilus milk or
of the milk plus definite amounts of lactose given daily. Cases of
obstinate constipation almost invariably require some added lactose.

B. acidophilus is, according to all available information, an organism
which does not elaborate toxic or other harmful products. It is classed
as a strictly nonfermentative and nonputref active micro-organism. It
is present in the intestine of nursing infants in relatively large numbers,
and at all times is an inhabitant of the inte.stine of children of all ages
and of the adult. It is demonstrated with difficulty, however, in the
feces of persons subsisting on the usual mixed diet of which protein
forms a large part, owing to its being suppressed by the other intestinal
micro-organisms which flourish on the usual mixed diet. Administra-
tion of sufficient lactose (or dextrin) stimulates the proliferation of the
relatively few aciduric bacilli. Ordinary milk accomplishes the same
result, but enormous amounts of the milk are usually required to
bring about the change. Pure viable cultures of S. acidophilus accom-
plish the .same thing, with or without added lactose, but when applied
alone definite minimum quantities of the cultures or suspensions, which
must be determined for each individual, are necessary.

Mable cultures of B. acidophilus when used in sufficient amounts
and under the right conditions have important therapeutic properties.
They are of particular merit in the treatment of chronic constipation
and of diarrheas, as the numerous experiments thus far conducted by
the writers conclusively show. Furthermore, B. acidophilus .should be
of marked benefit in the treatment of other ailments which are directly
or indirectly referable to disturbed function of the digestive system and
uf the chief organ of elimination; the results already obtained on
several individuals bear out this promise.

A word of warning should be sounded at this time. The principle
(in which the acidophilus treatment is based has been clearly set forth
in the various publications from this laboratory. It is only when
these principles are adhered to that favorable results are to be expected.

RETTGER-CHEPUX—BACILLUS ACIDOPHILUS 367

In the first place, this is not a cure for all kinds of ailment, nor will all
cases of disturbances falling within its category necessarily respond
to the acidophilus treatment. In the second place, it is not an elixir
in the sense that Metchnikoff's B. hulgaricus was for a while supposed
to be. The ingestion of relatively few acidophilus bacilli will not lead
to implantation and bodily improvement. A minimum amount of
bacterial culture is necessary to bring about these results, and for this
reason the writers have adopted the acidophilus milk culture as the
surest and most eflfective means of attaining the desired results.
So-called acidophilus tablets and powders can be of no use whatever
in effecting transformation of flora and relief from trouble, either
wlien taken as such or when used for the production of acidophilus
milk. As has been fully pointed out elsewhere, the preparation of
viable and satisfactory acidophilus cultures is impossible without the
absolute sterilization of the medium to be used, and without the aseptic
use of pure cultures of the organism as inoculum.

The fullest benefit from the acidophilus treatment can be derived
only "when the patient is under practically daily observation and when
thorough bacteriologic examinations of the feces are made at frequent
intervals. These examinations are of extreme importance in determin-
ing the amount of acidophilus milk or milk and lactose that are to be
taken from day to day. Unless at least a reasonable degree of trans-
formation of the intestinal flora to the aciduric type is effected little
should be expected from the clinical standpoint. In any application of
the principle of acidophilus treatment immediate results must not
always be expected. In a number of instances in which the treatment
became effective we have failed to obtain a favorable reaction until
after one to two weeks' application, and in certain few cases there was
no favorable response until a month after the initial administration
of the acidophilus milk or the milk and lactose. This is to be expected
when the results to be sought must of necessity be indirect, as in the
positive eczema case. Yet, such cases may clear up and in the end
prove to be some of the most satisfactory.

Various plans are now being considered for making B. acidophilus
milk generally available and under such conditions as to assure its
viability and purity.

THE EFFLUX OF BLOOD FROM THE CAROTID

ARTERY OF THE DOG AND ITS EXPRESSION

BY A GENERAL EMPIRICAL FORMULA*

H ALBERT L. DUNN, B.A.. M.A.

MINNEAPOLIS

The study of the normal and pathologic significance of the blood
pressure has been a subject of interest for many years. It was not
until mechanical instruments were available, however, that any accurate
measurements of blood pressure were taken. Ludwig. in 1847, first
obtained a blood pressure tracing. In 1882. almost forty years later,
\'. Basch invented the tonometer. Roy and Adami in 1890 modified
this for external use and Riva Rocca in 1896 manufactured the first
practical clinical instrument for taking blood pressures.

During the last two decades, the clinical literature has been filled
with blood pressure observations on both man and animal. The normal
blood pressure of man has been established by a score of investigators ;
Alvarez,^ Barach and Marks,= Cook,^ Dawson,-* Fisher," Frau ^^'olfen-
sohn-Kriss,^ Goepp,' Greene,** Kammerer." Lee,^" MacKenzie,^'
Michael," Sallom,^^ Smith." Sorapure," W'eysse.^'" and W'oley.'" The

* From the Physiology Department. University of Minnesota.
I.Alvarez, W. C. : Blood Pressure in University Freshmen. Arch. Int.
Med. 26:381 (Sept.) 1920.

2. Barach, J. H. : Blood Pressures, Arch. Int. Med. 13:648 (May) 1914.

3. Cook, H. \V. : Clinical Value of Blood Pressure Determinations as a
Guide to Stimulation in Sick Children. Am. T. M. Sc. 125:483. 1903.

4. Dawson: Physical Training Effect on Pulse and Blood Pressure, Am. T.
Phys. 42:590. 1917.

5. Fisher: Blood Pressure. Proc. Ass. Life Ins. Med. Directors of .\merica,
1915. pp. 90. 246; 1917, p. 203.

6. Frau Wolfensohn-Kriss. P.: Ueber den Blutdruck in Kindesalter. .-Xrch.
f. Kindcrh. 52-53:332. 1910.

7. Goepp, R. M. : Blood Pressure as a Prognostic Factor, Penn. M. J. 22:
295 (Feb.) 1919.

8. Greene, R. N. : Aviation and Blood Pressure, Lancet 1:63, 1918.

9. Kammerer : Blood Pressure in the Trenches, Miinchen. med. Wchnschr.
64:849. 1917.

10. Lee, R. I.: Blood Pressure Determinations, Urinary Findings and Dif-
ferential Blood Counts in a Group of 662 Young Male Adults, Boston M. &
S. J. 173:541, 1915.

11. MacKenzie: Blood Pressure. Proc. .^ssn. Life Ins. Med. Directors of
.■\merica. N. Y.. 1917. p. 221.

12. Michael, M. : A Study of Normal Blood Pressure in Children, .^m. T.
Dis. Child. 1:272 (March) 1911.

13. Sallom. A. K. : Standardization of Blood Pressure. New York M. I.
92:620 (Sept. 24) 1910.

14. Smith, B.: Blood Pressure Studies of 500 Men, L A. M. A. 71:171
(July 20) 1918.

15. Sorapure, V. E. : Blood Pressure and Physical Fitness of the Soldier,
Lancet 2:841 (Dec. 21) 1918.

16. Wcysse: Diurnal Variation in .Arterial Blood Pressure, Am. J. Physiol.
37:. 330, 1915.

17. Wolev. H. G.: The Normal \'ariiition of the Systolic Blood Pressure,
J. A. M. A. 55:121 (July 9) 1910.

DUXN—BLOOD PRESSURE AXD BLOOD FLOW 369

series by Goepp included 9,996 determinations; that of Fisher, 12,647;
and that of MacKenzie, 31,934. A recognized normal blood pressure
has been established also by several large insurance companies. The
Xorthwestern National Life Insurance Company, for instance, uses
an average scale ranging between 120 systolic pressure at 20 years to
136 systolic at 60 years. Much has been done to correlate high blood
pressures with certain diseases, for instance, the toxemias of pregnancy,
hypertension and arteriosclerosis.

Although blood pressure has been studied extensively, no attempt
has been made to establish a correlation between it and blood flow.
The present work is an effort to secure such a correlation. By means
of quantitative methods and the graphic analysis of the results, a
relationship between the arterial blood pressure and the efflux of
blood from a cannula of known size was obtained. The data of this
work are entirely from dog experiments. Any application of this
relationship to direct arteriovenous transfusion can be used only after
further observations on human material are made. This work was
done under the physiology department at the University of Minnesota
and material assistance was rendered by Dr. F. H. Scott and Dr. R. E.
Scammon. Credit is also due to Mr. J. F. Borg and Mr. H. R.
Smithies, who assisted in some of the experiments.

In addition to obtaining the relationship of blood pressure to
blood flow, the total blood volume of each dog was determined. These
data are reserved for a later publication.

MATERI.^L

The dog was selected as the best laboratory animal to be used for
the experiment. Eighteen dogs were used, differing in sex, variety,
and size. Xo distinction was made between sex and variety since
these factors did not bear on the problem. The size of the animal
was important, however, since it was necessary for the carotid artery
of the dog to be larger than the lumen of the cannula used for the
experiment. The dogs were all in good health and had not been used
for previous experimentation of any kind. The cannules used for the
problem were manufactured by the \\'ilson and Wilson Company,
Boston. These cannules are made from the same material as are their
nickle plated steel needles and they correspond to them in gage size.
The ends of the cannules are made blunt, however, to eliminate
experimental error.

METIICDS

A uniform technic was employed for all of tne experiments. From
twenty to thirty beakers were prepared for the collection of the blood
of the animal. Into each of these 5 c. c. of sodium citrate solution were

370 ARCHirES OF IXTERXAL MEDICIXE

measured by means of a buret. A bulb, T-shaped cannula was inserted
into the left carotid artery and one arm connected to a mercury
manometer for recording arterial blood pressure.

The internal surface of the metal cannula was coated with liquid
petrolatum to prevent the formation of a blood clot. All but a thin
film of the petrolatum was washed out by hot water. A rubber tube
was fastened to the cannula in order to convey the blood from the
carotid artery to the beakers. This tube was filled with a solution
of physiologic sodium chlorid solution so that the blood flow for the
first ten seconds would not be inaccurate. After finishing all the
Iireliminary preparations the last thing done before starting the experi-
ment was to insert this cannula in the right carotid artery. The
kymograph was started and a normal curve of blood pressure was
taken. Then the clip was removed from the right carotid and the blood
was permitted to i\o\\ through the cannula. At the end of each ten
seconds the blood flow was deviated to another beaker and the time
was indicated on the kymograph tracing. When the animal ceased
to bleed the arteries were cut and all of the blood was drained from
them. No attempt was made to perfuse the animal. The amount
of blood in each beaker was accurately measured to 0.3 c. c. The mean
blootl pressure for the respective intervals of ten seconds was obtained
from the tracings. The determinations for all the data necessary in the
experiment were now complete, except for the estimation of the area
of the cannula lumen. The diameter of the lumen of the cannula was
measured by means of a vernier caliper, accurate to one hundredth
millimeter. The cross section area of the lumen was calculated by
using the formula for the area of a circle: .^ren = 3.1416 (r=) in
which r is the radius in mm.

ACCUR.\CV OF THE D,\T.\

. It is essential to estimate the accuracy of the <lata in order to obtain
a relationship between the three variables, blood pressure, blood flow
and the size of the lumen of a cannula. The cross section area of
tlie lumen of the cannula can be determined with considerable exactness.
.\11 the cannules were measured in the .same manner. The diameter
of each was obtained by averaging thirty readings of the vernier
caliper. By this means an accurate determination of the cross section
area of the cannula was obtained. The blood flow per ten seconds
is also an accurate measurement since it is not distorted in any way
by experimental error. The blood pressure, however, is not as exact
an observation. The necessity for an absolutely accurate baseline was
not obvious while the experiments were being carried out. The error
in its determination occurred because the third arm of the three way
cannula was not kept horizontal to the carotid artery. The variation
ranged from 2 to 10 mm. Hg.

DU NX—BLOOD PRESSURE AND BLOOD FLOW 371

TREATMENT OF THE DATA

The accuracy of the data is very important for the successful
interpretation of the physiologic relationship between the blood pressure
and the blood flow. The significance of the results, however, can be
obtained only by an application of graphic and mathematical analysis
to the data. The methods used in this series of experiments are three
in number: (1) the construction of field graphs and drawing curves
by inspection; (2) the determination of a general empirical formula,
and (3) the comparison of the inspected absolute curves to the respec-
tive curves derived by the general empirical formula.

1. The construction of field graphs, together with the establishment
of curves showing the central tendency of the data, was the first method
of graphic analysis which was attempted. Each graph (Figs. 1, 2, 3,
4, 5 and 6) includes the data which were obtained by the use of a
certain cannula. In each figure the abscissa represents blood flow in
c. c. per ten seconds, while the ordinate is the blood pressure of the
animal in nun. of Hg. The curve drawn by inspection does not signify
a curve of mathematical accuracy but merely represents the best
expression of the judgment of the author as to the central tendency
of the data.

Figure 8 is also a field graph. In this instance the abscissa repre-
sents the area of the lumen of the cannula. The ordinate shows the
average blood flow per ten second interval for the respective cannulas.

2. After the field graphs were made and the inspected curves were
drawn for each, a general formula was calculated expressing the blood
flow in terms of the blood pressure and the cross section area of the
cannula. This general formula was obtained by the method of trial
and error. In determining the equation, the values of the cross section
area of the cannula were kept unchanged. Blood pressure values
were also kept as nearly similar to the respective observed determina-
tions as the nature of the data would permit. The best general formula
can be expressed as follows: Blood flow (c. c.)=0.17 (area [sq.
mm.]) (Blood pressure [mm. Hg.]) in which blood flow is the efflux
of blood from a cannula per ten seconds of time, area is the cross
section area of the lumen of the cannula in sq. mm., blood pressure
is the average blood pressure in mm. of Hg., and the number 0.17 is
an empirically determined constant.

3. The comparison of the absolute values of the curves drawn by
inspection with the corresponding determinations of the general formula
is demonstrated graphically and numerically. On each absolute graph
(Figs. 1, 2, 3, 4, 5 and 6) both the inspected and the calculated curve
are inserted, the fomier bv a solid line and the latter by a broken line.

ARCHIVES OF IXTERXAL MEDICISE

ZA&oL

Fig. 1. — A field graph and curves expressing the relationship between blood
flow and blood pressure. A cannula was used which had outlet 1.131 sq. mm. in
area. Abscissa : efflux of blood in cubic centimeters per 10 seconds of time.
Ordinate : blood pressure in millimeters of Hg. Individual cases are indicated
by solid dots (for Experiment 1) and by circle-crosses (for E.xperiment 2).
The solid line represents a curve drawn by inspection. The broken line is a
curve drawn to the general empirical formula: Blnod flow (c.c.) =:0.17 (area
[sq. mm.]) (blood pressure [mm. Hg.]).

DUXX— BLOOD PRESSURE AXD BLOOD FLOW

KHJ

— i —

1

1 I

130

-

^ /

V

leo

-

e /

-

110

-

. /

-

100

-

/

-

90

-

///

-

60

-

-

70

-

//'

-

60

-

/

/A

-

50

"

//

/• •

~

40

-

//

-

30

"

/

/ •*

~

10

" /

9

-

10

/

-

c

/,

i_

1 1

1 1 1 1 1 1 1

c

) 5

10

15 £0

25 30 35 40 45 50 55

~B.c

Fig. 2. — A field graph and curves expressing the relationship between blood
flow and blood pressure. A cannula was used which had an outlet 2.378 sq.
mm. in area. Abscissa: efflux of blood in c. c. per 10 seconds of time. Ordinate:
blood pressure in mm. Hg. Individual cases are indicated liy solid dots (for
Experiment 3), by circle-crosses (for Experiment 4) and by open circles (for
Experiment 5). The solid line represents a curve drawn by inspection. The
broken line is a curve drawn to the general empirical formula: Blood flow
(c.c.)=:0.17 (area [sq. mm.]) (blood pressure [mm. Hg.1).

374 ARCHIVES OF IXTERXAL MEDICIXE

In Figure 7 these curves are grouped to a uniform scale. The
curves drawn by inspection are paired with those which were calculated
and the corresponding pairs numbered according to the respective
graphs.

The numerical comparison of the respective values which are
taken from the curves drawn by inspection and those calculated by
the general empirical formula is seen in Tables 1 and 2.

Table 1 contrasts in parallel columns the observed and the cal-
culated values of both the average blood flow and the average blood
pressure for each experiment.

Table 2 shows the observed and the theoretical values of the
average blood flow and the average blood pressure for each graph.

EXPERIMEXT.AL OBSERVATIONS

A summary of the observations made on the data is divided for
convenience into four subdivisions: (1) a consideration of the absolute
data which are presented by means of field graphs and tables; (2) an
enumeration of averages obtained from this absolute data; (3) a
comparison of the observed and calculated values, and (4) the signi-
ficance of the relationship between the blood flow and the size of the
cannula outlet.

1. The absolfite data available for this study are set forth in the
field graphs (Figs. 1. 2, 3, 4, 5 and 6) and in Table 1.

Figure 1 is a field graph based on two experiments. Each of the
dogs selected weighed 4.54 kg. The cannula used was gage 16 and
had an opening 1.131 sq. mm. in area. The curve, drawn by inspection,
shows that 26 c. c. of blood flows from the cannula in ten seconds
when the blood pressure is 140 mm. of Hg. This amount of blood
outflovk' steadily decreases until at 25 mm. of arterial pressure only
3 c. c. of blood flows from the cannula per ten seconds of time. The
curve indicated by dashes is the general empirical formula : Blood flow
(c. c.)=0.17 (area [sq. mm.]) (Blood pressure [mm. Hg.]). It is
slightly lower than the curve drawn by inspection.

Figure 2 is a field graph based on three experiments, the dogs
■weighing 5.67, 13.4 and 7.72 kg., respectively. The cannula used in
these experiments was gage 14 in size and had a cross section area of
2.378 sq. mm. The curve drawn by inspection shows that 55 c. c. of
blood leave the cannula in ten seconds when the blood pressure is 140
mm. of Hg. Tlie arterial pressure descends rapidly to 40 mm. at
which point the efflu.x of blood in ten seconds is 20 c. c. The curve
drawn by calculation from the general formula is a straight line which
falls slightly below the inspected curve at a blood pressure above
100 mm. of Hg. and slightly above the inspected curve at an arterial
pressure below 100 mm. of Hg.

DUXX— BLOOD PRESSURE ASD BLOOD FLOW

lOOcc

Fig. i.—:\ field graph and curves expressing the relationship between blood
flow and blood pressure. A cannula was used which had an outlet 3.597 sq.
mm. in area. Abscissa : efflux of blood in c.c. per 10 seconds of time. Ordinate :
blood pressure in mm. of Hg. Individual cases are indicated by solid dots
(for Experiment 6), by circle-crosses (for Experiment 7) and by open circles
(for Experiment 8). The solid line represents a curve drawn by inspection.
The broken line is a curve drawn to the general empirical formula : Blood
flow (c.c.)=fl.l7 (area [sq. mm.]) (blood pressure [mm. Hg.]).

ARCHIVES OF IXTERXAL MEDICIXE

110 \z6cc

Fig. 4. — A field graph and curves expressing the relationship between blood
flow and blood pressure. A cannula was used which had an outlet 5.2 sq. mm.
in area. Abscissa: efflux of blood in c.c. per 10 seconds of time. Ordinate:
blood pressure in mm. of Hg. Individual cases are indicated by solid dots
(for Experiment 9), by circle-crosses (for Experiment 10) and by open circles
(for Experiment 11). The solid line represents a curve drawn by inspection.
The broken line is a curve drawn to the general empirical formula: Blood
flow (c.c.) =^0.17 (area [sq. mm.]) (blood pressure |nim. Hg.]).

DUSX— BLOOD— PRESSURE AXD BLOOD FLOW 377

Figure 3 expresses the absolute relationship between blood pressure
and blood outflow per interval of time when the cannula has a cross
section area of 3.597 sq. mm. Three experiments were carried out
to establish this relationship, the dogs weighing 9.54, 20.45 and 15.2
kg., respectively. The blood pressure decreases more rapidly in amount
than does the blood flow per unit of time. At 127 mm. of arterial
pressure there is approximately 80 c. c. of blood flow from the carotid
in ten seconds. The blood pressure decreases rapidly to 25 mm. and
the blood flow to 23 c. c. per ten seconds. The calculated curve is
at all times higher than the curve drawn by inspection.

The absolute data set forth in Figure 4 varies more widely from the
empirical formula than do those of any other field graph. Three dogs,
weighing 28.1, 10.13 and 21.8 kg., respectively, were used to establish
this relationship. The cross section area of the cannula was 6.07
sq. mm. The inspected curve is distinctly concave descending rapidly
from a blood pressure of 118 mm. of Hg. and a blood outflow of 120
c. c. per ten seconds of time to a blood pressure of 50 mm. of Hg. and
a blood outflow of 68 c. c. per ten seconds. The descent is then more
slow until the efflux of blood per ten seconds of time reaches 37 c. c.
at 25 mm. of arterial pressure. The curve of the general empirical
formula is definitely higher than the inspected curve although it is
approximately correct for the higher blood pressures.

Figure 5 shows the relationship of blood pressure to the efilux
of blood when a cannula with an outlet area of 7.548 sq. mm. was
employed. Four dogs weighing 12.3, 22.7, 20.4 and 27.3 kg., respec-
tively, were used. The individual cases are scattered in this figure
although the central tendency is quite obvious and is expressed by the
curve drawn by inspection. This curve shows that 220 c. c. of
blood leave the cannula in ten seconds of time when the blood pressure
is 159 mm. of Hg. The curve descends steadily to 37 c. c. of blood
outflow at 25 mm. of arterial pressure. The curve drawn by the general
empirical formula is slightly higher than that drawn by inspection.

The experiments set forth in Figure 6 were made on three dogs
weighing 26.4, 20.0 and 25.1 kg., respectively. The cannula used to
drain oflf the blood had a cross section area of 9.512 sq. mm. The
central tendency of the individual cases in this figure is expressed by •
the inspected curve. This curve shows that 200 c. c. of blood flows
from the cannula in ten seconds when the arterial pressure is 135
mm. of Hg. and that the amount of blood outflow per ten seconds
of time descends steadily to the amount of 35 c. c. at a blood pressure
of 25 mm. of Hg.

The absolute data for each experiment are summarized in Table 1.
The number of observations on each dog and the number of experi-
ments for each figure are given. Tiie body weight of the dog before

ARCHIVES OF ISTERXAL MEDJCIXE

£00 ZZCkic

Fig. S. — A field graph and curves expressing the relationship between blood
riow and blood pressure. A cannula was used which had an outlet 7.548 sq.
mm. in area. .Abscissa: efflux of blood in c.c. per 10 seconds of time. Ordinate:
blood pressure in mm. of Hg. Individual cases are indicated by solid dots
(for Experiment \2), by circle-crosses (for Experiment 13), by open circles
(for Experiment 14) and by circle-dots (for Experiment 15). The solid line
represents a curve drawn by inspection. The broken line is a curve drawn to
the general empirical formula: Blood flow (c.c.)=0.17 (area [sq. mm.]) (blood
pressure [mm. Hr.] ).

DUXX— BLOOD PRESSURE AXD BLOOD FLOW

tea e£Ocr

Fig. (■>. — A field graph and curves expressing the relationship between blood
flow and blood pressure. A cannula was used which had an outlet 9.512 sq.
mm. in area." Abscissa: efflux of blood in c.c. per 10 seconds of time. Ordinate:
blood pressure in mm. of Hg. Individual cases are indicated by solid dots
(for Experiment 16), by circle-crosses (for Experiment 17) and by open circles
(■for Experiment 18). The solid line represents a curve drawn by inspection.
The broken line is a curve drawn to the general empirical formula: Blood
flow (c.c.) =0.17 (area [sq. mm]) (blood pressure [mm. Hg.]).

ARCHIVES OF IXTERXAL MEDICINE

TABLE 1. — Absolute Data for Individual Experiments
Blood Volume (C.c.)=0.17 (Area [Sq. Mm.]) (Blood Pressure [Mm. of Hg.])

Of

Expert-

Number

in
Experi-

Weight
of
Dog,
Kg.

Total
Volume

of
Blood,

Average Blood Flow,

Average Blood Pressure,
Mm. of Hg

Number

of

Graph

Observed

Calculated

from
Observed

Blood
Pressure

Observed

Calcul ited

ObsMved

Average

Blood

Flow

13

4.54

200

10.30

10.58

55.0

54.0

2

20

4.54 ■

234

10.05

10.98

57.2

62.3

6

5.67

329

22.30

16.30

40.3

55.4

Z

4

19

13.40

806

33.60

27.85

69.0

83.5

■'

9

7.72

381

28.30

31.12

77.0

70.S

12

9.34

488

31.40

27.35

44.7

51.4

3

■ 17

20.43

1,074

52.00

47.90

78.3

84.7

15

15.20

825

46.40

36.90

60.4

75.8

17

28.10

1,333

61.70

55.E0

54.0

59.7

4

6

10.13

408

50.70

47.40

45.8

49.0

11

21.8

1,215

75.60

72.10

69.8

73.0

11

12.3

633

48.«>

71.90

66.0

36.9

5

7

22.7

948

83.00

69.40

54.0

64.8

9

20.4

790

67.00

68.50

53.4

52.4

15

9

27.3

1,204

86.10

73.40

57.2

67.2

14

26.4

1,509

77.60

78.40

48.6

48.0

6

17

9

20.2

737

66.30

75.0

46.4

34.8

9

23.1

1,206

88.30

97.0

60.0

54.6

T.^BLH 2.— .\i!SOLUTE Uat.\ for Emu (iRAPn
(C.c.)=0.17.(Arf.a [Sg. Mm.] ) (Blood Pressurk (Mm. of He,.])

Blood Pressure (Mm. of Hg. =

0.17 Area (Sq.Mm.)

Average Blood Pressure.

Cannula

O.c.

ofHg

Numbers

1 Calculated

Calculated

i from

from

of

of

mate

Observed Observed

Observed

Lumen,

Average

^B^^'o"!^

Mm.

Sq.Mm.

Pressure

Volume

land 2

0.60

1.131

16

10.2 1 10.8

66.2

53.0

30.2 26.7

«6.1

74.7

Band 6

1.07

3.697

13

44.4 38.8

6S.0

6.070

11

64.2 59.6

9 and 10

7.548

69.1 70.8

66.2

63.7

11 and 12

1.74

9.612

1 -

74.6 82.6

61.1

DUNN— BLOOD PRESSURE AND BLOOD FLOW

> 1 1 1 1 1 1 1 1 1
/

//
//

! 1 1 1 1 1

/ /
/ /

140

/ //

/ / /

130

// //

/ / /

// //

/s/ /

lEO

r 7

/ / / ^
/ // /'

/ U

/

/ //s/

1 10

A

/ // /

\ I 1 J

n

/// /

V / // /

/ / ^

100

■ ///

/

/ .
/ /

/ /

90

90

/ / /
/ /

/
/

- / // //#

' /

70
60

1 / /// /

50

// / / / ////

-

40

1 / ////
\\\i / /7//

-

30

-\ 1 / // ///
III ////
111 //p

"

ZO

V if

III l/ir/

10
0

\ 1 iJf

1 1

1 1 1 1 1 1

20 40 60 80 100

\m '-

140 100 180 ?00

Fig. 7. — A series of curves illustrating tlie relationship between blood
pressure and the efflux of blood per 10 seconds of time. These curves are
taken from Figures 1, 2, 3, 4, 5 and 6. Abscissa: Blood flow in c.c. per
10 seconds. Ordinate: blood pressure in mm. of Hg. The solid lines indicates
curves drawn by inspection. The broken lines arc curves drawn tn the gen-
eral empirical formula: Blood flow (c.c.")=0.17 (area fsq. mm.]) (blood pres-
sure [mm. Hg.]).

382 ARCHIVES OF IXTERXAL MEDICIXE

the operation is indicated in kilograms and the total amount of blood
drained from the arteries during the experiment is noted in c. c.

2. An enumeration of arithmetic means obtained from this data
has been carried out for the individual experiments (Table 1) and for
these experiments grouped according to the cannula used (Table 2).

3. A comparison of the observed and the calculated absolute values
can be made by a survey of the collected absolute curves (Fig. 7) or by
inspection of the adjoining columns of observed and calculated deter-
minations in Figures 1 and 2.

4. The curves taken from the field graphs are collected for the
purpose of comparison in Figure 7. Considering the variability of
the blood pressure due to experimental error, the inspected curves from

9 lOsqjivTi

Fig. 8. — A curve to show the relationsliip between blood flow and the size
)f the cannula outlet. Abscissa: average efflux of blood in c.c. per 10 seconds
)f time. Ordinate: the area of the cannula outlet in sq. mm. The solid line
s a curve smoothed by inspection.

four field graphs (Fig. 1,2, 5 and 6) fit closely the general empirical
ftjrmula. The inspected and the absolute curves from Figures 3 and 4,
however, show a marked deviation, especially in the intermediate
\alues of the curve.

In Table 1 the observed absolute values of both blood flow and
blood pressure can be compared to the calculated values of blood flow
and blood pressure respectively.

Table 2 demonstrates the correlation of the observed blood flow and
the observed blood pressure with the respective calculated blood flow
and blood pressure values.

5. The significance of the relationship between blood flow and the
size of the cannula outlet is portrayed in Table 2 and Figure 8. The
absolute averages of observed blood flow per ten seconds of time
(Table 2) increases with the size of the cannula. In Figure 8 this

DUXX— BLOOD PRESSURE AND BLOOD FLOW 383

relationship is expressed graphically. The abscissa represents the cross
section area of the opening of the cannula in sq. mm. ; the ordinate
gives the average efflux of blood per ten seconds of time. The curve
based on this absolute data is logarithmic in type. The amount of
blood flowing from the cannula in a .given length of time increases
rapidly at first as a larger cannula is used. When cannula gage 11 is
employed (which has a cross section area of 6.07 sq. mm.) the blood
flow per unit of time increases at a slower rate. The efflux of blood
reaches a maximum average quantity of 75 c. c. per ten seconds when a
cannula, 9.512 sq. mm. in area, is used.

The significance of this relationship between the flow of blood and
the lumen of a cannula is not understood at the present time. It
may be explained in several ways, none of them entirely satisfactory.
At any rate, the general empirical formula is definitely influenced
by this factor. A point is reached at which the size of the cannula
makes little difference in the amount of blood flowing from an artery.

SUMM.\RY

The relationship of the blood pressure, the efflux of blood from
the carotid artery of a dog, and the lumen of a cannula, the cross-
section area of which ranges between 1 and 10 sq. mm., can be expressed
by the general empirical formula :

Blood flow (c. c.)=0.17 (Area [sq. mm.]) (Blood pressure [mm.
Hg.]).

STUDY OF- BLOOD SUGAR CURVES FOLLOWING
A STANDARDIZED GLUCOSE MEAL*

W. H. OLMSTED, M.D., and L. P. GAY, M.D.

ST. LOUIS

The following study is concerned with the eftort to demonstrate the
main factors which influence the duration of hyperglycemia after a
glucose meal. More than 200 cases have been studied critically. There
is extensive literature on the subject of blood sugar curves after various
sorts of carbohydrate meals. In some studies standardized meals were
given, in others not. Many clinicians have assigned diagnostic impor-
tance to an increased hyperglycemia following glucose ingestion.

The interpretation of the value of sugar curves depends on the
following factors: (1) The technic of the adminf-stration of the
glucose meal; (2) the collection of blood samples; (3) the method
of doing the blood sugar determination, and (4) the wide application
of the test so as to learn the many factors which influence these curves.

We believe it necessary, in order to support our conclusions, to
discuss the first three of these points in detail.

Standardised Glucose Meal. — The standardized Janney^ glucose
meal was used. Glucose is the sugar of choice because it is most
readily absorbed and because there are data as to the rate of its
absorption from the gastro-intestinal tract. Fisher and Wishart -
and Janney ' have shown, by different methods, that from 66 to 8Q
per cent, of injected glucose is absorbed in the course of two hours.

The question of absorption is an important one. The curves here
studied would indicate that absorption is fairly constant for the
individual. This is brought out by the similarity of repeated curves
(Table 1) on the same individual. The constancy of these curves is
rather striking, especially where the symptoms and signs have not
greatly changed. It is probable, therefore, that the absorption rate is
fairly constant, at least for each individual. Again, in analyzing two
hundred curves only six were found with a sudden increase in hyper-
glycemia at the end of the second hour, there having been no
hyperglycemia the first hour. In other words, nearly all curves reach
the maximum at the end of the first hour and the second hour levels
are u.sually lower, or only slightly higher, than the first hour ones.
This would show that absorption is quite rapid.

* From the Metabolic Unit of the Department of Medicine, Washington
University School of Medicine.

1. Janney: Proc. Soc. Expcr. Biol. & Med.. 15: 1917-1918. Janney and
Isaacson: J. A. M. A. 70:1131 (April 20) 1918.

2. Fisher & Wishart: T. Biol. Chcm. 13:49. 1912.

3. Janney: J. Biol. Chem. 22:191, 1915.

OLMSTED-G AY— BLOOD SUGAR 385

Sansum and Woodyatt * have shown that the maximum intravenous
tolerance rate of man and animals without glycosuria is 0.85 gm. per
hour per kilo of weight. Intravenous tolerance methods, although
scientifically desirable, are not practical. . The known facts for deter-
mining a standard alimentary dose are: The above noted intravenous
tolerance and the average absorption rate of 66 per cent, in two hours.
The theoretical dose would be (2x0.85x -W^ ) or about 2.5 gm.
sugar per kilo. Janney has recommended 1.75 gm. per kilo as a

TABLE

;.\TED CURVES

Diagnosis

Date

Blood Sugar Values, per Cent.

Curve
Classifi-
cation

Fasting

1st Hr.

2dHr.

3dHr.

12/21/20
2/ 4/21
2/ 8/21
2/14/21

12/13/20
1/10/21
1/20/21
12/ 8/20
12/13/20
6/ /20

ioi-n/20

0.092
0.092
0.086
0.085
0.090
0.090
0.090
0.100
0.140
0.100
0.090
0.16S
O.lTo

0.124
0.120
0.190
0.185
0.110
0.110
0.160
0.183
0.192
0.110

5S

0.065
0.085
0.150
0.1T5

0!085
0.170
0.195
0.140
0.090
0.080
0.220
0.220

0.OT6
0.063
0.130
0.150
O.065
O065
0.140
0.175
0.140
0.080
0.090
0.113
0.095

s

s

•y

Manic-depressive psychosis

s

III
III

s

II

TABLE 2. — Showing Composition of Glucose Meal

Weight, Pounds

Glucose, Gm.

Lemon Juice, O.c.

Water, C.c.

72
80
88
96
104
112
120
128
136
144
152
160

\fe

184
192

66
72
78

^ 1

96
102
108

114 j
120 1
126
132
138
144

100

140

154

130

172

140

196

m

238

aoo

280

230

322

33S

Standard dose. This amount of glucose for a man weighing 150 pounds,
for instance, would amount to 120 gm. of glucose, or 480 calories. Such
a man in basal state plus 10 per cent, increase for the specific dynamic
action of glucose would be burning from 70 to 75 calories per hour.
This dosage of glucose is, therefore, greatly in excess of caloric needs
under ordinary circumstances and would show, with an absorption
efficiency of 66 per cent., the glycogenic function of the individual or
his abilitv to store the excess of an amount of sugar, commen.surate

4. Sansum and Woodyatt: J. Biol. Chem. 30:155. 1917.

386 ARCHIVES OF IXTERXAL MEDICIXE

with his weight, absorbed in a unit of time. Sansum and Woodyatt *
have also shown that injecting animals in excess of the tolerance rate
with varying concentrations of glucose made no difference in the amount
of glucose in the urine, but that it did make a difference in the height
of the blood sugar. It is, therefore, better technic to use a 40 per cent.
solution of glucose using water and lemon juice as a solvent.

It seems obvious that one should not give 100 gm. glucose to an
individual weighing 200 pounds and a like amount to one weighing 100
pounds and expect duplicate results in blood sugar concentration. It
has been fairly well established that blood volume varies approximately
with weight. Normally sugar is stored both in the liver and in the
muscles. Palmer ° found that in the diabetic animal the amount of
glucose in the tissues varied with the hyperglycemia. It is possible,
therefore, that sugar storage might take place in some abnormal
conditions in other tissues besides the liver and muscles, and it would
seem that the weight of the individual is our best index to his available
space for storing glucose.

T.\BLE 3. — Showixc. the Abse.xce of Rel.vtionship Betwee.n Weiuht
.WD THE Type of Curve

Number o£ Individuals

Weiglit in Pounds *

Normal Curve ' Subnormal Curve

90-100...... , i 3

100-110 8 5

110-120 ' 2 6

120-130 6 5

130-140 6 2

140-lSO I 6 5

150-160 .• ' S 2

160-180 1 1 2

18O-200 5 2

200 plus 2

* What the average weight of aU hospital patients is, is not known, but the table shows
the size ol the patient does not influence the curve.

In some of the latest work in blood sugar determinations '' after a
glucose meal, a constant amount of glucose was used for all individuals.
The results would have been more constant if the weight of the
individuals had been taken into consideration (Table 3). The same
can be said for the amount of water given with the meal.

Blood Sugar Determinations. — The blood was drawn with a 5 or 10 c.c.
syringe and introduced into a test tube containing a few crystals of potassium
oxalate and gently shaken. The first 'specimen was taken with the patient in
a basal state, twelve hours after the last food. After the collection of the
blood, the standardized meal was given. The blood was drawn again at the
end of one, two and three hours. The third hour specimen was found to be of
great importance. It is necessary that the blood should be precipitated imme-
diately after taking the sample.

5. Palmer: J. Biol. Chem. 30:79, 1917.

6. Allen, Wishart and Smith: Arch. Int. Me.l. 24:.SJ3 (Oct.) 1919. Boothby ;
I. A. M. A. 77:252 ( lulv 23) 1921.

OLMSTED-GAV— BLOOD SUGAR

387

Meyers and Bailey's ' modification of Benedict's first method was used : 3
c.c. blood was added to 12 c.c. saturated picric acid solution and a few crystals
of picric acid. The blood was thoroughly shaken and filtered after five
minutes. The standard glucose solution was made up of 1 mg. glucose to S

T.ABLE 4. — CoMP.\RisoN of Blood Sug.\r V.^lues
Shaffer's Methods

Benedict's and

Case

Fasting 1st, 2d and 3d Hours

Classification

Benedict's Method,
per Cent.

Sbailer's Method,
per Cent.

By Benedict

By Shafler

^

0.089
0.135
0.123
0.100

0.077
0.149
0.125
0.064

N*

N

2

0.087
0.159
0.136
0.117

0.096

o!l43
0.125

U

11

S

0.103
0.169
0.135
0.095

0.119
0.074

N

N

*

0.104

^^^

0.064

0.096
0.090
0.099
0.095

S

S

5

O.066
0.163
0.156
0.117

0.099
0.143
0.138
0.127

II

III

S

0.095
0.148
0.167
0.137

0.090
0.157
0.166
0.109

III

II

"

0.143
0.195
0.138
0.142

0.115
0.195
0.142
0.114

III

III

8

o!l90
0.206

0.098
0.176
0.181

■ III f

III ?

8

0.102
0.096
0.088

0.095
0.082
0.088

s

8

10

0.089
0.118
0.091

0.079

. 0.106

0.093

s

^

11

0.126
0.236
0.154
0.139

0.113
0.191
0.156
0.135

m

III

12

0.082
0.224
0.226
0.143

0.072
0.230
0.189
0.101

III

II

normal; S, subnormal; II, 2 hour sustained; III, 3 hour sustained.

c.c. saturated picric acid; this solution being part of the same sample of
saturated picric acid that was used to precipitate the blood.

Cowie and Parsons' have shown how much more sensitive picric acid
solutions are to such substances as acetone, diacetic acid and cpinephrin than

7. Meyers and Bailey:

8. Cowie and Parsons

J. Biol. Chem. 24:147. 1916.
Arch. Int. Med. 26:333 (Sept.) 1920.

388

ARCHIVES OF IXTERXAL MEDICI XE

to sugar itself. This work explains the very high blood sugars obtained in
the cases of diabetes mellitus, but in the absence of acetone bodies the method
is fairly accurate.

Since the publication by Shaffer and Hartmann ' of their iodometric method,
a considerable number of curves have been determined by both methods. In
only minor respects has the classification of curves been changed by values
obtained by the Shaffer method. In fifty blood sugar determinations done by
both methods varying from 0.06 to 0.30 per cent., the averages at different
levels by the two methods agree within a few per cent.

'1

/^

.iS

/

'/^"^

\ '

.10
-Dt

/

\

\

/

^--.^^r;

.%

iSr

IS^s ^^

Fig. 1. — In each curve the two lines represent the limits between which the
curves of that type falls. Ordinates show percentage of blood sugar. .-X, Normal
curve; B, subnormal curve; C, Type 1 1 curve; D, Type III curve.

Classification of Curies. — In dealing with a considerable number
of curves some sort of classification is necessary. This introduces the
question of terminology. Sugar tolerance work began by- feeding
sugar by mouth and watching for its appearance in the urine. If an
individual could take 100 gm. gluco.se and show no sugar in his urine,
his "tolerance" would be considered normal ; if he could take more than
100 gm. without glycosuria his "tolerance" was increased, and if he
showed glycf)suria on taking 100 gm. his "tolerance" was decreased.

9. Shaffer and Hartmann: J. Biol. Chem. 45:365. 1921.

OLMSTED-GAV— BLOOD SUGAR 389

It is impossible to transpose this term "tolerance" into the ter-
minology of blood sugar because a higher "tolerance" means a low blood
sugar curve and a low "tolerance" means a high blood sugar curve.
It has been our experience that the use of the word "tolerance" only
leads to confusion and it would, therefore, seem to be better to use
the term "blood sugar curve after glucose meal" or "blood glucose
curve."

In consideration of the classification of curves it has not been
considered important to include fasting hyperglycemia because it has
been found that a fasting hyperglycemia is not a common condition,
except where there is a loss of power to oxidize glucose, or in the
presence of ashyxia or severe toxemia. As a rule, fasting hyperglycemia
has not been found in endocrine or neurologic cases.

Normal curves were constructed on five normal individuals and
also on about forty patients in the hospital who showed no demonstrable
cause for a disturbed glycogenic function. These normal curves agree
with those of other observers " in that after the normal fasting level
there is a hyperglycemia at the end of the first hour of fiom 0.14
to 0.19 per cent, and at the end of the second hour the blood sugar
level is within normal limits, or from 0.08 to 0.12 per cent. The third
hour is still within normal limits from 0.06 to 0.12 per cent. We have
assumed a higher normal hyperglycemia at the end of the first hour
than other observers. However, this seemed justified when the pro-
cedure was so well controlled and the clinical data carefully studied.
Many of the normal curves show a marked hypoglycemia at the end of
the second and third hours. This hypoglycemia may be to the extent of
from 0.06 to 0.08 values — so low that mere changes in blood volume
would scarcely account for them. We cannot offer an explanation
but have observations to show that by the end of the fourth hour the
blood .sugar values return to normal levels.

Abnormal curves have been divided into two main classes. The
sustained curve, showing an abnormally sustained hyperglycemia, and
the subnormal curve, which shows no normal hyperglycemia and even a
hypoglycemia after a glucose meal. The curves showing sustained
hyperglycemia we have divided into two groups : those showing hyper-
glycemia the second hour but with a return to normal levels the third
hour; and those showing hyperglycemia at the end of three hours. The
reason for this division of sustained curves into two groups will
appear later.

Subnormal curves show a hypoglycemia or a normal fasting blood
sugar. No hyperglycemia follows the administration of a glucose
meal. The failure of the api)earance of hyperglycemia may be due to
one or both of two possibilities: Either a delayed absorption rate or an

10. Haminon and Hirschman: .^rch. Int. Med. 20:761 (Dec.) 1917.

390 ARCHIVES OF INTERNAL MEDICINE

increased glycogenic function. There are data " to show that in one
condition, hypothyroidism, there is no delayed absorption in spite of
the subnormal curves. Delayed absorption may occur in some condi-
tions, such as hypopituitarism.

DISCUSSION OF M.\TERI.AL

In the cases studied we noted age ; weight ; pulse rate and tempera-
ture; diagnosis (as obtained from the history sheet) ; gonads; children;
menses; sexual power and desire; sympathetic symptoms; sweating;
vasomotor instability ; emotional tendencies, fear, anxiety ; reflex
excitability; gastro-intestinal symptoms ; pituitary : sella (roentgen ray) ;
bones ; hair ; eyegrounds ; visual fields ; secondary sexual characters ;
thyroid: vascular activity in gland itself; tremor; external ocular
movements ; size ; exophthalmos ; skin ; special tests : basal metabolism ;
goetsch; hemoglobin.

In this study emphasis has been laid particularly in the selection
of cases, on the so-called suspected endocrine disturbances of the
thyroid and the pituitary; on the fatigued states, and on the hysterias
and true dementias.

Many other conditions show abnornial curves, but the nature and
constancy of their influence on the glycogenic function is even more
uncertain than the above mentioned conditions. Such conditions are
any mild toxemia, such as that in low grade bacterial infection ;
malignancy, etc., acidosis of any origin; drugs, such as opium and its
derivatives, or salicylates. It is probable that these conditions can
disturb blood sugar curves; certainly they affect general metabolism
to some degree.

In this type of case the curves presented are not as numerous
as one would wish. A few furunculosis cases (Fig. 2) show sustained
curves. Some carcinomas of the gastro-intestinal tract, especially when
metastases have taken place, show the same curve. Others with
localized carcinoma show a normal curve. A normal curve is usually
found in arthritis, but in the presence of fever or after foreign protein
injections there is a high curve. If sugar curves are to be of value
from an endocrine or a neurologic point of view, such conditions as
may disturb glycogenic function should be avoided. Too little is
known about them and there is no reason to believe their influence
on the glycogenic function is a constant one. It seems more probable
that a sustained curve in such conditions is a ])art of the general
effect of incidental toxemia rather than a specific characteristic of a
definite metabolic disturbance.

11. Janney and Isaacson: Arch. Int. Med. 22:160 (Aug.) 1918. Janney
and Henderson, Arch. Int. Med. 26:297 (Sept.) 1920.

OLMSTED-GAY— BLOOD SUGAR :V)\

Focal Infection.— Ptmhtrion'^- has shown the effects of low grade
inflammatory infection on delaying glycogenesis. Not only in cases
of arthitis, but in other focal inflammatory processes, he found higher
curves than in his normals. He did not use the standardized glucose
meal. In a few cases he observed a return to normal curves after the
removal of the foci of infection.

The manner in which toxins may influence the height of blood
sugar curves is open to much speculation.

1. The toxin may directly stimulate the action of the diastase of
the liver and muscles, or it may inhibit their glycogenic power/'^
Langfeldt '* has recently shown in vitro the optimum pn at which liver
diastase works in the presence of thyroid extract and epinephrin.
Toxins from foci may possibly disturb the hepatic acid-base equilibrium.

Fig. 2. — A and B curves are of arthriti.s cases. Curve A was taken two
days following the intravenous injection of foreign protein. B. normal curve
obtained in most cases of arthritis. C and D, curves of cases of furunculosis.

2. Focal toxins may act on glycogenolysis through their eff^ect on
suprarenal medulla directly or through autonomic reflex.

3. Focal toxins may also act on higher cerebral centers.

The association of fatigued state with focal infection is often noted.
The curves of such conditions will be discussed later. Again, the focal
toxins may disturb the mixture of food stuffs burned in the cell. The
well known protein-sparing property of carbohydrate, especially in
long sustained fevers, suggests that carbohydrate is burned most
readily and possibly is mobilized to protect protein. The possibilities
of the effects of toxins on glycogenesis and glycogenolysis have not

12. Pemberton and Foster: Arch. Int. Med. 25:243 (Feb.) 1920.

13. Lusk: Science of Nutrition, p. 522, quoting Rosenthal, who showed
that injection of diphtheria toxin prevented glycogen formation.

14. Langfeldt: J. Biol. Chem. 46:381, 1921.

392 ARCHIJ-ES OF IXTERXAL MEDICI. \E

been exhausted, but enough has been mentioned to show the complex-
ities of the possibiHties. Certainly at present it is better to suppose
that disturbance of glycogenic function in focal infection is a manifes-
tation of the effects of infection just as hyperpyrexia or esthenia. The
rationale of restricted diet in these cases can be questioned: Why
deprive these patients of the protein sparing property of carbohydrate?
Why feed typhoid fever patients carbohydrate and deny it to the
arthitic? The great losses of weight seen in chronic arthitis would
suggest that such patients are greatly in need of carbohydrate in
abundance. The metabolism of arthitis does not diflfer from that of
any other chronic focal infection. Certainly there is no loss of power
to oxidize glucose, nor is there reason to believe that products of
carbohydrate oxidation have a deleterious effect on periarticular inflam-
matory processes. Even if the toxins of the agent of infectious arthitis
do cause sugar mobilization, that in itself should not suggest carbohy-
drate denial as a therapeutic indication. The experience of this clinic
with low carbohydrate diet in arthitis has been quite disappointing.

Thyroid Diseases. — The influence of the thyroid gland on sugar
curves has long been appreciated. The internal secretion of the thyroid
excites two influences: (1) the delaying of glycogenesis, or an increased
glycogenolysis ; (2) an increased or stimulated metabolism.

When thyroid is fed carbohydrate is burned rapidly. One would
therefore suppose that curves of exophthalmic goiter patients (Fig. 3)
would be high but fall quickly, the rapid fall being associated with the
increased metabolism. This is brought out by the fact that in spite of
the height of the curves, normal blood sugar levels were reached by
the end of three hours. Basal metabolism was performed on many
cases. As found by Janney, there is no relationship between the height
of metabolism and the height of the blood glucose curve. If the
metabolism were extremely high one would suppose that the glycogen
stores would be exhausted continually and the sugar, which with a
lower metabolism would remain mobilized, is burned up; the result being
a lower curve than is found in milder cases. We found this to be
the case. One of the lowest curves in Figure 3 is from a patient having
a basal rate of -|- 100 per cent. The reverse also is true ; patients
showing the highest curves had basal rates of about + 50 per cent.

The curves in exophthalmic goiter cases clearly indicate the nature
of the dietetic treatment of hyperthyroidism: (1) To protect protein,
a high carbohydrate intake; (2) to avoid hyperglycemia, the feeding
in hyperthyroid cases should consist of many and small meals.

With clinical evidence of lack of thyroid secretion, curves were
obtained which substantiate those already published " (Fig. 4). Janney
did not find the constancy in hypoglycemia after the glucose meal that is
shown in Figure 4. Cases under treatment are not included in this

OLMSTED-G AY— BLOOD SUGAR

393

chart. In some cases the curves very quickly become high under
treatment, while others remain low. It has been shown experimentally
in animals whose thyroids have been removed that there is no delay
in intestinal absorption." If this be true, the low curves in hypo-
thyroidism must be due to increased sugar storage.

Curves of hypothyroidism.

Pituitary Cases. — The posterior lobe of the pituitary gland has been
shown to affect the glycogenic function ;'^ in acromegaly there is a low
"tolerance" while in hypopituitary disease the "'tolerance" is high.^°
Glycosuria following e-xperimental stimulation of the pituitary has been
shown to take place reflexly to the splanchnic area and also after
all known nervous paths have been cut, indicating a true hormone
glycogenolysis. No cases have been observed clinically or metabolically

15. Gushing: The Pituitary Body and Its Disorders, Lippincott, 1911.

16. Weed, Gushing and Jacobson : Bull. Johns Hopkins Hosp. 24:40. 1913.

394 ARCHIVES OF IXTERXAL MEDICIXE

showing high curves, but a large number of cases diagnosed "hypopitui-
tarism," "dyspituitarism" and "polyglandular syndrome" show low
curves. It is in this type of case that the factor of delayed absorption
may play a part. We have observed delayed water absorption in
some of these cases.

Mild Diabetes. — The necessity of pancreatic hormone for glycogen
formation was early demonstrated in perfusion experiments.^* It is
not known whether an increased secretion by the islands of Langerhans
ever occurs, but a decrease in pancreatic hormone has two effects ; loss
of glycogenic power and loss of ability to oxidize glucose. It would,
therefore, be reasonable to suppose that even if there is loss of oxidative
power to a small degree, or, in other words, a very mild diabetes, the
hyperglycemia would be sustained to a more marked degree than any
other condition affecting glycogenesis.^" That such is the case is
shown by the curves in Figure 5. One of the greatest uses for blood
sugar curves is in doubtful cases of mild diabetes. With a normal
fasting blood sugar and a carbohydrate tolerance of from 150 to 200
gm. the curves following a glucose meal are quite distinctive, and differ
from any other curve seen in cases of glycosuria. At the end of three
hours the hyperglycemia is commonly above 0.3 per cent.

Renal Diabetes. — The so-called "renal" diabetic shows glycosuria
with normal fasting blood sugar levels. Two cases have been studied
carefully. Both gave subnormal curves. In one case the threshold
glycemia seemed to be 0.075. Great care must be taken to distinguish
between the emotional glycosuria and this type of glycosuria. The
emotional patient's curve rises to above normal limits the first hour
and may be sustained still longer. The curves of the "renal diabetic"
here observed are quite flat (Fig. 5).

Mental States.- — It is not proposed here to enter into discussion as
to whether disturbances of the higher cerebral centers act on glycogenic
function through reflex action on the chromafin-sympathetic system.
It is simpler to accept Cannon's ^' hypothesis that there is a reflex
stimulation of epinephrin formation in some mental conditions. The
purpose here is to make clear the very considerable influence of various
disturbances of the mental state on blood sugar curves. There is,
however, one condition which would give distinct evidence of the
effect of suprarenal medulla on sugar curves ; namely, Addison's
disease. With hypofunction of the suprarenal medulla and the absence
of other factors influencing them, low curves should be obtained. Two
cases have been followed for several months. The first patient had
tuberculosis of the lungs and gave a normal curve; the second patient,

17. Cannon: Bodily Cliangcs in Pain, Hunger, Fear and Rage. New York,
D. Appleton, 1920.

OLMSTED-G AY— BLOOD SUGAR

395

the pathology of whose suprarenal was unknown, gave a subnormal
curve (Fig. 6). The first case shows the effect of the bacterial intoxi-
cation as well as the deficiency of medullary-adrenal secretion.

Cannon found that pain, rage and fear in animals caused, in a
considerable number of cases, the appearance of glycosuria. The same

Fig. S. — The upper curves are those of mild
Compare with Figures 3 and 8. The lower curv
Compare with Figures 7 and 8.

of diabetes melhtus.
of "renal" diabetes.

Fig. 6.— Curves of two cases of -Addison's disease. The ui)per of the two
is from a patient who had an active tuberculous lesion which would tend to
raise the curve.

has been observed in man, especially in states of excitement and after
severe mental effort. We have studied the blood sugar curves of a
considerable number of cases diagnosed as neurasthenia after a search
was made for organic lesions; cases of hypochondriasis; hysteria;

396 ARCHIVES OF IXTERNAL MEDICIXE

epilepsy, both of organic origin and ordinary type; dementia praecox
and manic depressive insanity/^ and the outstanding fact is that no
prediction can be made as to the nature of the curve from diagnosis
alone. It may be possible for the psychiatrist or neurologist to deter-
mine what the particular mental condition is that stimulates glycogenoly-
sis. Some of the interesting facts are as follows :

Hysterical individuals usually give a normal curve (Fig. 7) in
spite of their intense emotional state. Hypochondriacs and manic
depressive patients (Fig. 8) show, in the majority of cases, high
curves. Neurasthenics and dementia praecox patients may show any
type of curve. The uncertainty of the response in these cases makes the
interpretation of blood sugar curves difficult.

Siiiinuary of Factors Influencing Curies. — To summarize these
factors influencing glycogenesis and glycogenolysis in muscle and liver
the following outline may help.

Glycogenesis ; necessary hormones :

1. Pancreas,

2. Parathyroid ( ?)
Glycogenolysis ; increased by :

1. Increased pn of muscle, liver or blood. Found in such

pathologic conditions as starvation acidosis, nephritic acidosis,
etc.

2. Increased secretion of thyroid hormone.

3. Increased secretion of pars nervosa of the pituitary.

4. Increased activity of sympathetic-chromafin system may
occur with :

(o) Reflex stimulation from cerebral, peripheral or splanch-
nic areas.

{b) Blood bom stimuli, such as infections, malignant, in
pernicious anemia, leukemia, etc.

5. Substances in the blood acting directly on glycogen stores
in the muscles and liver, such as any of 4b.

So far as known the pancreas has the most definite and profound
influence on formation of the glycogen. The evidence for the para-
thyroids lies in the fact that their removal causes glycosuria.^" There
seem to be many more factors stimulating sugar mobilization. Anything
increasing H ion concentration of blood or locally in the tissues .seems
to stimulate glycogenolysis. This has been shown by intravenous
injections of acids and in perfusion experiments.

18. For permission to study these cases we are indebted to Prof. Sidney I.
Schwab.

19. Underbill and Hilditch: \m. T. Phvsiol. 25:66. 1909. Underbill and
Blatherwick: \m. J. Cbcm. 18:87. 1914.

OLMSTED-GAV— BLOOD SUGAR

397

The livers of experimental animals can be almost freed from
glycogen by feeding thyroid.-" Although respiratory quotients show
no decrease in burning power for glucose during feeding,-' it is
undoubtedly the strongest stimulus known to sugar mobilization. The
increased metabolism accompanying thyroid feeding tends to lessen

Fig. 7. — Cases of hysteria, many of whicli show intense emotional excite-
ment. Seventy-five per cent, of uncomplicated cases of hysteria give normal
curves.

Fig. 8.— Cases of manic depressive insanity. Compare these curves with
those of hyperthyroidism.

the glycemia by the rapid burning of sugar. Hyperthyroid blood sugar
curves are high but steep, and show an interesting distinction from the
high curves due to psychic disturbance which are not high but tend to
be sustained. The explanation of increased glycogenesis when thyroid

20. Cramer and Krause :

21. Cramer and McCall :

Quart.
Quari

J. Exper. Physiol. J1:S9, 1917.
J. Exper. Physiol. 12:81, 1918.

398 ARCHIVES OF INTERXAL MEDICIXE

hormone is decreased is only guesswork. If one considers the hormones
of thyroid and suprarenal as opposed or balanced against the pan-
creatic hormone, disturbance of this balance increases or decreases
glycogenesis. The increased glycogen storage coincident with thyroid
deficiency might be taken as evidence that the pancreatic hormone
overacts when not counterbalanced by thyroid.

TABLE S.— Summary of Cases

Diagnosis

Total
No. ol
Cases

Curve Classification

N

II

III

s

Endocrine:

Hyperthyroidism and exophthalmic goiter

Hypothyroidism and myxedema

Dyspituitarism; hypopituitarism: polyghindu-

19
10

15

2

15
16

2
i
9
2
12
4

12
7
2

4

2
2
1
2
2

1
2

1

3

1

1?

1

»

1

1
2

0
2
1
0
3
0

5
0
0

2
0

0
2

1
1
0

?
0

s

0
0
0

1

12
0

1
0

1

0

1

2

1

s

1

3
3
0

J

2
0
0

0
0

1
1

0

2

4
0

0
0

2
3
0

0
0
2.

1
2

3
4
2

0

1
0
0

t

2

0
0
0

0
0

0
0

2?
9

Addison's disease .

Neurologic:

Organic Central Nervous System Lesions:
Encephalitis

Neuritis

Psychoses:

Manic depressive insanity

Focal Injection:

0

0

Keratitis

1

Neoplasms:

Miscellaneous:

Arteriosclerosis

* Diagnoses with only one curve are not included.

In the above outline nervous effects on glycogenolysis are indicated
as reflex through the agency of adrenalin. It must be borne in mind that
this question is still a disputed one. It is also to be remembered that
blood borne stimuli to glycogenolysis may act through adrenalin. These
possibilities are quite hypothetical and are mentioned only as such.
Again it is not known whether the many toxic substances disturbing the
glycogenic balance toward the side of increased sugar mobilization act
directly on the liver and muscles; on the nervous system or on the
suprarenal medulla. All are possibilities.

OLMSTED-GAY—BLOOD SUGAR 399

CONCLUSIONS

1. The basis for the standardization of the technic of the administra-
tion of the glucose meal is pointed out. The necessity for such
standardization is made clear.

2. The discussion of the many known factors which influence blood
glucose curves shows the importance of the consideration of all of them
when such curves form part of any study.

3. The pathologic conditions in which the form of blood glucose
curve is usually (within certain limits) constant, are: (1) hyper-
thyroidism and hypothyroidism; (2) hypopituitarism, and (3) diabetes
mellitus.

4. There are certain conditions which, in general, show increased
curves after the glucose meal. The curves obtamed in such conditions
do not even approximate the fair degree of constancy found in the
above mentioned conditions. Our present knowledge of glycogenic
function in these conditions is rather meager. In this class belong the
effects of infectious toxins; those of cancerous origin; those supposedly
found in pernicious anemia and leukemia ; Hodgkin's disease, etc. Here
also belong conditions of the mental state. "Functional" disturbances,
usually spoken of as neurasthenia, very definitely disturb the height
of blood sugar after glucose meal.

ARCHIVES OF IXTERXAL MEDICIXE

CORRECTION

In the paper by Killian and Kast on "A Study of Significant Chemical
Changes in the Blood Coincident with Malignant Tumors" in the Archives of
Internal Medicine. December, 1921, the statement is made, page 813, that "the
comparative decrease in the amounts of these (nonprotein nitrogen compounds)
has been ascribed to an increased need for nitrogen for the new growth,
whether it be malignant tumor or embryo." This statement follows a refer-
ence to the work of Theis and Stone, so that many would infer that the above
explanation of the low nitrogen figures was offered by these writers. It should
be stated that Theis and Stone offered no such explanation of the figures in
this connection, but that the explanation was advanced by several workers
verbally to us, and we inadvertently included it in our paper without proper
explanation.

John A. Kii.li.^n and Ll-dwig Kast.

BOOK REVIEWS

ACUTE EPIDEMIC E.XCEPHALITIS (LETHARGIC E.XCEPHALITIS).
An Investigation by the Association for Research in Nervous and Mental
Diseases. New York: Paul B. Hoeber, 1921. 258 pages, 36 illustra-
tions. Price, $2.50.

This little book is worthy of attention for two important reasons : First,
on account of its form, as it embodies a novel idea in the organization of cen-
tralized effort by a large body of investigators focused on one disease, and in
the method of presentation of the subject to the reader. The book consists of
the papers read at the first meeting of the Association for Research in Ner-
vous and Mental Diseases held in New York in December, 1920, together with
discussions and the final conclusions of a commission of distinguished neu-
rologists who conducted the meeting and edited the book. At the meeting each
reader of a paper was questioned by the commission to which the papers had
been submitted beforehand. Anyone interested in the method of conducting
scientific meetings and reporting their proceedings will be benefited by glanc-
ing over this volume, even if he is not especially interested in the subject.

Secondly, this is the first book which covers all phases of this new disease
in a readable manner. The history of the disease, etiology, symptomatology,
diagnosis and morbid anatomy are presented by well-known investigators.
There are altogether thirty-si.x contributors but, thanks to the good editorial
work of the commission, the subjects are presented in logical sequence and
without- needless repetition. A bibliography of eighteen pages of the most
important articles is appended. As the literature already has grown to unwieldy
proportions, the appearance of this book is timely, placing all essential data
at the elbow of anyone desirous of authoritative and condensed information.

The commission which edited the book was composed of VV. Timme, Pearce
Bailey. L. F. Barker, C. L. Dana. Ramsav Hunt, Foster Kennedv, G. H. Kirby,
H. T. Patrick, B. Sachs, W. G. Spiller, Israel Strauss, E. W. Taylor, F. Tilney
and T. H. Weisenburg.

DIABETES: A HANDBOOK FOR PHYSICIANS AND THEIR
PATIENTS. Philip Horowitz. M.D. Published bv Paul B. Hoeber,
New York. July, 1920. Pp. 1-186.

For convenience in handling his cases, the author has recognized four
clinical types of diabetes, "mild, moderately severe, severe and juvenile." The
diagnostic symptoms of each type are enumerated so that a given case of
diabetes could readily be classified. The diet the author recommends is given,
and the subsequent additions follow. The book contains very complete analytical
tables showing the composition of foods, and lists of such menus and recipes
as are valuable in handling diabetics. In another chapter questions of hygiene
and exercise are discussed briefly. The book is closed with a description of
tests which are necessary in following intelligently the progress of the patient.
No generalizations of the principles used in the formulation of the diets are
given. Hence, the reader has no way of judging whether a given diet is
adequate in protein or calorics or whether the proper relationship between
fats and carbohydrates has been maintained. It can be seen that such a collec-
tion of diets falls short of giving either the physician or patient an adequate
conception of the fundamentals of diabetic management. Within the last
year studies directed towards the rationalization of diet formulation have
appeared. If these be accepted, then the present hand book may ie regarded
as out of date.

402 ARCHIVES OF INTERNAL MEDICINE

DISEASES OF THE DIGESTIVE ORGANS, WITH SPECIAL REFER-
ENCE TO THEIR DIAGNOSIS AND TREATMENT. By Charles
D. Aaron, Sc.D., M.D., F.A.C.P. Lea & Febiger, 1921.

The third edition of this book is a comprehensive volume devoted to dis-
orders of the entire gastro-intestinal tract — the plan of work following the
physiologic path from the mouth downward. The usual diseases are treated
clearly and fully, with full descriptions of the various laboratory and chemical
procedures and their interpretation, many of which the general practitioner
has never heard of — but with many of which he should be familiar. In gen-
eral, the material is presented in a form resembling that of Osier's "Principles
and Practice of Medicine," with definitions, etiology, pathology, symptoms,
prognosis and treatment under separate headings. Colored plates, engravings
and roentgenograms are used liberally to augment descriptions. Dietetic prin-
ciples in health and disease, the high caloric feeding for typhoid; vitamins —
their importance and distribution, and the functional disturbances of the ner-
vous systern, vagotonia and sympathecotonia. are clearly and fully discussed.
Throughout the book are tables and paragraphs giving a type of information
often desired and difficult to find — the analyses of various springs and waters,
as compared with the more famous European Resorts ; the composition of
many proprietary preparations, both of food and from the various pharmaceu-
tical houses. The book commends itself to both the specialist in diseases of
the digestive tract and to the general practitioner and surgeon.

Archives of Internal Medicine

OBSERVATIONS ON PAROXYSMS OF
TACHYCARDIA *

H. M. MARVIN, M.D.. and PAUL D. WHITE. M.D.

BOSTON

1. The Frequency of Paroxysms of Auricular Fibrillation. — The
widespread use of instruments of precision in the diagnosis of cardiac
arrhythmias during the past few years, and the resulting improvement
in diagnosis, have demonstrated that paroxysmal auricular fibrillation
is by no means a rare clinical condition. That it is encountered in a
certain variable proportion of hospital patients has been pointed out by
a number of observers, most of whom have expressed their belief that
the condition occurred more commonly than was recognized, but the
highest percentage of cases yet recorded is that of Levine,^ who found
that 14.1 per cent, of his group of patients with auricular fibrillation
had shown at some time the transient form. This author includes in
his series four patients who showed auricular fibrillation only during
the transitional stage between auricular flutter and normal rhythm; if
these be excluded, his percentage becomes 10.9, which still remains the
largest published 'figure.

In sharp contrast to this comparatively infrequent occurrence, it is
our experience that paroxysmal auricular fibrillation is found in private
practice with almost the same frequency as the permanent form, and
about as often as paroxysmal auricular tachycardia. In a recent series
of 250 cases with cardiac symptoms or with signs of heart disease
(consultation case of one of us), there were fifteen cases of paroxysmal
auricular tachycardia, seventeen cases of paroxysmal auricular fibrilla-
tion, eighteen cases of permanent auricular fibrillation, and four cases
of paroxysmal flutter. One of the cases of paroxysmal flutter showed
also on occasion the coarse type of paroxysmal fibrillation, and is
included in both groups.^ The total number of patients with auricular

* From the Cardiographic Laboratory. Massachusetts General Hospital.

L Levine. S. A. : Auricular Fibrillation : Some Clinical Considerations, Am.
J. M. Sc. 154:43. 1917.

2. It seems highly probable that paroxysms of flutter and of fibrillation may
occur in the same patient at different times more commonly than reports
would indicate. Theoretically, such variations might be expected, in view of

404 ARCHIVES OF IXTERXAL MEDICIXE

fibrillation in this group is thirty-five, and seventeen of these, or 49
per cent., showed the transient form.^

Eleven of the seventeen patients with paroxysmal auricular fibrilla-
tion were more than 50 years of age (eight were past 60 years). The
age incidence in this series is thus in general accord with the observa-
tions of Levine,^ Heard and Colwell,'' Fox " and Robinson.' Krumb-
haar* reported six cases, the patients aged 53, 38, 40, 18, 33 and 35
years, respectively. Our two youngest patients were aged 22 and 27
years, respectively ; the etiology of their paroxysms is not clear. Both
showed cardiac enlargement on roentgen-ray examination. In contrast
to the age incidence of this group of cases, we find that more than two
thirds of the patients with paroxysmal auricular tachycardia (eleven
out of fifteen) were under 50.

The common types of heart disease in which we have seen the
paroxysmal or transient form of auricular fibrillation are the same as
those in which the permanent form of arrhythmia is commonly seen.
They are most frequently arteriosclerotic (ten of seventeen) ; less often
rheumatic (three of seventeen) and thyroid (two of seventeen). A
recent article by Sniith ^ has emphasized this close correspondence.
Rarely, this paroxysmal type of fibrillation may be found during the
course of acute pericarditis; we have seen three such cases (only one
included in the present group of 250 cases) and Krumbhaar has reported
one case. Severe acute infections (particularly pneumonia), hyper-

the ready transitions from flutter to fibrillation after digitalis, and from fibril-
lation to flutter after quinidin. The recent work of Lewis and his associates."
which has demonstrated the common origin of the two conditions in a circus
movement in the auricle, lends further support to the belief that one patient
may exhibit both mechanisms.
^ 3. Lewis. T. : Observations on Flutter and Fibrillation : Part 2, The Nature
of .'\uricular Flutter. Heart 7:19L 1920. Observations on Flutter and Fibril-
lation : Part 9. The Nature of Auricular Fibrillation as It Occurs in Patients.
Heart 8:193. 1921.

4. It is to be remembered that this figure represents the incidence of tran-
sient auricular fibrillation among patients seen in a consulting practice,
where opportunity has been afforded of seeing them earlier in disease than in
the usual hospital practice. It is not. therefore, to be compared with previous
reports based on hospital records of patients who presented themselves, as a
rule, only when forced to do so by a failing heart. Perhaps in general private
practice the ratio of paroxysmal to permanent auricular fibrillation would be
found even greater.

5. Heard, I. D., and Colwell, A. H. : Transient Auricular Fibrillation. Penn.
M. J. 24:59 (Nov.) 1920.-

6. Fox. G. H. : The Clinical Significance of Transitory Delirium Cordis.
Am. J. M. Sc. 140:815, 1910.

7. Robinson, G. C. : Paroxj-smal Auricular Fibrillation, Arch. Int. Med.
13:298 (Feb.) 1914.

8. Krumbhaar. E. B. : Transient Auricular Fibrillation, Arch. Int. Med.
18:263 (Aug.) 1916.

9. Smith. F. M. : Clinical Observations on Paroxysmal .'\uricular Fibrilla-
tions and Flutter, .Am. J. M. Sc. 162:1.?, 1921.

MARVIX-IVHITE—TACHYCARDIA 405

thyroidism and digitalis in large doses, may be responsible for the onset
of the new mechanism. It seems highly improbable that nervous stimu-
lation alone may be responsible for the clinical condition ; no clear-cut
case is on record, although three of our cases were somewhat suggestive
when first seen.^" In our experience, and in the present series, perma-
nent damage or severe toxicity have been the basis for paroxysms of
auricular fibrillation.

The one patient mentioned above who showed at various times
paroxysms of auricular flutter and of coarse auricular fibrillation has
been under observation for a period of seven years, during which time
she has been followed carefully by clinical, roentgen-ray and electro-
cardiographic observations. One of her electrocardiograms has been
published. ^^ This patient has shown no important symptoms at any
time except during the attacks, she has no murmurs that can be dis-
tinguished, and she has worked as a nurse for six years with only
four days off duty because of her cardiac condition in spite of very
frequent paroxysms. She is now in good health. A cervical rib was
removed from the right side of the neck in 1915 and was followed by
considerable improvement. A teleroentgenogram of the heart in 1915
showed considerable enlargement ; a similar plate taken six years later
showed the same degree of enlargement, in the sairie chambers, notwith-
standing frequent paroxysms of tachycardia in the interval. ^-

2. The Diagnosis of Paroxysmal Tachycardia of Ventricular Origin.
— Within the past few years there has been considerable interest
manifested in the subject of paroxysmal tachycardia due to ectopic
impulses arising in the ventricular tissue. Although a number of
reports of such instances have been published, Robinson and Herr-
mann '* in a recent revie\v were able to find only six undoubted
instances, and six which were probable, in which the pacemaker
responsible for the new rhythm lay in the ventricles. These authors
have called attention to the necessity of obtaining electrocardiograms,

10. Xervous stimulation may be and often is, of course, the exciting factor
in i)ro(h!cing paroxysmal auricular fibrillation in a diseased heart.

11. White. P. D., and Stevens, H. W.: Ventricular Response to Auricular
Premature Beats and to Auricular Flutter (Fig. 5). Arch. Int. Med. 18:712
(Nov.) 1916.

12. Another patient who has shown paroxysmal auricular tachycardia very
frequently between 1914 and 1921 has not been incapacitated at all, and is in
good health at the present time. He also shows cardiac enlargement of
unknown cause. He has 1)een reported as an unusual case of paro.xysmal
auricular tachycardia." an exception to the general rule of an absolutely
abrupt onset and offset of the attack, but nevertheless an undoubted instance
of paroxysmal auricular tachycardia, possibly of nomotopic type.

IX White, P. D. : Clinical Observations on Unusual Mechanisms of the
.\uricular Pacemaker, Arch. Int. Med. 25:420 (April) 1920.

14. Robinson, G. C. and Herrmann. G. R.: Paroxysmal Tachycardia of Ven-
tricular Origin and Its Relation to Coronary Occlusion, Heart 8:59. 1921.

406 ARCHIVES OF IXTERXAL MEDICIXE

and have included in their Hst only those cases from which such curves
have been published. In order that the diagnosis may be established
beyond question, they direct attention to the following requirements :

(1) The electrocardiogram must give definite indications that the cardiac
impulses producing the high ventricular rate are arising in the ventricles, and
this can be shown most clearly when a succession of auricular impulses can
be made out, occurring independently of, and at a slower rate than, the com-
plexes of ventricular origin. (2) The ventricular complexes must be abnor-
mal in form. (3) The presence of isolated ectopic ventricular beats between
paroxysms is in favor of the tachycardia being of ventricular origin, especially
if the ectopic beats and those composing the paroxysms are of similar form.

To the six cases previously reported, Robinson and Herrmann have
added four, bringing the total number of undoubted cases to ten. It
has seemed, to us worth while to report a further instance of this com-
paratively rare type of tachycardia, partly to increase the number of
recorded cases, but chiefly to call attention to the possibility of error
in making the diagnosis, even with electrocardiograms, unless control
records have been obtained from the patient between paroxysms.
Under a certain condition, which we shall illustrate, it is possible to
obtain records of auricular paroxysms which resemble closely those of
ventricular origin.

The patient from whom the electrocardiograms shown in Figures
1, 2, 3 and 4 were obtained was a young woman, aged 21, whose past
history was uneventful, except for the occurrence of diphtheria and
pneumonia in early childhood and occasional mild attacks of tonsillitis
during several successive winters. There had been no symptoms of
cardiac disease whatever preceding the onset of the paroxysms of
tachycardia; the patient had been attending business school for two
years and had taken daily walks of several miles for exercise. Her
first paroxysm occurred six weeks prior to her entry to the hospital,
as she was stooping to pick something from the floor. It lasted for
approximately two minutes, during which time she felt dizzy, her head
seemed "hot and swollen," and she was conscious of the tachycardia.
The attack ended as suddenly as it had begun, but a second one occurred
two hours later and lasted for approximately fifteen minutes. During
this second attack her symptoms were as before, but in addition there
was marked shortness of breath, which was a feature of all subsequent
attacks. After the first day there was a period of freedom for about
two months, when the attacks recommenced, and it was during this sec-
ond period of paroxysms that the following records were obtained.

Figure 1 is an electrocardiogram obtained April 7. Leads I, II and
III are shown, and the only abnormality in the curves is a single
premature beat arising in the ventricle, which is recorded in Lead II.
April 21, we were fortunate in securing the onset of one of her numer-

MARriX-irHlTE—TACHYCARDlA 407

ous paroxysms (Fig. 2). The plate shows three strips of Lead II,
and satisfies in itself all the requirements mentioned by Robinson and
Herrmann for the diagnosis. Isolated ectopic beats of ventricular origin
are seen in A and B, and these are similar in form to the beats which
compose the succeeding paroxysm. The ventricular complexes are high'y
abnormal in form. Finally, auricular complexes occur in B after every
second ventricular beat, the rate of the ventricles being 262.5 per minute
and that of the auricles one half this rate. Similar curves, in which the
auricles beat at exactly half the rate of the ventricles, P waves appear-
ing at corresponding points on every second ventricular complex, have

^-^=^-=^^^^-M-^-4-4-4^^^ ^4-

^ in niiriiiniii^i>^^iM^MWi iit i rlil.hnj i_.jiijri

^ — - : : ^-%=|=H41 1 riT

j i . . . . ■ . ; . . .- 1 i - : , j : i-

=^;i :; it; ::::::::;; i i ; 1:

z^ : : - : ; ^; : : : ; i; : ■ '\ ■ ' ; ■: '- ':-

; -: ^ i: ill': : Li ^ t ^ ; - L i ^ i li L r

-^ ':': S L ':': L ':*:': L ':':': '^ S ':':' t '-

ii M : j ; ^ ; £ : • ^ j ^ ' : 1- ; i !

^^^^^^^^H

Fig. 1.— Electrocardiogram of M. R. Leads I. II and III. .N'.rmal com-
plexes throughout except for one ventricular premature beat in Lead II. (In
this and all succeeding electrocardiograms, distances between ordinates repre-
sent 0.2 second, distances between abscissae 10'' volts.)

been published by Hart,'' and more recently by Robinson and Herr-
mann " (their Figs. 9, 11 and 13). There is a close resemblance between
Hart's Figure 4 and our Figure 2, except that the R waves in our
illustration show notching at the tip and the P waves are more promi-
nent than in Hart's ca.se. It is to be noted that during the brief part of
the paroxysm shown in B the rate is not absolutely regular, nor are the
complexes precisely alike. The time intervals, as measured from peak

15. Hart, T. S.: Paroxy?

Tachyc

ARCHIVES OF INTERXAL MEDICJXE

MARVIS-UHITE-TACHYCARDIA 409

to peak of the R waves in seconds are as follows: 0.257, 0.198, 0.212,
0.237, 0.243, 0.243, 0.243, 0.253, 0.245, 0.248. The variation in rate,
as shown by these intervals, is largely confined to those beats which
initiate the new rhythm; from the fourth beat onward the greatest
variation from cycle to cycle is 0.010 second. Thus the rate for the
entire stretch of the paroxysm is 262.5, but the rate for the last eight
beats is 250. The difference in form of the complexes is apparently
of a progressive nature ; the second and third beats, for instance, are
practically free from notching or slurring of the ventricular complex,
the succeeding five beats show very definite slurring of the upstroke
of R and notching of the beginning of the downstroke, while the last

Fig. 3. — Electrocardiogram of M. R
paroxysm of tachycardia of ventricular >
I and III, and 225 in Lead II.

Leads I. II and III. taken during a
igin, in wiiicli tlie rate is 209 in Leads

three recorded complexes apparently foreshadow the final form which
is to be assumed, in which a heavy slurring of the upstroke and a
slighter degree of slurring of the downstroke of R appear as the promi-
nent features.

The lower part of the figure, marked C, was taken within a few
seconds after B and shows the form of ventricular complex which
characterized the second lead in all subsequent electrocardiograms. It
will be noted hat the P waves, which appeared so prominently after
every second ventricular complex in B, have now almost disappeared.

ARCHIVES OF INTERNAL MEDICINE

MARllX-llHI TE—TA CH ) 'CA RDIA

The last feature of the ventricular paroxysms to which attention
should be directed is shown in Figures 2 and 3, and consists of the
change of rate between the early and later parts of the paroxysm.
Thus in Figure 2 the rate in B is 250 or 262.5, according as we include
or exclude the first three beats, while in C, taken a few seconds later,
the rate is only 220. Similarly, in Figure 3, the rate in Lead II is 225
per minute, while in Leads I and III the rate is from 209 to 211. These
two figures are from different paroxysms on the same day.

- — s

--^E

^

fr

W^Jim

II

\$m

mm

ill

§w^

': 1-1

; i ik

f^J

I A ; A__jyi

tt

^

l^,^«

^Mhw ^

JKp

tRtt

M

roi>

T^i^

^tt"nT'n'r

i^ i '-

: *r==-=^

^=^1

Fig. 5.— Electrocardiogram of L. K. Leads I. II and III. Taken during a
paroxysm of tachycardia, in which the rate was 190 per minute. Note the
general resemblance between Leads I and II of this figure and Leads II and I,
respectively, of Figure 3.

The same patient showed also unusual paroxysms of auricular
origin. A plate taken just after that reproduced in Figure 2 and just
before that in Figure 3 recorded two such paroxysms, only one of which
is here shown (Fig. 4). This curve is from Lead II. and shows in its
early part complexes of normal form, with a rate of 7? per minute.

412 ARCHIfES OF IXTERXAL MEDICIXE

In the middle of the curve are seen two exactly similar premature beats
arising in the ventricle. Immediately following the second of these
premature beats there occurs a short paroxysm of tachycardia consisting
of five beats, in which the normal sequence of chamber contraction is

Fig. 6. — Electrocardiogr
lowly. Same paroxysm a

: L. K. Lead I. taken with
shown in preceding figure.

late traveling

Fig. 7.— Electrocardiogram (Leads I, II and III) of L. K.. showing normal
rhythm, the ventricular complexes of which arc typical of those associated with
defective conduction along the right hranch of the .\-\' bundle. Rate. 100 per
minute. Wntricular complexes are similar to those in Figure 5.

maintained.'" The rate is 125 per minute. After the conclusion of
the i)aroxysm, the rate falls at once to its original level.

This case, then, is one in which there have been recorded at difTercnt
times isolated ectopic beats of ventricular origin, paroxysms of tachy-
cardia of ventricular origin, and paroxysms of tachycardia in which

16. The P wave of the paroxysm is slightly hij.luT than the

MARVIX-UHITE— TACHYCARDIA 413

the pacemaker lay in the auricle's. In at least one instance, the rate for
the first few beats of the ventricular paroxysms was higher than the
rate subsequently maintained. In the only instance where the onset
of the new rhythm was recorded, there was a slight, progressive change
in the form of the first eight ventricular complexes.

By way of contrast to the above example of paroxysms of tachy-
cardia arising in the ventricle, is an electrocardiogram (Fig. 5) from
another case which at first sight appears also to be of ventricular origin.
Leads I and II of this figure, for example, are quite similar in form
to Leads II and I, respectively, of Figure 3. (Corresponding leads
cannot be compared because of obvious gross differences.) !■ The
general outline of each complex is that of an ectopic beat arising in
the ventricle. This appearance is even more striking in Figure 6, which
is from Lead I of the same patient, taken with the plate traveling at a
slower rate.

That this paroxysm does not owe its origin to impulses arising in
the ventricle, however, is made apparent at once by inspection of

Fig. 8. — Electrocardiogram of L. K. Lead 1. The end of a paroxysm is
recorded. Note that the first, second and fourth ventricular Ijeats after the
end of the tachycardia are in response to impulses from the normal supra-
ventricular focus, yet are similar in all details to the heats composing the
paro.xysm.

Figure 7, which was obtained about five minutes after Figure 6, and
shows the normal rhythm. The three leads are given, and a comparison
of this curve with that in Figure 5 shows that the ventricular complexes
in both are similar in all three leads and we know them to be of supra-
ventricular origin because in Figure 7 each ventricular beat is preceded
by an auricular wave. This electrocardiogram is of the type associated
with defective conduction along the right branch of the A-V bundle.

Further proof of the auricular origin of the paroxysm is obtained
from Figure 8, in which the termination of a paroxysm is shown. The
first, second and fourth beats after the end of the tachycardia arc
preceded by waves due to auricular activity, and are clearly to be

17. .\.s a matter of fact these very differences would help to difTere
the two conditions. Figure 5 being much more likely an example of iiitr
tricular block.

414 ARCHIVES OF IXTERXAL MEDICIXE

considered as ventricular responses to impulses from the auricles, yet
these ventricular responses are in all respects similar to the beats
composing the paroxysm.

It would be impossible to assert with confidence that the paroxysmal
tachycardia shown in Figure 5 is of auricular origin without the further
knowledge obtained from other records. From the standpoint of
accurate diagnosis and a more complete understanding of the electrical
events in the heart associated with tachycardias of paroxysmal nature,
it is important to make this diflferentiation.

Fig. 9.— Electrocardiogram of E. J. H. Leads 1, II and III. Paroxysmal
auricular tacli.vcardia. .-Xii ectopic l.cat arising ni tlie ventricle is shown in
Lead III. Xote that it does not disturl) the dominant rliythni.

Lewis '** has called attention to the possibility of sudden interference
with conduction in one branch of the A-V bundle during the period of
rapid heart action, and has published a record illustrating aberration only
during the paroxysm. Such curves, of course, are similar to those
obtained from patients with pre-existing bundle branch block, for the
mechanism of their production is exactly alike. Lewis concludes that
"in the human subject, jiaroxysms presenting anomalous ventricular

18. Lewis, T. : Mechanism and Graphic Registration of the Heart Beat.
Xcw York. R. Hoebcr, 1921. p. 259.

MARnX-WHITE— TACHYCARDIA

415

complexes may be produced in one of two ways ; these paroxysms either
arise in the ventricle itself, or, arising in the auricle, the excitation
wave pursues an abnormal ventricular course." This abnormal ven-
tricular course may be due to transient interference with conduction,
as in his case, or to permanent interference, as in the case reported
above. Reference has already been made " to an electrocardiogram
published by one of us in 1916, which shows the sudden development
of aberration of the ventricular complexes during a paroxysm of auric-
ular flutter.

3. The Occurrence of Ectopic Ventricular Beats in Auricular
Paroxysmal Tachycardia. — The interruption of paroxysms of auricular
tachycardia by ectopic beats arising in the ventricle is of rare occur-
rence ; we have been able to find but one published electrocardiogram
illustrating such an event. ^'' \\'e have recently seen a case exhibiting
ectopic beat* with such frequency as to make it one of considerable
interest.

4-i-M 1-'

_ J _ 1 _j

F==

g

. — ii— — ^~-^

"i

^=s

^

W

i

p

i

^yjfc

i

f^

fw

1

i

1

Fig. 10. — Electrocardiogram of J. D. S. Paroxy.smal auricular tachycardi;
Rate 200. Alternation in size of the QRS complexes is shown.

The patient from whom the curve shown in Figure 9 was obtained
was a laborer, aged 57 years, who had been subject to attacks of
tachycardia for about three years. The paroxysms occurred usually
three or four times a day, for from two minutes to several hours. His
only symptom during the attacks was slight palpitation, which had
never been of sufficient severity to cause him to stop his work for even
a few minutes. An electrocardiogram taken between attacks showed
ectopic ventricular beats similar to those which appeared during the
attacks, and indicated also partial A-\' heart block (the P-R interval
measuring 0.233 seconds) and intraventricular block.

Figure 9 was taken during an attack in which the heart rate was
166.6 per minute. It shows one ectopic ventricular beat. Several such
beats were recorded during the paroxysm. It is to be noted that the
ectopic beat does not disturb the rhythm ; it occurs almost at the precise

19. Agassiz, C D. S. : Paroxysmal Tachyeardia .Accompanied by the Ven-
tricular Form of Venous Pulse, Heart 3:193, 1912.

416 ARCHirES OF IXTERXAL MEDICI XE

instant when a beat was to be expected, and the beat which follows falls
at its proper point. In other words, the same features mark a ven-
tricular premature beat in paroxysmal auricular tachycardia as in the
normal rhythm arising in the sino-auricular node ; in both instances, the
distance between the complexes embracing the ectopic beat measures
the same as that between any two rhythmic beats.

This case is recofded because of the rarity of such published curves
and because it is of some importance to recognize that the finding of

Fig. 11. — Electrocardiogrnm of J. D. S. Normal rhythm from same patient
wliose previous record is shown in Figure 10.

ventricular premature beats in a case which presents the features of
auricular paroxysmal tachycardia does not militate against that diag-
nosis. Among the records of more than thirty cases of auricular
paroxysmal tachycardia in the files of our laboratory, this is the only
instance in which ventricular ectopic beats have been found in the
midst of a paroxysm.

4. Altcr)iatioi! of the R W'a^'cs in Paroxysmal Tachycardia. — Alter-
nation in the amplitude of the radial pulse in cases of i)aroxysmal

MARriX-lVHITE—TACHVCARDIA 417

tachycardia occurs with such frequency that it is no longer regarded
as unusual or important. Alternation in the height of the R waves in
this condition, however, is not common, in so far as one may judge by
published records. There have been many curves showing variation
in amplitude without alternation, although the vast majority of
cases exhibit comple.xes which are precisely alike in size and shape
over long periods. Lewis has placed on record an excellent example
of simultaneous alternation of the R waves of the electrocardiogram
and the radial pulse, in which the large ventricular complexes corre-
spond to the small pulse waves." Harf " has recorded a striking instance
of this condition, his curves being obtained from a case of paroxysmal
tachycardia of ventricular origin.-"

The electrocardiogram shown in Figure 10 represents Lead IL
The rate is 200 per minute. The patient from whom it was obtained
was subject to paroxysms of the usual description, and the only feature
of interest in the curve, other than the alternation, is its close resem-
blance to auricular flutter. A tracing taken after the paroxysm showed
curves which were normal in all respects (Fig. 11 showing Leads L II
and III).

SUMM.iVRY

1. Paroxysmal auricular fibrillation is a common type of paroxysms
of tachycardia, and is seen in practice as frequently as paroxysmal
auricular tachycardia and permanent auricular fibrillation.

2. Paroxysmal auricular fibrillation is found most frequently in old
age, the result of cardiosclerosis. It is also found in rheumatic and
thyroid hearts, in acute pericarditis, severe acute infections, and follow-
ing digitalis.-'

3. Paroxysms of tachycardia may occur at frequent intervals for
years without incapacitating the subject and without increasing the
degree of cardiac damage appreciably.

4. Paroxysmal tachycardia of ventricular origin is very rare
Another case is added to the ten undoubted cases already reported.

5. Electrocardiographic study is essential in the accurate diagnosis
of ventricular paroxysmal tachycardia, and even with electrocardio-
grams the condition must be differentiated from auricular paroxysmal
tachycardia with bundle branch block.

6. \'entricular ectopic beats may occur in auricular paroxysmal
tachycardia .without disturbing the dominant rhythm.

7. Rarely, alternation of the Q R S complexes of the electrocardio-
gram may be found in paroxysmal tachycardia. (Alternation of the
radial pulse in this condition is common.)

20. Lewis, T.: Mechanism of the Heart Beat, 1911. p. 274.

21. It also has been reported after other poisons.

RENAL GLYCOSURIA^

D. S. LEWIS. M.D.

MONTREAL, CANADA

There are four cardinal points in the diagnosis of this interesting
anomaly: (a) a glycosuria without hyperglycemia; (h) little, if any,
relationship between the carbohydrate intake and the amount of glucose
excreted in the urine; (c) the absence of the signs and symptoms
characteristic of diabetes mellitus, and (d) a long period of observation
during which the patient shows no tendency to develop diabetes mellitus;
Joslin 1 lays particular stress on the last criterion, which is the most
difficult to carry out. A critical review of the literature is made by
Goto,- Bailey,^ Strouse * and Lewis and Mosenthal.^ When preparing
data connected with the report in 1915, Lewis and Mosenthal found less
than ten cases which were described in sufficient detail to warrant their
acceptance as instances of true renal glycosuria, but since that time
at least nine other cases '^ have been noted. With the more careful
observations of the blood sugar this depression of tlie "leak point" for
glucose is being recognized with greater frequency.

In this paper a further note is recorded on the case reported in
1915^ and studies of two other instances observed in the metabolism
clinic of the Royal Victoria Hospital are presented in some detail. The
clinical findings in the first patient may now be said to fulfill all four
requirements as he has been observed for a period of six years ; the
second and third cases which have been observed for twelve and
fifteen months, respectively, can be regarded as answering the first
three tests, but a final decision will not be possible without a further
period of observation.

Methods Employed. — Sugar identified as glucose, by fermentation;
osazone crystals; and synchronous determinations of the amount of
sugar by polariscopic and copper reduction methods. Urinary sugar:
Benedict's standard quantitative methods. Blood sugar: Lewis-
Benedict method, unless otherwise stated in the text.

* From the Metabolism Clinic of the Royal Victoria Hospital.

1. Joslin, E.: Treatment of Diabetes Mellitus. Philadelphia. 1917. p. 64.

2. Goto. K.: Alimentary Renal Glycosuria, Arch. Int. Med. 22:96 (July) 1918.

3. Bailey. C. V.: Renal Diabetes. Am. J. M. Sc. 157:221, 1919.

4. Strouse, S.: Renal Glycosuria, Arch. Int. Med. 26:768 (Dec.) 1920.

5. Lewis. D. S., and Mosenthal, H. O. : Renal Diabetes, Bull. Johns Hop-
kins Hosp. 27:133, 1916.

6. Beard. H.. and Grave. F. : Renal glycosuria. Arch. Int. Med. 21:705
(Tune) 1918. Allen. F. M.: Wishart. M. B.. and Smith, L. M.: Three Cases
of "Renal Glycosuria." Arch. Int. Med. 24:523 (Nov.) 1919. Paullin. J. E. :
Renal Glycosuria. T. A. M. A. 75:214 fjulv 24) 1920. Marsh, P.: Renal
Glycosuria, Arch. Int. Mod. 28:54 fjulv) 1921.

LEWIS— REX AL GLYCOSURIA 419

REPORT OF CASES
Case 1. — W. P. W." (medical No. 34774), was first studied in September,
1915. On admission to the Johns Hopkins Hospital his urine contained 2.06
per cent, sugar, and the daily output averaged 25 gm., with a blood sugar
ranging from 0.08 to 0.11 per cent. He gave a normal curve following the
ingestion of 100 gm. glucose. Further studies by Mosenthal, in 1916. con-
firmed the findings. In August. 1921, the patient reported that he had con-
tinued in excellent health since his discharge from the hospital. He had
gained 12 pounds in weight: he had survived a severe "flu" infection, and
was taking a full and unrestricted diet. His urine still contained about 2 per
cent, sugar, but he had no thirst, polyuria, or any other symptoms of diabetes
mellitus.

The following data of recent studies on this individual liave been
supplied b)' Dr. F. M. Hanes of Winston-Salem. N. C, and prove
that his condition has shown no essential change. The diagnosis made
in 1915 has been verified by the subsequent course of the case.

The functional condition of the kidney has attracted considerable
attention in renal glycosuria. Klemperer ' in his first description of
what he termed "renal diabetes," stated that the sugar always dis-
appeared from the urine with the onset of a nephritis. On the other

T.-\BLE 1. — Response of Blood Sugar to 132 Gm. Glucose

Junes. 1921 Blood Sugar, per Cent. Remarks

Fasting 0.09 After 24 hours fasting

First hour 0.15

Second hour 0.12

Third hour 0.10

The urine passed during the twelve hours preceding the test contained 3.5 per cent.
sugar (Benedict).

hand, Liithje,^ Tachau " and Naunyn " suggested a direct causal
relationship between the nephritis and the glycosuria, and the first
two gave e.xamples of proved renal glycosuria in which the onset
seemed to be associated with the appearance of a nephritis. Frank "
has described a glycosuria following the toxic nephritis produced by
mercury, uranium, chromium and cantharadin, but there has been little
evidence of severe kidney disease in a majority of the reported cases.
Bailey ' described one case of severe parenchymatous nephritis, with
a renal glycosuria. The blood sugar was comparatively high, and after
75 gm. glucose it rose to 0.4 per cent., returning to the fasting level

7. Klemperer, G.: Ueber regulatorische (ilvkosurie und renalen Diabetes,
Berl. klin. Wchnschr. 33:571. 1896.

ft Liithje, H. : Beitrag zur Frage des renalen Diabetes, Miinchen. med.
Wchnschr. 38:1471, 1901.

9. Tachau. H. : Beitrag zum Studium des Nierendiabetcs, Deutsch. Arch. f.
klin. Med. 104:448. 1911.

10. N'aunyn, B. : Der Diabetes Mellitus. Wien., 1906, p. 136.

11. Frank. E. : Ueber experimentelle u. klinische Glykosuricn renalen
Ursprungs. Arch. f. exper. Path. u. Pharniakol. 72:387. 1913.

420 ARCHUES OF IXTERXAL MEDICI. \E

six hours after the meal. This, he states, is a type of curve often seen
in nondiabetic cases of nephritis. In this patient the sugar output was
remarkably constant, and there was no sign of a true diabetes.
Mosenthal and Lewis ^- also report a case of arteriosclerosis and primary
contracted kidney, in which a glycosuria appeared while under obser-
vation. This patient excreted 22 per cent, of phthalein in two hours,
and his blood urea was 0.749 gm. per liter. With a fasting blood sugar
of 0.10 per cent., the urine contained 0.13 per cent, sugar, and after
the ingestion of 100 gm. glucose the blood sugar rose to 0.26 per cent,
in 90 minutes, and returned to the fasting level at the end of three hours.
The following cases are examples of this type of renal glycosuria.

Case 2.— May 5, 1920, T. C, (metabolism Xo. 50). Chinese boy, aged 18.

History. — About four weeks before admission he had a chill and fever.
He passed very little urine. His legs began to swell, and he had to stop work.
The swelling gradually spread to his trunk, arms and face. Xo further
history was obtainable. He was admitted to hospital May 9 and transferred
to the Metabolism Clinic May 21.

Physical £.ramma/iO)i.— Temperature, 98; pulse, 70; respiration, 20 (on
admission). Patient was an adult Chinese of about stated age. He lay com-
fortably in bed. There was marked general anasarca : pupils equal and active ;
teeth in fairly good condition; tonsils not enlarged; tongue coated; no gen-
eral glandular enlargement; thyroid not enlarged: no signs of hyperthyroidism.
The chest wall was edematous. There was a bilateral hydrothorax and many
moist rales were heard over both lungs. The heart was regular in rhythm,
and extended 7 cm. to the left of the midline. The sounds were well heard.
There were no murmurs or accentuations. The vessel walls were not thick-
ened. Blood pressure: 110/65. The abdomen was tense: the walls wxre
edematous : there was a marked ascites ; the liver and spleen were not palpable.
The genitalia were much swollen, there was no urethral discharge. The reflexes
were active. The eyegrounds were normal.

Uriuc: .■'icid; cloudy: specific gravity. 1.026: albumin. 20 gm. per liter; no
sugar. Microscopic Examination ; Granular, hyalin and fatty casts. White
blood cells and an occasional red blood cell.

Tests of Kidney Function.— Blood urea. 0.583 gm. per liter; plasma chlorids,
5.65 gm. per liter; phthalein tests, 8 per cent, in two hours.

Blood: Wassermann test negative. Red blood cells, 5,910.000: white .blood
cells, 5,600; hemoglobin, 85 per cent. (Talquist).

Diagnosis. — Chronic diffuse nephritis ; general anasarca.

Diary. — In view of the severity of the nephritis the patient was
kept in bed on a salt-poor diet, with an average daily carbohydrate
intake of 125 gm. His condition gradually improved, and he lost 16
kilos in weight, but with ;in imrcascd diet hl^ eilema returned and
his general condition became much worse. At this time an estimation
of the blood proteins showed a gloiuiHn-albuniin ratio of 3.^)1 :l.vW,
which is a complete reversal of the usual proportions. In view of
Epstein's '" reports, the diet was changed September 10, to one with

12. Mosenthal, H. O.. and Lewis. D. S.: The D:X Ratio in Diabetes Mel-
litus, Bull. Johns Hopkins Hosp. 28:187. 1917.

13. Ep,stein, A. A.: Oedema in Chronic Xephritis. .\m. J. M. Sc. 154:638
(Nov.) 1917.

LEltlS—REXAL GLYCOSURIA 421

low fat and relatively high protein values. There was little change in
the water balance, but the total nonprotein nitrogen of the blood
mounted from 40 mg. per hundred c.c. to 89 mg. per hundred c.c. in
six days, and the diet was discontinued soon afterward. Suddenly,
September 22, a glycosuria appeared. The first day the output was
6.99 gm. ; the ne.xt day it was 7.39 gm., and for the succeeding eight
months sugar was absent in only five twenty-four-hour specimens, and
in individual specimens of si.x other days. During this entire period
the output has never exceeded 16.8 gm., and it has been between 5 and
10 gm. 151 times in a total of 230 determinations (Table 2), while
the concentration in the urine has been from 0.5 to 0.75 per cent, in two
of ever)' three examinations. In other words, the output and concen-
tration have shown a high degree of constancy.

T.^BLE 2. — Range of Coxcentr.\tion -\nd Tot.^l Excretion
OF SuG.^R IX Urine

Concentration in
Urine, per Cent.

Xumbcr of
Analyses

Per Cent, of
Total Xo.

Output in Gm.
per Day

Xumber ol
Analyses

Per Cent, of
Total No.

Above 0.75

38
151
41

16.5
6,1.7
17.8

Above 10.0

lO.O - 5.0

Below 5.0

38
151
41

16.5

i-is

TABLE 3. — Independence of Carbohydrate Intake and Output

Carbohydrate

Per Cent.

Gm.

174

0.86

6.45

•:o9

0.51

3.82

144

0.52

3.74

87

0.60

2.70

»4

204

0.40

6.68

204

D. 0.28

1.12

Again, the total output of glucose seems to be independent of the
carbohydrate intake, the higher rates of excretion often being associated
with the lower diets and vice versa. Table 3 will serve to illustrate
the independence of intake and output.

Similarly, his excretion varies from 0 to 8.16 gm. during the period
from November 16 to December 16 with an unchanged diet which
contained protein, 50 gm. ; fat, 57 gin., and carbohydrate, 234 gm.
Throughout the period of observation the changes in the output are
featured by their sudden onset and disappearance. In Table 3 the
output is 6.68 gm. one day, 1.12 gm. the next day, then sugar free for
twelve hours, and on the following day is -reestabli.shed at its old level
of 5.71 gm. On the other hand, during any particular day the rate of

422 ARCHIVES OF IXTERXAL MEDICIXE

excretion from hour to hour seems to be very constant. Two hourly
collections are presented in Table 4, and show a maximum variation
in concentration of only 0.16 per cent., and an hourly output from
0.50 to 0.36 gm.

In September, 1921, he was studied again, and at this time he was
fasted for three days before becoming sugar free. On a gradually
increasing carbohydrate diet traces of sugar reappeared with 30 gm.
carbohydrate given as green vegetables, and a measurable quantity
(0.18 per cent.), with a diet containing 40 gm. of carbohydrate in the
form of potato.'

A glucose curve was also carried through with the following results,
which are quite typical of those seen in nondiabetic cases of nephritis.
There is a relativ;ely slow rise to the maximum (0.241 per cent.), at one

TABLE 4. — R.\TE of Sugar E.xcretio.v from Hour to Hour
March 7, 1921. Diet: Protein, TO gm.; fat. SI gm.: carbohydrate, 2S0 gm.

Time Volume in C.c. Spec. Gr. Percent. Gm. Gm. per Hour

Sa.m.-lOa.m 116 1.022 0.71 0.82 0.41

10 a.m. -12 m 138 1.020 0.63 0.87 0.43

12m. - 2p.m 152 1.021 0.66 1.00 0.50

2p.m.- 4p.m 142 1.022 0.69 0.98 0.49

4p. m.- 6p. m 170 1.020 0.56 0.95 0.47

6 p.m.- 8p. m 140 1.021 0.67 0 94 0.47

8 p.m.- 8 a. m 775 1.019 0.55 4.26 0 36

Total 1.633 9.82

T.AlBLE

5.— Response

OF

Blood

AND UbIXI

E TO 100 Gm. Glucose

Urine

Blood
Sugar-

Time

Vol.. C.

C.

Sp. Gr.

Per Cent.

Gm.

Gm. per Hr. Remarks

95

0.35

100 gm. glucose

in 300 c.c. of

10:15 a. m.

0.208

lemonade at

10:30 a.m.

80

1.020

0.60

0.48

0.32 9:30 a. m.

11:00 a.m.

0.241

11:30 a. m.

0.151

88

1.016

0.91

080

0.80

0.096

167

1.011

0.32

0.54

0.54

2:30 p.m.

0.087

202

1.013

0.20

0.40

• Folin. 0., a

md Wu.

H."

and a half hours and the normal level is reached again in three hours.
The specimen of urine passed before the ingestion of glucose showed
that sugar passed through the kidney with a blood concentration in the
vicinity of 0.092 per cent. A previous determination had .shown a
glycosuria of 0.11 per cent, with a fasting blood sugar of 0.068 per
cent. ; therefore, no question can be raised as to the extreme depression
of the renal threshold.

In the course of ten months during which the glycosuria has been
observed in hospital, there have been periods during which the patient

14. Folin. O.. and Wu. H.: A System of Blood Analysis. Suppl. 1. I. hi
Chcm. 41:367 (March) 1920.

LEWIS— REXAL GLYCOSURIA 423

has shown large changes in his body weight. A study of the fluid
exchange and the rate of sugar excretion has shown no constant
relation between the total fluid excreted and the amount of sugar in
the urine. During one period there was a considerable fall in the sugar
output tvith a retention of water, while during a subsequent period of
diuresis the sugar showed no corresponding increase. This independence
of sugar and fluid output has been noted in a majority of the published
cases.

Case 3. — C. M., (metabolism No. 56), aged 74; bookkeeper.

History. — This man was admitted on account of a severe attack of scurvy
occasioned by a deficient diet, the result of his financial straits. The sugar
was found during the routine analysis of the urine. Subsequent questioning
failed to reveal any of the usual symptoms or signs of diabetes mellitus. There
was no thirst; no craving for sweets: no polyuria; no loss of weight. Nothing
suggesting hyperthyroidism, and beyond a mild eczema of the hands, he had
been very healthy. The family history was negative as regards metabolic
disorders.

Physical Examination. — Temperature, 98 F. ; pulse, 80 ; respiration, 20 ( on
admission). A well nourished man ; who appears much younger than stated age.
There is fluid in both knee joints, and the right olecranon bursa is filled with
blood-stained fluid. There are intramuscular hemorrhages in the thighs and
calves of both legs. The gums are much cut up where several teeth have
been extracted recently on account of "pyorrhea" and are quite spongy around
the remaining teeth. The pupils react to light and accommodation. The chest
shows a moderate emphysema and bronchitis. The heart is slightly enlarged,
there is an extrasystolic arrhythmia, the sounds are well heard, and there is
a soft apical systolic murmur, poorly transmitted to the axilla. The vessel
walls are definitely thickened. Blood Pressure: 160/84. The abdomen is
negative, the genitalia are negative.

L/rine: .-Xcid ; specific gravity, 1.021; albumin, faint trace; sugar, 2.5 per
cent, .'\cetone and diacetic acid are absent. Microscopic examination shows a
few leukocytes and occasional hyalin and granular casts.

• Tests of Kidney Function. — Blood urea. 0.435 gm. per liter : uric acid, 5 mg.
per hundred c.c. ; creatinin, 1.64 mg. per hundred c.c. ; plasma chlorid, 6.19 gm.
per liter. The nephritic test diet gives a normal curve, and the phthalein excre-
tion is 64 per cent, in two hours.

Blood: Erythrocytes, 4,900,000; leukocytes, 7,200; hemoglobin, 90 per cent.
(Talquist). Wassermann test negative.

Diagnosis. — (1) Scurvy, (2) arteriosclerosis and arteriosclerotic kidney, (3)
glycosuria.

Diary. — The scurvy cleared up rapidly with the ordinary anti-
scorbutic foods, and the glycosuria was then studied. On ordinary diets
with no limitation of carbohydrate he excreted from 33.2 to 53 gm.
sugar daily, with a fasting blood sugar of 0.115 per cent., and a
digestion sugar'' of 0.133 per cent. Four days of a low calory diet
(protein, 43 gm. ; fat, 46 gm., carbohydrate, 29 gm.), followed by
starvation for two days and three days of protein 40 gm., only sufficed
to reduce the daily sugar excretion to 15.6 gm. The attempt to render
the urine sugar free was then abandoned. On subsequent days, with a

IS. Blood taken one and one-half hours after food.

424 ARCHIVES OF IXTERXAL MEDICINE

gradually increasing diet, his output ranged from 20 to 57.5 gm. and the
blood sugar from 0.112 to 0.129 per cent, when fasting, and from 0.121
to 0.161 per cent., one and half hours after food. In Table 6 a summary
is presented of the diets, sugar excretion and bloud sugars, fasting and
digestion, during his first admission.

This table shows the independence of carbohydrate intake and
output. The highest excretion being immediately after the starvation
period, when the intake was at its lowest level. Allen ^^ has directed
attention to this apparent inability of the organism, be it normal or
diabetic, to handle a sudden increase in the carbohydrate intake after a
period of starvation or of low carbohydrate feeding. This also is

TABLE 6.— Sl-.\im-\ry of Diet. Urine and Blood Sugar of C. M. <(Case 3)

Duration of
Period

5 days
12 days
8 days
5 days
4 days

Diet

-*

Urine Sugar

Blood Sugar

Protein

C-40
90
100
100
101

Fat Carbohyd.
House diet

0 0
131 90
130 100
125 150
125 200

Per Cent. Gm.
2.2-3.8 33.2-53.4
0.6-1.3 16.0-23.1
1.0-3.1 20.3-57.5
1.2-l.V 22.3-38.1
0.8-1.6 28.7-39.0
1.6-1.7 25.3-48.8

Fasting
0.115
0.1S8
0.112
0.116
0.117
O.IH

Digestion
0.133
0.144
0.141
0.161
0.121
0.136

TABLE 7.

— Reac-

riON- OF Si

GAR IN

Blood and Urine

TO Special Diets

Duration of
Period
3 days
15 days
6 days

3 days

4 days

Diet

Urine Sugar

Blood Sugar-

Protein

SO
40-80

SO

Fat Carbohyd.
House diet

SO 250
0 0

100 0
so 250

Per Cent. Gm.
1.40-2.60 30.1-33.3
0.90-2.60 37.7-42.0
1.02-1.70 17.4-20.7

0.92-2.01 13.3-16.1
l.(»-5.40 37.5-54.0

Fasting
O.0S6
0.068
0.088
0.046
0.078
0.089

Digestiont
0.141
0.13«
0.093
0.06>
0.087}
0.242

* Blood sugars. Folin Wu method.
t Blood 1% hours after food.
: Severe acidosis. Van blyke 33.2 at end
after two days of balanced diet.

third day, but rose

60.5 volumes per cent.

shown in Table 7, where the sudden change from a pure fat diet to
a liberal regime containing 250 gm. of carbohydrate is associated with
a marked digestion hyperglycemia (0.212 per cent.) and one of "the
highest sugar outputs (54 gm.) ever found in this case. During the
second admission studies were made of the effects of high carbohydrate,
high protein and high fat intakes, with the following results.

After a preliminary period the patient was given a fixed diet for
fifteen days with increasing fluid intakes of 1,000, 2,000. 3.000 and
4,000 c.c. per diem, but under these conditions the output showed only
minor variations, an evidence of the lack of relation between the
diuresis and the sugar leakage. \\'ith a pure protein dietary, there was
a gradual drop in the blood sugar to the extremely low level of 0.046

Three Cases of "Renal Glycosuria." .Arch,

16. .Mien. F. M.. and Associates:
Int. Med. 24:52,3 (Nov.) 1919.

LEWIS— REXAL GLYCOSURIA 425

per cent., but even on this day he excreted 19.42 gm. sugar. So far
there had been no sign of acidosis, but on the third day of a pure fat
diet he developed an alarming acid intoxication. The bicarbonate
reserve dropped to 33 volumes per cent., and he became drowsy. The
diet was changed at once to a more varied one, and in two days all
sign of the acidosis had disappeared ( bicarbonate reserve 60.5 vokunes
per cent.)

Response to Added Glucose. — Three blood sugar curves were
carried out, in each case the usual dose of 100 gm. was given in
lemonade.

T.A.BLE 8. — Response of Urine .\nd Blood Sug.^r to 100 Gm. of Glucose

Hour
June 16. 1920

Blood
Sugar

8:45 a.m.
9:00 a.m.
9:30 a.m.

0.119

o.m

per Hr. Remarks

Test taken afti r two days

— of fasting and three of

carbohydrate-free diet.

1.S5 100 gm. glucose at !i:00

a.m.

March 16, ISil

8:45 a.m.

0.068

9:30 a.m.

0.122

10:00 a.m.

0.136

10:.'iOa.m.

0.1T8

12:00 m.

0,083

2:00 p.m.

O.061

March 23, 1921

8:45 a.m.

9:00 a.m.

0.127

After two weeks of pro-

tein, 80 gm.; fat. 80 gm.:

carbohydrate, 260 gm.

2.77

ICO gm. gljcose at 9:03

1.87

39«

0.68

.4fter three days protein.
40-60 gm.; caiboliydrate

free. 100 gm. glucose

0 51

at 9:00 a.m. Voided at

10:15 a.m.

The second curve is the one most nearly approaching normal. The
maximum rise in blood sugar is a little high and its appearance is
slightly delayed, but this can be accounted for by the arteriosclerosis
and low grade nephritis '" as shown by the albuminuria, casts, etc.
The slow return to normal levels (three hours) can also be explained
on the same grounds.

The first and third curves are decidedly abnormal, but are similar
to those reported in other cases in which the test period was preceded
by -starvation or restricted carbohydrate intake. The fact remains,
that in each case the kidney excretes sugar, while the glycemia is
within normal limits, and in the third test the urine contained 2.7 per
cent, sugar with a blood sugar of 0.064 per cent.

17. Janney. X. : Discussion. J. A. M. A. 75:217 rjuly 24) 1920.

426 ARCHIVES OF IXTERXAL MEDICIXE

In this case the urine was collected in two hourly specimens during
the day and a single specimen at night, and the hourly output was found
to be remarkably constant. It was highest in the morning, fell during
the afternoon and reached its lowest point during the night, showing
in this regard a close resemblance to the output of albumin in that
other anomaly of kidney action, orthostatic albuminuria.

TABLE 9-R.

ATE OF Si

UGAK Excretion'

FROM Hour to

Hour

Sugar

Volume

Specific

March 7, 1921

in C.c.

Gravity

Per Cent.

Gm.

Gm. per Hr.

Eemarks

8-10 a.m.

95

1.034

4.59

4.36

2.18

Blood sugar: a. c,

0.104:

10-12 a.m.

150

1.030

2.86

4.29

2.15

p. c.

,0.134

12- 2 p.m

335

1.030

1.-4

5.82

2.91

Diet:

protein, SO

gm.:

ISo

1.028

1.99

1.S4

fat,

80 gm.: carbohy-

1 6 ?;S:

155

1.029

2.02

siia

1.5T

drat

e, 250 gm.

6- 8 p.m.

U5

1.030

2.40

3.48

1.T4

8p.m.-8a.m.

1.0.50

1.021

14.23

1.19

Total

2.U5

39.05

SUMM.ARY

Notes are presented on three cases of renal glycosuria. The first
patient, after six years of observation, still presents a marked glycosuria
without symptoms and with a normal amount uf sugar in the blood.
He is apparently in excellent health.

The second is a severe case of chronic diffuse nephritis, in which
a glycosuria appeared while under observation. The glycosuria has
been practically continuous since its onset twelve months ago. It
is small in amount, the largest quantity being 16 gm. ; it is largely
independent of the carbohydrate intake, it required three days starva-
tion before its disappearance, and reappeared on an intake of 30 gm.
carbohydrate as green vegetables. The amount of glucose does not
show any constant relation to the urinary volume. Synchronous sugar
determination on blood and urine show the presence of a glycosuria
with 0.068 per cent, sugar in the blood. The response to 100 gm.
glucose falls within the limits of a nondiabetic case of nephritis. There
are no other signs of a diabetes mellitus.

The third patient, aged 74, has a marked arterio-sclerosis, (arterio-
sclerotic kidney), and was first seen on account of scurvy. The
duration of the glycosuria is unknown. The glycosuria has been
continuous for the past fifteen months, the usual output varying from
30 to 50 gm., and the ordinary changes in the diet had very little eflfect
on the amount of sugar excreted. The sugar output was found to be
independent of the urine volume. Synchronous studies of the blood
and urine showed 2.7 per cent, sugar in the urine, with only 0.064 per
cent, in the blood. The blood sugar curve following ingestion of 100
gm. glucose is somewhat atypical, but can be explained by the presence
of arteriosclerosis and nephritis. There are no signs of diabetes
mellitus.

LEn-IS~R£XAL GLYCOSURIA 427

It is generally recognized that there are two types of renal glyco-
suria '" : the one of unknown or idiopathic origin in which the blood
sugar curve is of a strictly normal order, the other, is associated with
a chronic diffuse nephritis or an arteriosclerosis, in which case the
patient shows a remarkably high and prolonged rise in the blood sugar,
which is probably a retention phenomenon, or may be connected with
the high diastatic activity of the blood, so often seen in severe nephritis.
The first case is an example of the idiopathic type, while the second
and third are examples of the second group.

Acknowledgment is due to Dr. E. H. Mason who has kindly placed at my
disposal much of the data connected with the second and third cases ; also
to Miss Lane for technical assistance.

CHEMICAL STUDIES OF THE BLOOD AND URINE

OF SYPHILITIC PATIENTS UNDER ARSPHEN-

AMIN TREATMENT

WITH A NOTE OX THE MECHANISM OF EARLY ARSPHEN-
AMIN REACTIONS*

CHARLES WEISS, Ph.D.. and ANNA CORSON, B.Sc.

PHILADELPHIA

The chemical analyses of the blood and urine of five cases of
tertiary syphilis form part of an investigation conducted in collabora-
tion with Schamberg, Kolmer and Raiziss ' on the subject of the
causes and mechanism of the severe reactions occasionally observed
after the intravenous administration of alkaline solutions of arsphen-
amin. Since the clinical and pathological literature on arsphenamin
reactions has frequently been reviewed, we will concern ourselves here
with a small number of available papers that deal with strictly chemical
investigations made by modern acceptable methods.

REVIEW OF THE LITER.\TUkE

Very little literature is available concerning the effects of arsphen-
amin on metabolic processes. Marischler and Schneider,- who were
among the first to investigate this problem, unfortunately used the sub-
cutaneous route of administration. Since the journal in which their
paper is published is inaccessible, a brief summary of their work will
be of interest. They investigated three ca.ses of syphilis : one primary,
one secondary and one tertiary. The patients were kept on a constant
diet. An increased excretion of calcium oxid and phosphorus in the
urine and feces was found in all three cases, and in one case there
was an increase in tlic ])li(is])Iiorus to nitrogen ratio, with a currcspond-

* From the Dcrmatological Research Institute.

* Investigation aided by funds accruing from the preparation of arsphenamin.

* A preliminary note on this subject was publislied in tlie Proceedings of
the Society for Experimental Biology and Medicine 18:210. 1921.

1. Schamberg, J. F. ; Kolmer, J. A., and Raiziss. G. \V. : Experimental and
Clinical Studies of the Toxicity of Dioxvdiamino-Arsenobenzol Dihydrochloridc,
J. Cutan. Dis. 35:286, 1917. Schamberg, J. F.; Kolmer. J. A.; Raiziss, G. W.,
and Weiss. C. : Laboratory and Clinical Studies Bearing on the Causes of the
Reactions Following Intravenous Injections of .Arsphenamin and Neo-Arsphena-
min. Arch. Dermat. & Syph. 1:235. 1920. Weiss. C: Phenol Elimination
in the Dog After Intravenous Injection of Neo-.Arsphenamin, Proc. Soc. Exper.
Biol & Med. 17:10.3, 1920.

2. Marischler, J., and Schneider, N. : The Effect of Subcutaneous Injections
of Salvarsan on Metabolism. Lwowski Tvgodnik Lekarski (Lembcrg Medical
Weekly [Polish]) 7;6.3. PI, 1512.

IVEISS-CORSOX— BLOOD AXD URIXE IX SYPHILIS 429

ing decrease in the nitrogen output. These authors probably used
acid solutions of arsphenamin (although no data as to this are given
in the original article). They concluded that arsphenamin acts like
an acid, combining with the alkali of the body, especially with the
calcium of the bone cells.

Rowntree, Marshall and Chesney ^ reported a case of "tabes dorsalis
with acute and chronic nephritis" in whicii death resulted from arsphen-
amin poisoning. The total nonprotein and amino-nitrogen of the blood
were distinctly abo\e normal, being 150 and 12.4 mg. per hundred c.c,
respectively.

Rappleye * studied the effects of intravenous injections of arsphen-
amin on the blood urea and the phenolsulphonephthalein elimination
in paretic patients. In one series of cases, with normal urea values
just before injection (from 7.8 to 15.4 mg. urea nitrogen per hundred
c.c. of blood), only three out of tiie nine patients who were examined
one hour after the injection, showed very small increases of from 2.5
to 3.1 mg. urea nitrogen in 100 c.c. of blood. In two other series
of cases tested three and twenty-four hours after injection, respecti\ely,
no changes were detected.

Rappleye also tested the blood urea and kidney function in another
series of cases which had been under treatment for a long time, and
had received a total of from 11 to 32 gm. of diarsenol (the Canadian
brand of arsphenamin). Only two of the tour patients who had
received more than 19 gm. of diarsenol showed low dye elimination
(10 and 20 per cent., respectively). Although these patients "were
in bed and suffering from considerable edema and were in poor gen-
eral condition, showing occasional hyalin and granular casts in the
urine," their blood urea nitrogen values were normal. Of the remaining
si.K patients who had received less than 19 gm. of diarsenol. all had
normal blood urea nitrogen values, although five of them eliminated
low percentages of phenolsulphonephthalein (from 20 to 50 per cent.),
and three of the latter showed albumin and ca.sts in the urine. In
only three of the entire series of ten cases, were the phenolsulphoneph-
thalein and urea-nitrogen values comparable. These three gave
normal figures. In all others, the phenolsulphonephthalein elimina-
tion was much lower than normal, but there was no increase in the
urea nitrogen of the blood. Rappleye concludes that diarsenol has no
deleterious efYect on the kidneys when their functioii is good at the
outset.

3. Rowntree. L. B.: Marsliall. F.. K.. and Chesney. A. M. : Studies in Liver
Function, Tr. Assn. Am. Phys. 29:586. 1914.

4. Rappleye, W. C. : Xotes on the Effect of Intravenous Diarsenol, J. Lab
& Clin. M. 4:630, 1919.

430 ARCHIFES OF IKTERXAL MEDICI XE

Elliott and Todd " made similar studies on syphilitic young men
who were receiving weekly intravenous injections of arsphenamin.
They report as follows : Of twenty patients on whom phenolsul-
phonephthalein determinations were made both before and after a
course of six arsphenamin injections with a total dosage of 2.7 gm.,
five showed a reduction of from 10 to 17 per cent, each, while the
other fifteen remained practically unchanged. The urea-nitrogen con-
tent of the blood in these twenty cases showed an average of 14.3 mg.
per hundred c.c. before and 16.0 nig. after treatment, or practically
no change. One case showed an increase of 9 mg. No details are
given of these analyses.

In another series of nine cases, receiving injections twice a week
with the same total dosage (2.7 gm.), the results were as follows:
Before injection, the urea nitrogen of the blood was normal, varying
from 12.6 to 20 mg. per hundred c.c. of blood. After injection, it
increased slightly (2 mg. per hundred c.c.) in four of the patients, but
none of them had more than 20 mg. urea-nitrogen per hundred c.c.
(the upper normal limit). The phenolsulphonephthalein elimination,
on the other hand, was low to begin with — from 41 to 58 per cent,
before injection — and was decreased somewhat, being from 40 to 51
per cent, after injection. In three of these cases there were reductions
of from 11 to 17 per cent., while in two others there was 5 per cent,
reduction, without corresponding retention of nitrogen.

In one case of acute syphilitic nephritis, Elliott and Todd observed
high blood urea figures which declined rapidly under the arsphenamin
treatment. Albumin also disappeared from the urine. The phenol-
sulphonephthalein excretion, on the other hand, which was rather high
to begin with, declined somewhat during the treatment, in spite of the
obvious renal improvement. The authors concluded that "these find-
ings suggest that the admitted inadequacy of the phenolsulphonephtha-
lein test to detect acute nephritis applies also to syphilitic nephritis."

Bailey and MacKay," in an extensive chemical study of the blood
and urine of twenty-five cases of syphilis in which toxic jaundice had
developed tmder combined mercury and novarsenobillon (French brand
of neo-arsphenamin) treatment, observed:

(1) In eleven of the cases, bile pigments were found in the plasma,
and, as a rule, also bile salts.

(2) The greatest excretion of urobilinogen and urobilin was in the
jaundiced cases. In the absence of jaundice, the output of these pig-
ments was less than half.

5. Elliott, J. A., and Todd, L. C. : Eflfects of .Arsphenamin on Renal Func-
tion in Syphilitic Patients, Arch. Dermat. & Syph. 2:699. 1920.

6. Bailey. C. V.. and MacKay. A.: Toxic Jaundice in Patients Under Anti-
syphilitic Treatment. Arch. Int. Med. 25:628 CMay') 1920.

U'EISS-CORSOX— BLOOD AXD URIXE JX SYPHILIS 431

(3) In the voided urine, the relative excretion of urobilinogen to
urobilin was greatest in the well patients and least in the severe liver
cases, indicating a decreased excretion of oxidase in the latter.

(4) In twelve of the patients with no disorder of the liver, the
percentage of cholesterol varied from 0.117 to 0.210, with an average of
0.155, or within a fairly normal range. In all of the twenty-five cases
of toxic jaundice, the cholesterol value was strikingly high, varying
from 0.165 to 0.292 per cent., with an average of 0.235 per cent.
In twenty-one of the twenty-five cases the values were over 0.2 per cent.
Bailey and MacKay consider hypercholesteremia as being an early and
marked sign of toxic jaundice and a valuable indication of a precarious
state of the liver in this disease.

(5) The blood sugar and rate of excretion of sugar in the urme
were normal in all of these patients.

(6) As for the nitrogenous constituents of the blood of these
patients (who were on a high protein diet), the urea-nitrogen was
normal (from 8 to 20 mg. per hundred c.c. of blood) in eight, but
above normal (22 mg. and above) in the majority; one was as high
as 49 mg. per hundred c.c. of blood. These abnormal values were
ascribed, in part, to the high protein diet and the various physic restric-
tions placed on the patients (soldiers), and, in part, to impaired kidney
elimination, since they were accompanied by slight increases in the
creatinin and marked increases in the uric acid of the blood as well
as a decreased elimination of uric acid in the urine. The high blood
uric acid was attributed to an increased production of this substance
resulting from a destruction of liver nuclear substance. It is of interest
to note that casts and protein were found only occasionally in the
urines of a few of those patients who showed retention of nitrogen
in the blood.

While this manuscript was being prepared, Anderson's paper on
the effect of arsphenamin on kidney function appeared.' This author
reports chemical analyses of the blood in thirty-eight cases of syphilis
in which the total dosage of arsphenamin ranged from 8 to 21 gm.
with an average of about 14 gm. The treatment extended over a period
of two years and included inunctions or injections of various mercurials
in addition to arsphenamin injections. Of the thirty-eight patients
(we omit here the case which he characterized as nephropathic),
twenty-four had total nonprotein nitrogen values higher than 30 mg.
(the upper normal limit), four of these ranging from 40 to 46 mg.
per c.c. of blood. The urea-nitrogen and creatinin values were normal
in all, although one case showed albumin and globulin in the urine

7. Anderson, H. B.: Some Observations on the Use of .Arsphenamin. Am.
J. M. Sc. 162:80, 1921.

432 ARCHIVES OF IXTERXAL MEDIC IX E

(no casts). The phenolsulphonephthalein test gave the following
results. Of five patients with somewhat subnormal dye elimination
(from 45 to 50 per cent, in two hours) only one patient had a non-
protein nitrogen value of 46, the others being normal. Case 25 of
Anderson, diagnosticated as "tabes and nephritis" had a phenolsul-
phonephthalein elimination of 40 per cent, in two hours, total non-
protein nitrogen 37 mg. per hundred c.c. of blood, and urea and
creatinin normal. The urine showed a distinct trace of albumin but
no globulin and a few granular casts. Anderson draws the conclusion
that there is no evidence of kidney injury.

In this connection, some of the recent histopathologic studies of
Kolmer and Lucke " must be referred to briefly. These authors found
the following changes in normal rabbits and rats injected intravenously
with repeated small (therapeutic) doses of either neo-arsphenamin or
alkaline arsphenamin :

The liver showed small areas of focal necrosis and slight periportal fibrosis.
The latter was confined to the tissue about the bile ducts and blood vessels.
The kidneys revealed vascular and tubular changes characterized as "nephrosis."
More or less marked chronic passive congestion was found with moderate
hemosiderosis in the spleen; inconspicuous amounts of hemosiderin occurred
in the lung. Occasional vessels contained thrombi composed of partly or
entirely conglutinated or hyalinized erythrocytes. The lipoids of the supra-
renals were at first increased in quantity; later a slight exhaustion appeared.
Parenchymatous changes of mild degree were seen in the various organs.

PL.AN" AND METHODS OF I^•VESTlG.^TIO^■

As has been brought out in the review of the literature, previous
writers have limited their work to a single chemical analysis of the
blood and urine of the patient after a course of arsphenamin injections
had been given, with the hope of detecting kidney injury. The present
work was begun before any of the above publications had appeared.
Our object being primarily to study the mechanism of early arsphen-
amin reactions, we deemed it necessary to hospitalize our cases, and
to make frequent analyses of the blood and urine before injection and
then for a sufficient interval after treatment. In this way we hoped
to detect changes, which could, with some degree of certainty, he
ascribed to the action of the drug.

Five cases of tertiary syphilis with varying degrees of optic atrophy
were selected from the Skin Clinic of Dr. Jay F. Schamberg, Poly-
clinic Hospital, Graduate School of Medicine, University of Pennsyl-
vania. They were kept in a ward at the Polyclinic Hospital and given
the usual house diet, which was low in proteins and fats, rich in
carbohydrates and fairly constant from day to day. Their water intake

8. Kolmer. J. A., and Lucke. B. : E.xperimental Studies on the Histopatho-
logic Changes Produced bv .'\rsphcnamin and Keo-.'Xrsphenamin. Arch. Dermat.
& Syph. 2:289. 1920; Ibid.'3:48.v 1921.

IIEISS-CORSOX— BLOOD A.\D URIXE IX SYPHILIS 433

was also controlled. Blood specimens were taken in the first two
cases, three hours after a constant prescribed breakfast and in the last
three cases, before breakfast.

The following were investigated: (1) The urea and total non-
protein nitrogen, sugar and uric acid of the blood, by the methods of
Folin and. Wu.' The total nonprotein nitrogen was estimated by the
digestion and direct nesslerization technic, and urea nitrogen by the
aeration and titration method of Van Slyke and CuUen.'" (2) The
carbon dioxid combining power of the plasma was detennined by Van
Slyke's method. ^^ (3) Kidney function was determined by the phenol-
sulphonephthalein test of Rowntree and Geraghty,^^ the dye being
injected intramuscularly. (4) The daily twenty-four hour specimens
of urine were also analyzed routinely for sugar and albumin, and
microscopic examinations were made of the sediment. (5) The
hydrogen-ion concentration of the blood and urine was measured by
the colorimetric method of Sorenson, as developed by Bayliss ^^ and
Clark," respectively. (6) Total nitrogen in the daily twenty-four-hour
urine was estimated by the gross Kjeldahl method in the usual way.
(•7) Arsenic elimination in the urine was determmed by Green's micro-
titration method. These results are reported separately. ^^

Before beginning this investigation the various methods employed
were carefully tested out by running "recovery" and "control" experi-
ments. All determinations reported were done in duplicate, and only
checking results were accepted.

REPORT OF CASES ^'^

C.\SE 1.— T. M., aged 38; optic atrophy of left eye. This patient was in
good physical condition. He had been under continuous treatment for more
than two years, and had always suffered reactions. During the "control periods"
(before injection), normal values for urea, total nonprotein nitrogen and blood
sugar were observed. Oct. 26, 1920, before breakfast, the patient was given an
intravenous injection of 0.6 gm. of an alkaline solution of arsphenamin. dis-
solved in 120 c.c. distilled water. The patient had a very mild reaction —
nausea and vomiting, which continued until the next morning. Three hours

9. Folin. O., and Wu, H.: A System of Blood Analysis, J. Biol. Chem.
38:81, 1919; Ibid. 41:367. 1920.

10. Van Slyke, D. D., and Cullcn. G. E.: A Permanent Preparation of
Urease and Its Use in the Determination of Urea. J. Biol. Chem. 19:211. 1914.

11. Van Slyke, D. D. : A Method for the Determination of Carbon Dioxid
and Carbonates in Solution, J. Biol. Chem. 30:347, 1917.

12. Rowntree, L. G.. and Geraghty, T. J. : Described by Myers," pp. 103-104.

13. Bayliss, W. M.: The Neutrality of the Blood, Brit. J. Physiol. 53:
162, 1919.

14. Clark, W. M. : The Determination of Hydrogen Ions, Williams and
Wilkins Co., Baltimore, 1920,

15. Weiss, C, and Raiziss, G. W. : The Elimination of Arsenic in the Urine
of Syphilitic Patients After Intravenous Injection of Arsphenamin, Arch.
Int. Med. (1922) to be published.

16. Additional data are given in the preceding paper.'"

434 ARCHIVES OF INTERNAL MEDICINE

after the injection, the total nonprotein nitrogen rose to 32.9 mg., which is
slightly above the upper normal limits given by Myers." The urea remained
unchanged. No food had been taken during this interval. The blood sugar
increased from 100 to 123.8 mg. per hundred c.c. On the following morning,
the figures for urea and total nonprotein nitrogen (blood specimen obtained
three hours after a prescribed breakfast) continued to rise, the former reach-
ing 6.8 mg. above the normal range.

TABLE 1.— Showing Che.mical Ch.^nges in the Blood of Syphiutics
During Arsphenamin Treatment

Date

Total Nonprotein
Nitrogen *

Drea Nitrogen

Sugar

Bemarks

Case l: T. M.;

male: aged 38;

optic atrophy

10/25/20

24.2

11.1

85.1

30.0

13.2

100.0

t

32.9

13.2

123.8

VI

10/27/20

36.8 .

20.6

130.2

V

10/29/20

31.4

17.2

133.8

11/ 2/20

39.5

14.3

115.6

11/ 4/20

33.4

14.9

103.9

11/ 8/20

34.3

22.4

131.6

11/ 9/20

23.2

8.1

133.3

t

11/ 9/20

28.4

9.4

129.9

m

11/10/20

44.3

14.8

121.2

11/13/20

34.5

11.3

137.9

11/16/20

32.6

12.6

137.9

11/18/20

31.0

12.4

139.9

11/22/20

32.4

18.1

121.6

11/23/20

34.8

16.9

116.3

vt

11/24/20

38.2

18.3

117.6

11/26/20

30.3

13.5

106.1

Case 2: M. J.

male; aged 28;

optic atrophy

1/19/21

33.5

11.4

93.5

1/20/21

32.1

11.4

88.3

1/21/21

32.6

11.7

111.4

1/25/21

30.7

12.2

76.6

1/25/21

32.8

15.5

153.8

v:

1/27/21

31.5

11.3

113.0

1/28/21

31.5

12.6

89.9

1/31/21

30.7

14.2

137.9

2/2/it

38.8

18.6

96.4

G

2/ 4/21

17.5

119.0

21 7/21

30.9

14.3

107.5

• All figures are given in milligrams per hundred c.c. ol blood.
i Injection immediately after sample was drawn. . ,. ,.

I Patient received an intravenous injection ol 0.6 gm. arsphenamin three hours belore
this sample of blood was drawn.
V = "reaction"— patient vomited.

D = severe reaction with vomiting, diarrhea and pain in the legs.
G = gastric crisis. C = slight reaction, clillls.

During the succeeding days, the values fluctuated somewhat. One week
after injection a total nonprotein nitrogen value of 39.5 mg. per hundred c.c.
of blood was observed. The urea nitrogen of this specimen was normal
(14.3 mg.). The highest urea value, observed thirteen days after injection,
was 22.4 mg. with a corresponding nonprotein nitrogen figure of 34.28 mg.,
both indicative of a very mild nitrogen retention.

After the second injection, the reaction was much more severe. The patient
became very sick; he had vomiting, diarrhea, pain in the legs and oliguria
with bile tinged urine. The blood specimen taken three hours after this
injection (no food having been consumed) fhowcd increases similar to those
seen after the first dose. On the next morning, however, the highest value

17. Myers, V. C. : Practical Chemical Analysis of Blood, St. Louis, C. V.
Mosby Co., 1921, pp. 70-82, 103-104.

WEISS-CORSOX— BLOOD AND URINE IX SYPHILIS

435

for total nonprotein nitrogen was obtained, 44.3 mg. per hundred c.c. blood.
The urea-nitrogen remained normal, 14.8 mg. per hundred c.c. An increase of
2 gm. above the usual range was observed in the urinary nitrogen excretion
of the succeeding day. The urea and sugar continued to remain normal during
the next two weeks, but the total nonprotein nitrogen figures were slightly
above normal.

The third injection, now given, was followed by a mild reaction, and the
analytical figures were very much similar to those seen after the first dose.
The blood sugar values never assumed pathologic significance.

Case 2. — J. M., male, aged 28; total optic atrophy. This patient was in
good physical condition ; he had been under continuous treatment for over two
years. During the control period (one week) the values for urea and total
nonprotein nitrogen were or gradually became normal. A mild reaction (vomit-
ing) followed the first injection of 0.6 gm. of alkaline arsphenamin. The blood
specimen taken three hours after this injection (the patient having taken no
food) showed slight rises (from 2 to 3 mg.) in the urea and total nonprotein
nitrogen. The blood sugar was more than doubled but never assumed patho-
logic significance. No other significant changes were observed until eight
days later when the urea and total nonprotein nitrogen rose above their usual
values, reaching 18.6 and 38.8 mg. per hundred c.c. of blood, respectively. Both
gradually declined during the course of the next few days.

Case 3.— K. W., male, aged 39; locomotor ataxia; total optic atrophy. This
patient was not in good nervous or physical condition. During the control
period of observation (lasting twelve days) the values for urea-nitrogen, sugar
and uric acid were normal. The total nonprotein nitrogen values, however,

TABLE 2.— Biochemical Data on Case 3
K. W.; male; aged 39; optic atrophy

Total

Urea

Date

&'."^^'."

Nitrogen

Sugar

Uric Acid

Remarks

2/23/21

32.2

0.9

126.6

2/2S/21

3S.8

11.0

124.6

2/28'21

35.1

12.5

132.5

2.4

3/ 2/21

35.2

12.3

128.2

2.4

3/ 4/21

33.0

13.7

144.4

2.4

3/ 7/21

28.0

9.1

173.2

1.3

t

3/ 7/21

30.9

14.5

160.6

1.4

CI

3/ 8/21

31.4

11.7

173.5

3/10/21

28.9

8.6

160.6

2.4

3/14/21

29.1

6.0

S6.5

2.4

3/16/21

27.9

9.6

96.4

2.4

3/18/21

24.0

7.8

93.0

3/21/21

26.5

12.2

86.4

2.4

t

3/21/21

29.0

14.0

149.3

2.4

vt

3/22/21

35.2

14.9

147.6

2.0

3/24'21

24.6

8.9

107.0

3/24 '21

26.4

10.3

199.5

CJ

3/25/21

26.5

13.1

102.6

3/29/21

22.9

7.3

97.S

2.2

3/31/21

24.9

10.0

98.2

2.4

4/ 5/21

34.7

11.2

95.2

2.6

4/ 7/21

26.1

14.9

91.3

2.2

• All figures are given In milligrams per hundred c.c. of blood.
i Injection immediately after sample was drawn. . ,. ,.

t Patient received an Intravenous injection of 0.6 gm. arsphenamin three hours Doiore
this sample of blood was drawn.

V = "reaction"— patient had severe chills.
C = slight reaction— chills.

were often somewhat above normal, ranging from 28 to 35.2 mg. per hundred
c.c. of blood. The first injection of 0.6 gm. arsphenamin was not followed by
any untoward symptoms, except slight chills. The usual small increases in
the urea and nonprotein nitrogen (and uric acid) of the blood were observed
three hours after the injection. No other significant changes were detected.

436 ARCHHES OF IXTERXAL MEDICIXE ,

The second injection of 0.6 gm., given a fortnight after the first, was fol-
lowed by somewhat more severe chills. The blood findings were similar,
except that the blood sugar rose appreciably three hours after the injection.

Three days later a third injection was given and was followed by similar
results. The urea nitrogen content of the blood in this patient never rose
above IS mg. per hundred c.c.

C.'^SE 4. — G. R., male, aged 48; partial optic atrophy; diminished hearing;
tremor; Rhomberg positive. Loss of tactile sensation. General physical con-
dition fair. In this case and in Case 5 an effort was inade to determine whether
or not the alkali used to neutralize arsphenamin produced any appreciable
change in the carbon dioxid combining power of the plasma or in the hydrogen-
ion concentration (/>h) of the blood or urine. In addition to chemical analyses
of the blood," we made phenolsulphonephthalein elimination tests and careful
examination of the urine for albumin and casts.

During the control period of observation (one week) this patient showed
abnormal values for urea and total nonprotein nitrogen of the blood, from
24 to 30 and from 36 to 41 mg. per hundred c.c, respectively. The carbon
dioxid and sugar values were, however, normal. The phenolsulphonephthalein
elimination was not comparable with the nitrogen figures, being within the
normal range. The urine frequently showed a few granular and hyalin casts,
renal cells and leukocytes and an occasional red blood cell. There was no
proteinuria or glycosuria. The pn of the urine was. within normal range.

After the first injection of 0.6 gm. alkaline arsphenamin no untoward symp-
toms were observed. There were the usual small increases in the total non-
protein nitrogen and sugar of the blood three hours after the injection. The
urea, carbon dioxid combining power and the pa of the blood remained
unchanged. The urine taken iminediately after the injection was completed,
showed a very small increase in alaklinity. Forty-four hours- after the injec-
tion an appreciable increase in the urea and total nonprotein nitrogen of the
blood was observed. The former rose to 32.3 mg. and the latter to 44.1 mg,
per hundred c.c. blood. One week later a second injection of 0.6 gm. was
given. There was no clinical reaction. The increase in total nonprotein nitrogen
observed three hours after the injection was 4 mg. per hundred c.c: the small
increases in urea nitrogen and sugar of the blood were similar to those usually
noted. There was no increase in the pa of the blood. A specimen of urine,
taken immediately after the injection was completed, showed a very small
increase in alkalinity similar to that noted in the first injection. Two days
after the injection the urea dropped to normal although the total non-protein
nitrogen remained unchanged. At no time was any decrease in tlic phenol-
sulphonephthalein elimination observed.

C.\SE 5.— C. J., male, aged 53; partial optic atrophy. General physical con-
dition very good; patient complained of headche and occasional pain in luml)ar
region.

This patient showed abnormal total nonprotein nitrogen content of the
blood, although the urea and carbon dioxid combining power were normal.
He never showed any reaction (except very mild diarrhea) nor any subnormal
phenolsulphonephthalein elimination, although he regularly eliminated a few
hyalin and granular casts, leukocytes and red cells in the urine. The second
injection of 0.6 gm. alkaline arsphenamin (which we were able to follow more
carefully than the first) resulted in no appreciable change, other than the
usual tuarked rise in blood sugar (observed three hours after the injection).
The carbon dioxid combining power of the plasma remained unaltered.

The third injection showed a more appreciable rise in the total nonprotein
(but not in urea) nitrogen, and a similar increase in blood sugar tliree hours
after the injection.

18. Specimens of blood were taken Iicfore breakfast.

IVEISS-CORSOX— BLOOD AXD URIXE IX SYPHILIS

TABLE 3. — BiocHEMirAL Data on Cases 4 and 5
Case 4: G. B.; male; aged 48; optic atrophy

Blood Analyse

Kidney Function Tests

Total

Phenolsul-

Non-

phonephthalein

Drca

Remarks

tein
Nitro-
gen'

Xitro-

Sugar

Plasmat
CO:

per Cent.

Albu- Sediment

First

Total

Hour

2Hrs.

6/15/21

41.1

24.0

SS.9

58

S3

70

Sega- One of two gran i-

tive ; lar casts, slight

amount o£ sedi-

6/17/21

41.1

30.0

109.8

58

30

65

Xega-
tive

Many granular
casts, lew hyalin
casts, occasional
red blood cells.

i many white blood

j cells

6/20/21

36.5

23.9

101.8

58

45

60^3

Xega- 1 Many granular and

hyalin casts, few
renal cells and

6/22/21

36.1

29.6

105.3

63

Nega-
tive

Few casts and white
blood cells

6/22 '21
n/24/21

44.1

32.3

99.5

58

Nega-

Few casts, many

white blood cells

6/28/21

33.5

21.4

102.0

fiO

30

±60

Nega-

tive

Few casts and wh:tc
blood cells

t

6/30/21

39.2

19.4

92.0

53

50

65-70

^1?|-

of sediment, many
easts, white blood
cells, few epithel-

•

ial cells

Case 5: S. J.; male; aged 53; optic atrophy

7/ 7'21
7/ 7/21
7/12/21

7/14/21
7/18/21

117.3
137.9

94.8
120.9

Nega-
tive
Nega-

occa-

s, red
white

sional cast
blood cells,
blood cells
Fewgrr.nularc
white blood
renal cells,
spermatozoa

Occasional hyal n

cast and whit(
blood cells

Occasional grani

Occasional white
blood cells and a
few spermatazoa

' All figures for total nonprotein nitrogen, urea nitrogen and sugar are given in mill:-
grams per hundred c.c. of blood.

♦ Cubic centimeters of carbon dioxld reduced to oxygen, 760 mm. bound as bicarbonate
by 100 c.c. of plasma.

; Injected Immediately after sample of blood was drawn.
; Patient received an Intravenous injection of O.fl gm. arsphenamin three hours before this
sample of blood was drawn.

•■ Blood drawn during Injection after three fourths of dose had been administered.
± This phenolsulphonephthalein test was made the afternoon prior to injection.

D — very mild diarrhea.

438 ARCHIVES OF IXTERNAL MEDICIXE

THE MECHANISM OF EARLY ARSPHENAMIN REACTIONS

Evidence has been brought forth in the review of the literature as
well as from our own data, that : (a) arsphenamin does not exert any
selective injurious action on the kidneys ; (b) patients with injured
kidneys do not necessarily manifest arsphenamin reactions; and (c)
patients with good kidney function may suffer from severe reactions.

What, then, is the mechanism of arsphenamm reactions?

Numerous theories have already been suggested and these have
been reviewed by one of us (C.W.^) as well as by Hirano.^' The latter
suggested the hypothesis that arsphenamin causes a diminution in the
epinephrin content of the suprarenals and, therefore, of the circulating
blood thus producing shock to the organism. Work done in this
Institute by Drs. Lucke, McCouch and Kolmer (Journal Pharmacol.
Exp. Therapeutics, 1922 [in press] casts very grave doubt on Hirano's
findings and interpretation. That arsphenamin reactions bear no rela-
tion to true anaphylactic shock has been shown by the pharmacologic
studies of Hanzlik and Karsner.""

These authors, as well as Jackson and Smith,-' have shown that
in experimental animals, even therapeutic doses of arsphenamin
raise the pulmonary arterial pressure and dilate the right heart. Hanz-
lik and Karsner maintain that the symptoms observed after arsphen-
amin or neo-arsphenamin injections are due primarily to injury toJ:he
circulatory apparatus caused by the arsenic. (No distinction is to be
drawn between inorganic and organic arsenicals in their opinion.)
"Amelioration or partial protection afforded by adrenalin or atropin
is due entirely to improvement in the circulation."

Injury to the circulatory apparatus (and, perhaps, also destruction
of erythrocytes due to the hemolytic action of arsphenamin) probably
accounts for the small but constant increases in total nonprotein nitro-
gen of the blood observed by us within three hours after injection.

The recent toxicological data of Willcox and Webster - showing the
wide-spread distribution of arsenic in the organs of fatal cases of

19. Hirano, N. : Experimental Studies on the Nature of Anaphylactoid Reac-
tions Caused by tlie Repeated Intravenous Injection of Salvarsan, Kitasato
Arch. Exper. M. 3:1, 1919.

20. Hanzlik, P. J., and Karsner, H. T. : A Comparison of the Prophylactic
Effects of Atropin and Epinephrin in Anaphylactic Shock and Anaphylactoid
Phenomena from Various Colloids and Arepsenamin, J. Pharmacol. & Expcr.
Therap. 14:425, 1920. Effects of Various Colloids and Other Agents Which
Produce Anaphylactoid Phenomena on Bronchi of Perfused Lungs, loc. cit. 14:
449, 1920.

21. Jackson. D. E., and Smith, M. I.: An Experimental Investigation of
the Cause of Early Death from Arsphenamin, J. Pharmacol. & Exper. Therap.
12:221, 1918.

22. Willcox, W. H., and Webster, J.: The Toxicology of Salvarsan, Brit.
M. J. 1:473, 1916; The Analyst 41:231, 1916.

IVEISS-CORSOX— BLOOD A\D URIXE IX SYPHILIS 439

arsphenamin intoxication, as well as the histopathological studies of
Kolmer and Lucke * already alluded to (showing that practically very
organ is to a mild degree deleteriously affected during a course of
arsphenamin injections in experimental animals), lead us to suggest
that early arsphenamin reactions may not be due primarily to injury to
any specific organ alone but to a general tissue injury which may be
ascribed to the toxic action exerted by the drug or the products of its
oxidation or reduction in the tissues of certain hypersensitive cases.
That the liver also suffers injury is suggested by comparing our obser-
vation of increases in total nonprotein nitrogen, without corresponding
increases in urea-nitrogen of the blood, with similar data published by
Losee and Van Slyke -^ and Killian -* on eclampsia, and of Rowntree,
Marshall and Chesney ^ in various other diseases in which the liver is
known to be involved.

SUMM.\EV

Five cases of tertiary syphilis with varying degrees of optic atrophy
were studied. The details of the chemical analyses of the blood and
urine and the history of the cases will be found in Tables 1, 2 and 3
and in the text. Herewith are presented a brief summary of the
essential points noted.

1. Urea and Total Nonprotein Nitrogen of the Blood. — (a) Small
but definite increases in the nonprotein nitrogen of the blood (from
2 to 5 mg. per hundred c.c.) were observed three hours after prac-
tically every intravenous injection of 0.6 gm. doses of arsphenamin
(ten out of twelve injections). These increases cannot be accounted
for by the nitrogen content of arsphenamin (which is 5 per cent.).
The maximum rise accompanied the severest reaction (Case 1). The
increases in urea nitrogen were not always parallel to those in the
nonprotein nitrogen and often were absent or exceeded them.

(b) Blood specimens examined at intervals after every injection of
arsphenamin showed significant increa.ses above the original limits, only
in those cases (Cases 1 and 2), in which the reactions vyere most pro-
nounced. The nonprotein nitrogen in Case 1, which was from 24 to
30 mg. before injection, rose twenty-four hours after the injection to
44 mg. per hundred c.c. of blood. Case 2, with a less severe reaction,
showed a milder increase (from 33.5 before, to 38.8 mg. eight days
after injection). The urea nitrogen figures, however, remained
normal.

(c) In no case was the final total nonprotein nitrogen or urea-
nitrogen of the patient, when discharged (after one, two or three 0.6

23. Losee, J. R., and \'an Slyke. D. D. : The Toxemias of Pregnancy, Am.
J. M. Sc 153:94, 1917.

24. Killian, H.: Proc. New York Path. Soc. February. 1921.

440 ARCHIVES OF IXTERXAL MEDJCIXE

gm. doses of arsphenamin) any higher than when admitted. On the
contrary, many reductions were noted. All of the patients benefited
greatly by the low protein diet and hospital care, as well as by the
injections.

(d) Of the five cases studied( one (Case 4) had urea and total
nonprotein nitrogen values distinctly above normal before treatment
was begim. Yet this patient never showed untoward symptoms. On
the other hand. Case 1, with normal blood figures, reacted severely.
W'c cannot, therefore, in every case, ascribe arsphenamin reactions to
impaired kidney function alone, as suggested by \Vechselmann.=^

2. Blood Sugar. — Marked but not pathologic increases in blood
sugar occurred fairly constantly three hours after injection. Two or
three times (Cases 2 and 3) we noted that the blood sugars were
doubled, although no food had been taken during this interval. As a
rule, these increased values gradually subsided in the course of a
few days. Whether these sudden increases were due to stimulation
of the suprarenals, resulting from the action of the drug or from mere
fright, is a matter to be investigated.

Marked variations in blood sugar were also noted from day to
day in specimens taken at the same hour before breakfast. The nervous
state of the patient and the weather conditions seemed to be controlling
factors.

3. Uric Acid. — The uric acid of the blood was studied in Case 3
and it was found to be constantly normal during the investigation.

4. Carbon Dio.vid Combining Power of Plasma. — In cases 4 and 5
normal values were observed both before and during treatment. The
amount of alkali added to acid arsphenamin to produce the di.sodium
salt was insufficient to change either the bicarbonate reserve or the
hydrogen ion concentration of the plasma or of the urine. Ferannini ='
draws similar conclusions from his pharmacologic studies of the
respiratory rate in dogs.

5. Phcnolsidphoncphthalcin Elimination. — The elimination of this
dye was nornlal in each of the two patients studied (Cases 4 and 5)
although the former showed distinct signs of nephritis. There were
no changes after arsphenamin treatment.

The writers wish to thank Dr. Jay F. .Schamberg. Dr. John A. Kohiier and
Dr. George W. Raiziss of this Institute for their kind cooperation throughout
the work.

25. Wechselmann. W. : Ueber die Pathogenese der Salvarsantodesfalle,
Berlin, Urban and Schwarzcnberg, 1913.

26. Ferrannini. L. : Richerchc Sperimentali suU' azione farmacologia del
Salvarsan, Riforma mcd. 27:1065-1068, 1101-1106, 1126-1128, 1911.

. STUDIES IN THE VARIATION OF THE LENGTH
OF THE Q-R-S-T INTERVAL*

G. K. FEXN, M.D.

CHICAGO HEIGHTS, ILL.

I

REL-ATION TO HEART RATE AND TO CLINICAL CONDITIONS
ASSOCIATED WITH CHRONIC HYPERTENSION

\'ariations in the length of the different phases of the human
electrocardiogram have long been observed, and satisfactory explana-
tions have been offered for many of them by various writers. The length
of the entire ventricular complex is known to vary in different indi-
viduals and in the .same individual at different times but the explana-
tion of this phenomenon has received comparatively scant attention
at the hands of cardiologists.

The normal ventricular complex, the Q-R-S-T group, is usually
divided into subgroups, the Q-R-S interval and the S-T interval. Pro-
longation of the Q-R-S interval is due to defects in certain portions
of the conducting tissues. This fact has been established by animal
experiment and by carefully controlled clinical observation. There are,
however, certain cases in which the Q-R-S interval is normal but the
Q-R-S-T interval exceeds the accepted time limit. It is with such
cases that these studies have to deal.

Because of the great number of angles from which the subject
must be approached, it was considered advisable to report separately
on each investigation or group of investigations.

This report considers only those hearts in which the rhythm is
regular and the rate approximately constant and under 120 per minute.
The clinical conditions considered in this report may be termed chronic
as they have existed in each individual over a long period of time.
The mea.surements reported here were all made from Lead II of the
electrocardiogram.

The results of the work in other fields ot this investigation will
follow.

Lewis ' has defined the time relation of the Q-R-S group. He
states that this group must have a duration of no more than 0.1 second,
and that it usually con.stitutes less than one third of the entire ventric-
ular complex. A critical review of the literature fails to reveal any
carefully controlled work that places a definite time limit on the S-T

* From the Medical Service, St. Luke's Hospital. Chicago.
1. Lewis, T. : The Mechanism and Graphic Registration of the Heart Beat,
London, Shaw & Sons, 1920.

442 ARCHIVES OF IXTERXAL MEDICIXE

interval without taking into consideration other features of the heart's
activity. It is, therefore, safe to assume that the ventricular complex
may normally vary in its length, and while the Q-R-S group is rather
restricted in its variations, the S-T interval has much more latitude.
This assumption is borne out in the study of electrocardiograms taken
from normal individuals.

Garrod,- Thurston,^ Chapman,* Eyster," Einthuven ^ and others have
pointed out the relation of the duration of the systole to the heart
rate. Lombard and Cope ^ have devised a tormula by which the
systolic length may be predicted from the heart rate. Katz,* in care-
fully controlled animal experiments, has shown that the formula of
Lombard and Cope may be used to predict the systolic length in animals
whose heart rate is under 150.

In this work the prediction of the systolic length has been under-
taken using a modification of the Lombard and Cope formula. The
Q-R-S-T group is taken to represent ventricular systole. Wiggers
and Clough ' have shown that systole of the ventricle may be divided
into two periods, the isometric period and the ejection period. The
isometric period varies between 0.04 and 0.06 second, regardless of
the heart rate or systolic length and is comparable to the Q-R-S group
in that its limits of variation are narrow. Lewis ^ states that while
the Q-R-S group begins before the actual contraction of the ventricle,
the ventricular contraction has its inception sometime during the
recording of the Q-R-S and ends within 0.03 second of the completion
of the T. As the initial phase of ventricular activity is recorded 0.2
second or less before the contraction, the Q-R-S-T may be taken to
represent the period of ventricular activity with a probable error of
less than 0.05 second.

Lombard and Cope " devised the formula S = -~= to determine the
relation of systole to heart rate. In this formula S represents the
systolic length in seconds, R the heart rate per minute and K a con-
stant which they found varied with different positions of the body.

In studying the electrocardiogram it was found that the substitution
of cycle length in seconds for heart rate per minute greatly facilitated
matters. The heart rate is reciprocal of cycle length and may be
expressed by the formula R= -^i, R being the heart rate per minute

2. Garrod, A. H.: T. Anat. & Physiol. 5:17, 1871.

3. Thurston, Edgar: J. .'\nat. & Physiol. 10:494, 1876.

4. Chapman, P. M.: Brit. M. J. 1:511, 1894.

5. Eyster, J. A. E.: J. Exper. M. 14:594, 1911.

6. Einthoven, W. : Arch. f. d. ges. Physiol. 122:532, 1908.

7. Lombard, W. P., and Cope, O. M. : Am. J. Physiol. 49:140, 1919.

8. Katz, L. N.: J. Lab. & Clin. M. 6:291, 1921.

9. Wiggers, C. J., and Clough, H. D. : J. Lab. & Clin. M. 4:624, 1919.

FEXX—Q-R-S-T IXTEKJ'AL 443

and C the cycle length in seconds. The original formula now becomes
^ ^^ — V60 ■ To simplify the equation both sides are squared resulting

in S- =■ ^°L or 60C. Extraction of the square root results in S = :^''ii'

Up to this point the method followed was that suggested by Katz ' in
the plotting of his curves.

It now becomes necessary to determine the value of K for the
electrocardiogram. This was done by measuring the systolic and cycle
lengths of a number of electrocardiograms from normal individuals.
The values of S and C being known, the equation becomes K S =
V60C and it appears that 20 is the proper valuation of K. The man-
ner of arriving at this conclusion is shown in Table 1.

TABLE 1. — Manner of Determining Value of K in Formula KS^V60C

^fo.«-

Cycle Length

Q-I^S-T Interval

KS = V«>0

K

»

0.59

0.30

0.30 K = vTET

19.7

12

0.78

0.33

0.33 K = V«.8

20.8

17

0.68

0.32

0.32 K = V40.8

20.0

41

0.88

0.35

0.35 K = V 52.8

20.7

47

0.70

0.32

0.32 K = V 42.0

20.3

56

0.80

0.34

0.34K = vliJ"

20.2

81

0.80

0.36

0.36 K = V«0

19.2

138

0.84

0.36

0.36 K = V50.4

19.8

143

0.68

0.32

0.32 K = v'ioJ"

20.0

lae

0.68

0.33

0.33 K = V40.8

19.5

Since S = ^^j— and the value of K has been determined it is now
possible to predict the normal duration of systole for any heart rate
considered in this paper. The equation, may now be further simplified
and becomes S =Z:iil£or S = .39\/C. The last formula is the one which

20

is used in this study for the prediction of the Q-R-S-T interval.

Bazett,^" in analyzing the relation of systole to cycle length, pro-
jected images of his electrocardiograms on a large sheet of millimeter

10. Bazett, H. C: Heart 7:353, 1920.

444 ARCHirES OF IXTERXAL MEDICIXE

ruled paper. In tliis way very accurate measurements were made
possible. He evolved the formula S^ K\'C. He found that the value
of K was 0.37 for men and 0.40 for women. After the publication of
Bazett's results, the cardiograms used in this series were reexamined
and it was found that appro.ximately one third were taken from women
and two thirds from men.

Considering the fact that this work was done independently and by
a different method from that of Bazett the results correspond exceed-
ingly well.

It is apparent, then, that ventricular systole bears a definite relation
to heart rate or cycle length, and it is evident that this fact must be
considered before speculating on the significance of prolonged Q-R-S-T
intervals. Evidence is here presented to show that unusually long
Q-R-S-T intervals may be well within the calculated limits and the
prolongation only a relative one. Table 2 shows a number of tracings
in which the Q-R-S-T group measures 0.40 second or more and yet
the calculated systole corresponds very closely with the measured one.

A further study of this series of tracings reveals that quite another
condition may be present. The Q-R-S.T interval may have all the
appearance of being normal in its duration but the application of the
formula will show that it is prolonged beyond its predicted length. It
may be within the accepted normal limits yet absolutely prolonged.
Table 3 records a series of cardiograms in which the Q-R-S-T interval
falls well within the usual normal limits. None of these intervals is
prolonged beyond 0.40 second yet each of them is prolonged 0.05 second
or more beyond the predicted length.

In selecting from the studied material those cardiograms in which
the Q-R-S-T interval was definitely prolonged, none were chosen which
did not show a prolongation of 0.05 second or more. The clinical
conditions that seem to stand out preeminently as a cause for the
increase in systolic length are those conditions associated with high
blood pressure. Table 4 shows a number of tracings exhibiting this
prolongation and in all of these cases the blood pressure is well above
the normal.

Meakins " has published a series of cases in which the Q-R-S-T
interval is prolonged and he calls attention to the frequent occurrence
of high blood pressure. In Meakins' work, however, the relation of
the systolic length to heart rate has not been sufficiently emphasized
and it will be found that many of the apparently long systolic intervals
correspond closely to the predicted lengths.

11. Meakins, J.: Arch. Inl. Mccl. 24:-489 (Oct.) 1919.

FE\X—Q-R-S-T IXTERl'AL

TABLE 2.— Long Q-R-S-T Intervals Which Correspond Closely to
Their Predicted Lengths

Heart
1 Rate

Cycle
Length

Q-RS-T Interval

Patient

Measured

Calcu-
lated

Deviation

of
Measured

from
Calculated

Diagnosis

Mrs. N.

8:1

C. B.
T. L.

54
30
BO
30
49

1.12
1.20
1.20
2.0O
1.23

0.42
0.41
0.44
0.55
0.44

0.41
0.43
0.43
0.55
0.43

0.01+
0.02—
0.01 +

0
0.01+

Kormal individual
Normal individual
Diabetes mellitus
Complete heart block
Chronic arthritis

TABLE 3. — Q-R-S-T Interv.\ls Within the Accepted Normal Limit

Which Are Prolonged 0.05 Second or More

Beyond Their Predicted Length

Cycle

Q-R-S-T Interval

E. C. G.

Deviation

Ko.

Bate

Length

Measured

Calcu-
lated

Measured

from
Calculated

Diagnosis

34

83

0.72

53

86

0.70

0.39

0.32

0.07+

Mitral disease

55

7-

0.78

0.40

0.34

0.06+

Hypertension: arterioselero-

SO

M

0.64

0.38

0.31

0.07+

Hypertension; card ac liy-
pertrophy

12.'!

90

0.66

0.38

0.32

0.06+

13.1

107

0.07+

Hypertension; uterine

154

IW

0.60

0.37

0.30

0.07+

Diabetes mellitus

139

0.78

0.39

0.34

0.0-5+

0.30

27W

83

0.72

0.39

0.33

0.06.

Pernicious anemia

TABLE 4. — Q-R-S-T Intervals from Patient? With High Blood Pressure
Showing Prolonged Systole

Cycle

Q

-R-S-T Interval

E.C.G.

No.

Rate

Length

Measured

Calcu-
lated

of

Measured

from

Diagnosis

Calculated

28

80

0.75

0.40

0.34

0.06+

Hypertrnsion: mitral steno-

107

0.56

0.34

• 0.29

Hypertension; chron'c ne
phrltis; S. 210, D. 148

43

0.68

0.38

0.32

0.06+

Hypertension: mitral steno-

55

77

0.78

0.40

0.34

0.06+

Hypertension: senile arterio-
scleros s; S. I'M, D. 110

90

94

0.64

0.38

0.31

0.07 +

Hypertension; artcrioscl to-

1.33

108

0.56

0.36

0.29

0.07+

Hypertension; uterine
flbroid.s S. 200, D. 128

72

0.81

0.41

0.30

0.05+

Hypertension; chronic ne-
phritis: S. 202, D. 120

191

73

0.82

0.40

0.35

003+

Hypertension; diabetes;
.S. 200, D. I.tO

106

83

0.72

0.40

0.33

0 07+

251

82

0.73

0.40

0.33

0.017+

Hypertension; S. 220, D. 10(

In this table S. represents systolic pressure and D. diastolic pressi

446 ARCHIVES OF IXTERXAL MEDICIXE

Bowen ^- was the first to call attention to the augmented systoHc
length in the heart working against increased pressure. In his observa-
tions on normal individuals doing measured amounts of muscular work
he found that as cycle length shortened with increased heart rate, the
systolic length tended to become longer. As this observation was at
variance with any previously reported, he carefully checked his
apparatus to be sure that his results were real rather than apparent.
After assuring himself of the correctness of his observations, he
explained this phenomenon by comparing the heart to a simple pumping
engine which slows when it meets an increased resistance. The rise
in blood pressure which accompanies the beginning of muscular work
furnished the increased resistance against which the heart must work.

More recently Patterson, Piper and Starling,^^ working with heart
and lung preparations, have shown that heart volume, intracardiac
pressure, and systolic length all increase to meet an increased arterial
resistance. In their work it was possible to control all features so that
the diastolic inflow could be kept at a constant level and by increasing
the arterial resistance it was shown that the heart dilates and, in this
process, the individual muscle fibers become lengthened. If the law
of the heart muscle corresponds to that of skeletal muscle, that the
energy set free on contraction depends on the initial length of the
muscle fibers, then greater contractile stress and prolonged systole
should result. This was found to occur in the work of Patterson,
Piper and Starling.

These experiments serve very well to explan the prolonged
Q-R-S-T interval in the patients in this study exhibiting high blood
pressure. They also serve to show that ventricular systole measured
by the electrocardiogram may be favorably compared to that measured
by mechanical means.

It was observed, however, in this study, that many of the patients
whose Q-R-S-T interval is not prolonged exhibit greatly increased
blood pressures. Table 5 records a series of cases in which the blood
pressure is high and yet the measured and predicted systolic lengths
closely correspond.

A consideration of Tables 4 and 5 demonstrates clearly that factors
other than a simple increase in the arterial resistance must operate
to produce a prolongation of the Q-R-S-T interval.

In attempting to explain the lack of similarity in the behavior of
difTerent individuals, consideration must be given to the work of
Wiggers " dealing with the relation of systolic length to heart rate,
diastolic inflow and arterial resistance. lie states that the duration of

12. Bowcn. W. P.: Am. J. Physiol. 11:59, 1904.

13. Patterson, S. W.; Piper, H., and Starling, E. H. : J. Physiol. 48:465, 1914.

14. Wiggers, C. J.: Am. J. Physiol. 56:439, 1921.

FEXX—Q-R-S-T IXTERJ-AL 447

systole is prolonged by increased venous inflow, whether or not this
increase is accompanied by a rise in arterial resistance. He found
also that increase in the arterial resistance caused prolongation of
systole only when the pressure causing the resistance was applied on
the aorta near the semilunar valves. When the resistance was met with
in the peripheral circulation or even in the abdominal aorta no lengthen-
ing of systole took place ; indeed, the systolic length tended to become
shorter. He concluded, further, that the duration of systole at a con-
stant heart rate was dependent on the initial pressure in the ventricle
rather than on the initial length of its fibers.

TABLE 3. — Q-R-S-T Intervals from Patients W'nn High Blood Pressure
Showing no Prolongation of Systole

He.nrt

Cycle

Q-R-S-T Interral

E, C. G.

DeTiatioD

No.

Bate

Length

Measured

»

of
Measured

from
Calculated

Diagnosis

1

77

0.78

0.34

0.34

0

Mitral and aortic disease;
S. 202, D. 120

17

88

0.68

0.32

0.32

0

Mitral und aortic disease:
S. 182, D. 76

38

70

0.88

0.37

0.36

0.01 -f

Myocarditis; renal disease;
S. 200, J>. 140

60

84

0.72

0.34

0.33

0.01 -f

Hypertension: S. 202, D. 110

85

111

OM

0.28

0.28

0

Acute alcoholism; hyperten-
sion; S. 162, D. 88

103

117

0.52

0.28

0.28

0

Abscess of nose; S. 180. D. 112

126

96

0.62

0.31

0.31

0

Aortic aneurysm; S. 160, D. 70

151

0.96

0.38

0.38

Hypertension: S. 250, D. 120

152

81

0.74

0.34

0.34

0

Hypertension: chronic ne-
phritis: S. 238. D. 110

208

73

0.82

0.36

0.35

o.on-

Aortic disease; S. 182, D. 110

In this table S. represents systolic pressure and D. diastolic pressure.

In Striving to apply these conclusions to the results obtained in the
human electrocardiogram one must enter largely into the fields of
speculation. It is obviously impossible to measure venous inflow and,
while increased venous pressure may be estimated easily enough, an
increase in the venous pressure does not necessarily carry with it an
increased venous inflow into the ventricle. It would also be very
difficult to determine the location of the etiologic factor in the produc-
tion of an increased arterial resistance.

In spite of these difficulties it does not appear that increased initial
tension in the ventricle will account for the prolongation of the Q-R-
S-T interval in these patients. All of the patients, whose records
appear in this report, had been carrying their vascular overload for
a considerable period of time and they were subjected to no unusual
stress at the time their cardiograms were made. Most of them, in
fact, had been at rest for some time. Under these conditions it is
improbable that there should be any increase in the initial tension in the
ventricle for that particular heart.

448 ARCHIVES OF INTERXAL MEDICIXE

W'iggers does not state whether the heart volume failed to increase
when the pressure causing increased arterial resistance was applied
at some distance from the aortic valves, but it is possible that heart
volume would not increase if the pressure were applied for a com-
paratively short time.

In such an event the intervening arterial system would take care
of the increased resistance so that there would be no increase in initial
tension in the ventricle, as Wiggers states, and neither would there be
increase in initial length of its fibers.

If, however, the factor, producing the incieased resistance, con-
tinued to act over a long period of time, such as it must in the cases
of chronic hypertension reported here, sooner or later the increased
tension must make itself felt in the ventricle and with the increase in
intraventricular pressure comes an augmented initial length of its
fibers.

The conclusion is forced, therefore, that the duration of systole
in hypertension is a measure of the initial length of the ventricular
fibers and, to such an extent, a measure of its muscular dilatation.

It is possible to determine the effect on the Q-R-S-T interval of
rapidly rising or falling blood pressures. Such a series of cases is
under observation at the present time and will be reported on later.

It is probable that the variation in the behavior of different indi-
viduals is of prognostic value but these cases have not been under
observation a sufficiently long time to venture a definite opinion on the
prognostic significance of the Q-R-S-T interval. This will be reported
on later.

SU.MMARY AND CONCLU.SIONS

1. It is possible to predict with a reasonable degree of accuracy
the duration of systole in normal individuals.

2. Clinical conditions accompanied by high blood pressure are often
associated with prolongation of the Q-R-S-T interval.

3. It is probable that this prolongation is of prognostic value, but
from the data at hand at present a definite statement may not be made.

POSTOPERATIVE PULMONARY CO.MPLICATIONS *
ELLIOTT C. CUTLER. M.D.. and ALICE M. HUNT. R.N.

BOSTON
INTRODUCTION

It is with some hesitation that we again present this subject.' That
there remains, however, considerable divergence in the minds of both
surgeons and anesthetists as to the etiologic factors, and therefore, the
prevention and treatment of pulmonary complications there is no doubt.
The retiring president of the American Association of Anesthetists
stated this Spring that only 3 per cent, of anesthetists considered such
complications as embolic in origin and in the most recent textbook of
medicine • aspiration is put forward as the chief cause of postoperative
pneumonia. Previous studies had led us to believe that the majority
of these lesions are infarcts and an analysis of the cases subjected to
operation in this clinic during the past year brings additional evi-
dence to support this view. The recent literature contains many excel-
lent papers in agreement with this opinion, and we feel that it is
unfortunate that anesthetists and anesthesia should bear the blame for
such complications when the facts would seem to exonerate both, in
the majority of cases.

The cases subjected to operation in this hospital during 1920 will
form the basis for this report. Statistical reports are unquestionably
open to criticism, the use of hospital records possibly more so than in
other fields. For no matter how meticulously kept, when studied from
the viewpoint of a researcher, much will be found missing. Morfeover,
let alone the fact that many of the observations are made by another
worker, they are often inexact. This is an example of the practical
impossibility of cooperative research. It may be said, however, that our
previous interest in this subject ' has made it possible to obviate some
of the inaccuracies that can occur. In this clinic all patients are con-
stantly watched for the appearance of such complications and the roent-
gen ray is generously used as a control. Moreover, when such a lesion
does occur an additional diagnosis card is filed away for future refer-
ence so that all proven cases are easily available for study when desired.

As might be expected, this arrangement has enabled us to assemble
a steadily increasing percentage of cases. Whereas in 1916 of 3,490
cases we ' were able to identify only sixty-five with pulmonary compli-

* From the Surgical Clinic. Peter Bent Brigham Hospital.

1. a. Culter, E. C, and Morton, J. J.: Postoperative Pulmonary Complica-
tions, Surg., Gynec. & Obst. 25:621, 1917. b. Cutler, E. C, and Hunt, A. M.:
Postoperative Pulmonary Complications, Arch. Surg. 1:114 (Jan.) 1920.

2. Lord. F. T. : Posto|>erative Pneumonia, Nelson's Loose Leaf System Med.
1:296, 1920.

450 ARCHIVES OF INTERNAL MEDICINE

cations, during 1920 we found sixty-three cases among 1,604 cases, a
difference in respective morbidity of from 1.86 to 3.92 per cent. This
increasing morbidity is comparative to contemporary studies. McKesson ^
reported 3.03 per cent, morbidity among 39,438 collected cases in 1918.
That the increasing morbidity figure is due to better records is further
emphasized by the comparative drop in the mortality percentage.* In
1916 our cases showed a mortality percentage of morbidity of 50.7 as
opposed to our present figures of 7.9. The cases studied in this report
give an approximate morbidity of 1 in every 25 cases and an approxi-
mate mortality of 1 death in every 320 cases. Although this indicates
a great decrease, these figures should still cause concern.

As in previous reports, patients with bulbar palsy or a terminal
pneumonia, in association with definitely serious or fatal primary
lesions, are excluded. Otherwise, all patients submitted to some form
of operative procedure under anesthesia are included. The entire field
of pulmonarj' complications is again studied since it appears that such
studies are of greater value in that the etiology is often the same in the
many varied clinical lesions and the clinical classification is difficult
owing to a confusion in the signs and symptoms presented. Any
attempted comparative study of statistical reports is open to the further
criticism that in no two clinics are the conditions the same. That is
particularly true of this very subject, and the report of a clinic such as
this, in which the majority of patients are from the poorer classes of a
large city, is not suitable for accurate comparison with a clinic such as
that at Rochester, Minn.' Again, no comparison should be made with
special clinics, as those given over to gynecology." In a search, how-
ever, for the underlying pathologj', all the material can be used since it
is probable that no matter what the type of case or operation the
mechanism resulting in pulmonary complications is the same.

The result of the present study appears to corroborate a previous
report of ours^ in which embolism from the operative field seemed to
be the chief cause of pulmonary complications. In the present report
each case was carefully scrutinized with a view to determining the

3. McKesson, E. I. : Some Observations on Postoperative Lung Complica-
tions, Am. J. Surg. 32:16. 1918 (Quart, suppl. Anesth.).

4. The comparative studies at the Presbyterian Hospital with figures for
1898 (75), 1916 (81) and 1917 (82) show the same tendency to an increased
morbidity and decreased mortality due to more accurate records. In 1898
Schultze reported 0.38 per cent, postoperative pneumonias, whereas in 1916
Whipple reported 2.6 per cent, of this complication.

5. a. Beckman, E. H. : Pulmonary and Circulatory Complications Following
Surgical Operations, Mayo Clinic Papers, Philadelphia and London, W. B.
Saunders Co., 1910, p. 594. b. Complications Following Surgical Operations,
1912, p. 738. c. 1913, p. 776.

6. Pfanncnstiel, I.: Ucbcr die Vorzuge der Athernarkose, Zcntralbl. f.
Gynak. 27:8, 1903.

CUTLER-HUXT—POSTOPERATIIE LUXG LESIOXS 451

etiology of the lesion at the time of its occurence. Roentgen ray
studies confirm the t>-pe and extent of the majority of these lesions.

REVIEW OF LITERATURE

Except for recent reports the literature pertaining to this field was
fully covered in previous studies.' At that time reports covering the
whole field of pulmonary complications were few although there were
many excellent studies concerning the compHcations following lapa-
rotomy and the incidence of postoperative pneumonia alone. Table 1
includes the comparative figures for total pulmonary- complications in
the various clinics up to the present time.

TABLE 1. — Total Pulmonary Complications

Pulmonary Pulmonary

Morbidity Mortality Mortality

, * \ f *- — -^ per Cent.

Per Per of

No. Cent. Xo. Cent. Morbidity

No. of

Clinic

Author and Tear

Cases

Montreal General Hos-

pital

Armstrong.' 1906. ,

2,500

Combined statistics...

Von Llchtenb€rg,»

1908...

23,673

Mayo Clinic*

Beckman.«'1912...

5,635

Beckman '« 1913 ..

6,825

Massachusetts General

Cutler and Morton,
McKesson.' 1918...

,'■ 1917

3,490

39,438

Peter Bent Brigham

Hospital

Cutler and Hunt,'"

1920...

1,.'«2

Decker," 1921

5 9T6

Peter Bent Brigham

Hospital

* Amended to include pulmonary embolism.

The chronologic order of the data in this table demonstrates
graphically the increasing morbidity with its concomitant decrease in
mortality. The figures of Decker alone disagree. The reason for this
appears to be a failure to include minor complications. It is indeed
almost impossible to arrive at accurate data unless each case is studied
from this viewpoint and checked, at the time of occurrence, for future
reference.

The discussion of the etiology of these complications has brought
to light two opposing views, — one, that the anesthetic and the other that
embolism plays the chief role. That inhalation anesthesia produces
some irritation, and that the aspiration of mouth contents occurs there

7. Armstrong, G. E. : Remarks on Lung Complications After Operations
with Anesthesia, Brit. M. J. 1:1141. 1906.

"8. Ranzi, E. : Ueber postoperative Lungcnkomplikationen embolisher Natur,
Arch. f. klin. Chir. 87:380, 1908.

9. Lichtenberg, A.: Die postoperativen Lungcnkomplikationen, Centralbl.
f. d. Grcnzgcb. d. Med. u. Chir. 11:129, 1908.

10. Decker, H. R. : Postoperative Complications and Sequels of the Res-
piratory Tract, Penn. M. J. 24:391, 1921.

452 ARCHirES OF IXTERXAL MEDICIXE

can be no doubt (Hoelscher " and Kelly ^-). That a perfectly smooth
inhalation anesthesia reduces such undesirable sequels is vigorously
upheld (Poppert,^' Offergeld," von Lichtenberg,^' Ladd and, Osgood."
Magaw,^' Bevan/* Kroenlein,^'' Keen,-" Herb,-^ Henderson--). This,
however, fails to explain why, with local anesthesia, the proportion of
such complications is equally high (Gottstein,-^ Mikulicz,''* Henle ^^
and Sauerbruch -"). It also fails to explain why wfth anesthesia in
expert hands these complications continue to occur.-'

The statement - that aspiration is the chief cause of postoperative
pneumonia is not proven by the facts. There is an increasing mass of
evidence demonstrating the frequency of postoperative embolism and
its relation to pulmonary complications. W. J. Mayo ^^ states that

11. Hoelscher, R. : Experimentelle Untersuchungen iiber die Entstehung der
Erkrankungen der Luftwege nach Aethernarkose. .Arch. f. kltn. Chir. 57:175.
1898.

12. Kelly. R. E. : Anesthesia by the Intratracheal Insufflation of Ether,
Brit. M. J. 2:112, 617. 1912.

13. Poppert : Experimentelle und klinische Beitrage zur Aethernarkose und
zur Aetherchloroform Mischnarkose. Deutsch. Ztschr. f. Chir. 67:505, 1902.

14. Offergeld: Lungenkoinplikationen nach Aethernarkosen, Arch. f. kliii.
Chir. 83:505, 1907.

15. Lichtenberg. A.: Experimenteller Beitrag zur Frage der Entstehung der
Pneumonie nach Narkosen. Miinchen. med. Wchnschr. 53:2286, 1906.

16. Ladd, W. E.. and Osgood. G. : Gauze-Ether, or a Modified Drop Method,
with Its Effects on Acetonuria. Ann. Surg. ■16:460, 1907.

17. Magaw. A. : A Review of Over Fourteen Thousand Surgical Anes-
thesias. Surg., Gynec. & Obst. 3:795. 1906.

18. Bevan, A. D. : The Choice and Technic of the .Anesthetic, Tr. Am. Surg
Assn. 33:21. 1915; 29:177, 1911.

19. Kroenlein, R. V.: Discussion, Verhandl. d. deutsch. Gesellsch. f. Chir.
34:1.31. 1905.

20. Keen. W. W.: The Dangers of Ether as an Anesthetic, Boston M. &
S. J. 173:831, 1915.

21. Herb. I.: Ether: Simplicity in Its Administration, J. A. M. A. 66:1376
(April 29) 1916.

22. Henderson. F. : Ether Anesthesia, Collected Papers Mayo Clinic. Phila-
delphia and London, W. B. Saunders Co., 1913, p. 701.

23. Gottstein. G. : Erfahrungen fiber lokale Anasthesie in der Breslauer
Chirurgischen Klinik. Arch. f. klin. Chir. 57:409. 1898.

24. Mikulicz. T.: Die Methoden der Schmerzljetaubung und ihre gegenseitige
Abgrenzung. Verhandl. d. deutscli. Gesellsch. f. Chir. 30:560. 1901.

25. Hcnle: Die Methoden der Schmerzbetaubung und ihre gegenseitige
Abgrenzung. Verhandl. d. deutsch. Gesellsch. f. Chir. 30:240. 1901.

26. Sauerbruch. F. : Der Stand der Klinischen und Operativen Chirurgie,
Beitr. z. klin. Chir. 122:234, 1921.

27. This statement will be objected to by many. Only let these consider that
in most large institutions and especially in those from which reports come,
anesthesia is in really expert hands now. Yet these complications continue.
.Mso let them take into consideration that a betterment of operative technic
has also taken place and consider that a reduction in such complications may
be due to this factor quite as well as to the improvements in the administra-
tion of anesthesia.

28. Mayo, W. J.: Mortality and End Results in Surgery, Surg., Gvnec. &
Obst. 32:97, 1921,

CLTLER-HU\T— POSTOPERATIVE LUXG LESIOXS 453

minute septic emboli from the operative field are a common cause of
secondan' pulmonan- complications. He laments that these complica-
tions are tod frequently attributed to the anesthetic, and says that they
are quite as frequently found in cases in which local anesthesia is
employed. The recent illuminating papers of Ochsner and Schneider,-"
Hampton and Wharton,^" Capelle,^^ McCann ^= and Rupp ^^ demon-
strate both the mechanism of thrombosis and embolism subsequent to
operation and the great frequency of this condition as a source of pul-
monary complications. These ideas are by no means of recent origin,
and a study of the papers by Zahn,^*, Miller,^= Gebele,^'' Otte,^" Zur-
helle,^* Honians,^*" Michaelis,*" Henderson," Grant *' and Eisenreich "
reveals statistics and studies in support of this view. Zahn's paper,
written in 1897, discusses the part passive congestion in the lung during
operation may have as increasing the liability to infarction. In addi-
tion to such direct contributions are the many excellent reports both
clinical ** and experimental on fatal pulmonary embolism and pul-
monary infarction. The ex])crimental and pathologic studies of pul-

29. Ochsner, A. J., and Schneider, C. C. : Fatal Postoperative Puhiionary
Thrombosis, .'\nn. Surg. 72:91, 1920.

30. Hampton, H., and Wharton, L. R. : Venous Thrombosis, Pulmonary
hifarction and Embolism Following Gynecological Operations, Bull. Johns
Hopkins Hosp. 31:95, 1920.

31. Capelle: Einiges zur Frage der Postoperativen Thromboembolie, Bcitr.
z. klin. Chir. 119:485, 1920.

32. McCann, F.: Suggestions £01* the Prevention of Postoperative Throm-
bosis and Embolism, Brit. M. J. 1:277, 1918.

33. Rupp, A.: Postoperative Thrombose und Lungenembolie, Arch. f.,k!in.
Chir. 115:672 (March) 1921.

,34. Zahn, F. W. : Ueber die Folgen des Verschliissens der Lungenarterien
und Pfortaderaste durch Embolie, Verhandl. d. Gesellsch. deutsch. Naturforsch.
u. Aerzte 19:9, 1897.

35. Miller. R. B. : The Significance of Postoperative Pleurisy: Its Relation
to Pulmonary Embolism, Am. Med. 4:173, 1902.

36. Gebele : Ueber Embolische Lungcn .^ffektioncn nach Baucboperationen ;
Eine klinisch-experimentelle Studie, Beitr. z. klin. Chir. 43:251, 1904.

37. Otte, A.: Ueber die postoperativen Lungcnkomplikationen und Throm-
bosen nach .Aethernarkosen. Miinchen. med. Wchnschr. 54:2473, 1907,

38. Zurhelle, E. : Thrombose und Embolie nach Gynakologischen Operationen,
Arch. f. Gynak. 84:443, 1908.

39. Homans, J. : Postoperative Pulmonary Complications, Bull. Johns Hop-
kins Hosp. 20:128. 1909.

40. Michaelis: Zur Frage des Praeinonitorischen Symptoms von Thrombose
und Embolie, Ztschr. f. Geburtsh. u. Gynak. 70:285, 1912.

41. Henderson, F.: St. Paul M. J. 16:74, 1914.

42. Grant, H. H. : Thrombophlebitis and Pulmonary Embolism, Mississippi
Valley M. J. 30:217, 1918.

43. Eisenreich. O. : Embolism After Gynecologic Operations, Monatschr. f
Geburtsh. u. Gynak. 53:190. 1920.

44. Wilson. L. B. : Fatal Postoperative Embolism, Collected Papers Mayo
Clinic, Philadelphia, W. B. Saunders & Co.. 1912, p. 727.

454 ARCHIVES OF INTERNAL MEDICINE

nionary infarction are well covered by Welch,*^ McCallum,*" Karsner
and Ash " and Karsner and Austin.*^ That these postoperative lesions
are usually situated in the lower lobes, more frequently on the right side
than on the left, and thus comparable to pathologic studies on embolism
and infarction may be carried as an additional argument for this mecha-
nism. Indeed, to one familiar with the literature it is difficult to accept
aspiration as the chief factor in such complications unless infarction
can be disproved. The papers referred to present excellent discussions
of the clinical signs and suggest prophylactic measures. With these
ideas both our studies of 1919 and 1920 seem to agree and bring addi-
tional proof.

Etiologic factors other than the two cited may be considered of
secondary importance. A focus of pulmonary disease existing at the
time of operation unquestionably is the cause, at times, of a subsequent
complication. The flare-up of a quiescent tubercular lesion or of an old
bronchitis under inhalation anesthesia is unquestioned. Such lesions,
however, do not always light up, and the fact that equally following
operations under local anesthesia activity may be stirred up suggests
the possibility of embolism as the direct cause even in such cases. The
fact that the embolus reaches an already injured field makes its influ-
ence the more marked.

The importance of sepsis as a secondary factor cannot be denied,
and there are cases with septic abdominal lesions in which, following
operation, the lung is involved by direct extension. That a lymphatic
avenue is open for such extension has been shown by the anatomic
studies of Sabin," Cunningham^" and Miller.^^ The importance of
sepsis both in the initiation of and the setting free of thrombi is an old
story. The danger which it adds to embolism is manifest. Many cases
of lung abscess following tonsillectomy usually have been thought to be
due to the inhalation of mouth infectious material owing to position.

45. Welch. W. H. : An Area of Coagulative Necrosis Resulting from Shut-
ting Off of the Blood Supply in an Infarct, Papers and Addresses, Baltimore,
The Johi>s Hopkins Press 1:110, 1920.

46. McCallum, W. P.: Infarction, A Textbook of Patholog>-, Ed. 2, Phila-
delphia and London, W. B. Saunders & Co., 1920, p. 3i.

47. Karsner and Ash : Studies in Infarction, Experimental Bland Infarc-
tion of the Lung, J. M. Research 22:1912-1913.

48. Karsner and Austin: Studies in Infarction. T. .\. M. A. 57:951 (Oct.
10) 1911.

49. Sabin, F. R. : The Method of Growth of the Lymphatic System, Science,
New York 44:145, 1916.

50. Cunningham. R. S. : On the Development of the Lymphatics in the Lungs
of the Pig, Proc. Am. Assn. Anat., Anat. Rec. 9:69, 1915.

51. Miller, W. S.: Some Essential Points in the Anatomy of the Lung, Am.
J Roentgenol. 4:269, 1917. The Vascular Supply of the Pleura Pulmonalis,
Am. J. Anat. 7:389, 1908.

CUTLER-HUKT— POSTOPERATIVE LUXG LESIOXS 455

The recent report of W. B. Porter '^ concerning such lesions following
local anesthesia, and his suggestion of the logical embolic route is of
interest in this relation. And Burnham's ^^ paper on pleurisy and
empyema evidences the importance of a septic focus. In all his cases
in which empyema developed the original operation was for a septic
abdominal lesion.

Acidosis has been considered as having an influence in the produc-
tion of these lesions. There has been, however, no satisfactory expla-
nation of the mechanism by which this produces its effect, and some
investigators ^* have denied its existence. It is possible that where
asphyxia is produced there is some acidosis, and thereby the natural
resistance to infection is lowered. Whether this is due to the anesthetic ^^
or to the effects of the operation '^ is unknown.

Chilling and cold possibly produce some effect in a similarly remote
manner. The work of Miller and Noble,^" Hill ■'" and Mudd and Grant ^'
would appear to show that there is some lowered resistance to infection
with chilling. Moreover, such clinical investigators as Boothby,""
Homans,'' Keen,"" Armstrong,' Gerulanos,^^ and Decker^" have felt
that either external exposure or chilling by cold wet packs during opera-
tion played a part in the subsequent pulmonary complications.

These factors, by the production of a lowered immunity, may open
the body to attack by organisms commonly present. That Group IV
pneumococcus is commonly found in these cases (Whipple ^^ and Cleve-

52. Porter, W. B. : Pulmonary .-Xhscess Following Tonsillectomy Under Local
Anesthesia, Virginia M. Month. 47:606 (March) 1921.

53. Burnham, A. C. : Postoperative Pleurisy with Effusion and Empyema,
Surg., Gynec. & Obst. 19:468, 1914.

54. Caldwell, G. A., and Cleveland, M. : Observations on the Relation of
Acidosis to Anesthesia, Surg., Gynec. & Obst. 25:22. 1917.

55. Carter, W. S. : Effect of Ether on the Alkali Reserve, .'\rch. Int. Med.
26:319 (Sept.) 1920. Collip, I. B.: Effect of Surgical Anesthesia on the
Reaction of the Blood, Brit. J. E.xper. Path. 1:282, 1920. Jeanbrau, E., Cristol,
P., and Bonnet, V.: Anesthesia and Acidosis, J. d'urol. med. et. chir. 11:505
(May-June) 1921.

56. Farrar, L. K. P.: Acidosis in Operative Surgery; Occurrence During
Operation and Its Treatment by Glucose and Gum Acacia Given Intravenously,
Surg., Gynec. & Obst. 32:328. 1921.

57. Miller, J. A., and Noble, W. C: The Effects of Exposure to Cold on
Experimental Infection of the Respiratory Tract, J. Exper. M. 24:223, 1916.

58. Hill: The Influence of the Atmospheric Environment on the Respira-
tory Membrane, Brit. M. Research Committee, Ser. 32, p. 141, 1919.

59. Mudd, S., and Grant, S. B.: Reactions to Chilling of the Body Surface,
J. M. Research 40:5.3. 1919.

60. Boothby, W. M. : Postoperative Pneumonia (Discussion) J. A. M. A.
67:539 (Aug. 12) 1916.

61. Gcrulanos, M. : Lungen Komplikationen nach Operativen Emgnffen,
Deutsch. Ztschr. f. Qiir. 57:371, 1898.

62. Whipple, A. O. : A Study of postoperative Pneumonia in the Presby-
terian Hospital During 1915, Med. Rec. 89:581, 1916. A Study of Postoperative
Pneumonitis, Surg., Gynec. & Obst. 26:29, 1918.

456 ARCHIVES OF IXTERXAL MEDICINE

land'^^), might seem to be additional evidence that this mechanism
existed. It should, however, be recognized that Group IV pneumo-
coccus is commonly found in normal mouths (Stillman," Dochez and
Cole"'^). Provided, that a focus of pulmonary disease already existed,
this lowered resistance might enable organisms present to get a start
and thus a more virulent mixed type of infection might occur.'"'

It is difficult, however, to appreciate the importance of such sec-
ondary factors, and also that aspiration is the chief factor, when it is
so well known that the experimental production of pneumonia by the
bronchial route is extremely difficult. Blake and Cecil's ^' studies
alone seem to be successful by this method, except when enormous
dosage has been used.

The foregoing facts may be taken as a defense of inhalation
anesthesia and anesthetists. It is astonishing that so few of the latter
have attempted their own defense when the general concensus of
opinion now definitely favors embolism as the chief factor in the
production of these lesions. The earlier reports on "ether pneumonia"
cast a most unjust opprobrium on the anesthetic. That venous throm-
bosis, fatal pulmonary embolism and pulmonary infarction are common
there can be no doubt. McCallum " and Welch *" have given admirable
descriptions of the underlying pathology. The right lower lobe is most
frequently the site of both these clinical lesions and the pathologic types
recorded. Moreover, infarction can be the, only explanation in cases
in which local anesthesia is used. Why search for additional factors
when the pathologic and clinical studies emphasize the importance and
frequency of embolism, and when the clinical picture is far more fre-
quently of the type ascribable to this factor?

As is shown in the cited reports, and as the cases presented later
show, the onset in the majority of these cases is abrupt, dissimilar to
the onset of infection elsewhere, and the subsidence, except when the
embolus has arisen in a septic field or landed in a focus of pre-existing
pulmonary disease, is almost equally rapid. Moreover, the symptom
of pain, the small areas involved and the typical lobular distribution of

63. Cleveland, M. : Further Study in Postoperative Pneumonitis, Surg.,
Gynec. & Obst. 28:282. 1919.

64. Stillman, E. G. : A Contribution to the Epidemiology of Lobar Pneu-
monia, J. Exper. M. 24:651, 1916.

65. Dochez. A. R., and Cole, R. I.: Pneumococcus Infection, Forchheimer"s
Therapeusis of Internal Diseases. New York and London. D. Appleton & Co.,
5:472, 1914.

66. Wadsworth, A. B. : Experimental Studies on the Etiology of .\cute Pneu-
monitis. Am. J. M. Sc. 127:851. 1904. A Study of Experimental Organizing
Pneumonia, J. M. Research 34:147. 1918.

67. Blake, F. G., and Cecil. R. L. : Production of Pneumococcus Pneumonia
in Monkeys. J. Exper. M. 31:403, 1920. Pathology and Pathogenesis of Pneu-
mococcus Lobar Pneumonia in Monkeys. J. Exper. M. 31:445. 1920.

CUTLER-HUXT— POSTOPERATIVE LUXG LESIOXS 457

the lesion, with often cone-shaped roentgen-ray shadows, is so striking
as to leave little doubt as to the kind of lesion present.

PRESENTATION OF MATERIAL

From Jan. 1, 1920, to Jan. 1, 1921, 1,604 cases "« in this clinic were
submitted to some form of operative procedure under anesthesia.
Sixty-three of these patients developed a pulmonary complication that
might be attributable to the operative intervention or to the anesthetic.
There were five deaths among these sixty-three cases. In addition five
other patients died with pulmonary lesions following operation, but
three of these suffered from bulbar paralysis due to intracranial lesions
and two were old men, admitted "in extremis," with septic genito-
urinary lesions, who, after some simple emergency intervention devel-
oped septicemia with its concomitant terminal pneumonia. It appears
only fair to remove such cases from this discussion. Thus in 3.93 per
cent, of the cases (one in twenty-six) a complication developed and 0.3
per cent. (1 in 321) patients died from one of the complications. -

The conception that small emboli from the operative field are the
chief etiologic factors makes the classification simpler. All cases in
which this mode of occurrence was obvious can be placed in a distinct
class. We have elected to term these, cases of pulmonary infarction.
The further divisions are, lobar and bronchopneumonia, bronchitis,
pleurisy, pulmonary embolism, empyema, lung abscess and cases in
which there has been exacerbation of a tuberculous lesion.

We found that unless a separate division was made for the embolic
cases the greatest difficulty in classification arose. Thus many lesions,
in which this mechanism was obvious, simulated partly pleurisy, there
being pain and a friction rub, and partly pneumonia, since a small
patch of consolidation was demonstrable both on physical examination
and by the roentgen ray. In this report, therefore, we have placed in
the class termed infarction all those cases in which the mechanism of
embolism was undoubted (excepting the typical fatal pulmonary
embolism cases). Ranzi * called these small embolic lesions embolic
pneumonia, a somewhat confusing pathologic term. In the other classes
we feel there must be also some cases in which embolism was present,
but either the presence of sepsis in the wound or a preexisting pul-
monary lesion so confused the picture that classificatidu as infarction
was not justifiable.

Pathologists regard as jjulmonary infarction only those lesions that
go on to coagulative necrosis. Thus they may object to this nomen-
clature. The mechanism of these minor infarctions, however, is the

68. This does not include cystoscopies, dressings and the apphcations of
plaster cast unless anesthesia was employed.

458 ARCHIVES OF IXTERXAL MEDICIXE

same as with larger coagulative necrotic lesions and certainly both in
etiology and pathology the lesions are more nearly infarction than true
pneumonia. Observers in the past, including ourselves, by classifying
such lesions as pneumonia have certainly not clarified the issue. It is
our hope that the submitted evidence in regards to such lesions will lead
others to adopt the same classification. If this is followed it would
appear that a better understanding of pulmonar)' complications as a
whole must result.

TABLE 2.— Morbidity .\nd

Mortality of

\'.\RIOUS

Complications

Complications

No. of

16
2
2

32

Mortality

Percen-

Morbidity No. of tage of

Percentage Deaths Morbidity

0.06 1 lOO

0.43 1 14.2

0.99 0 0

0.12 0 0

0.12 0 0

0.06 1 lOO

1.99 0 0

0.12 2 100

3.92 5 7.93

Approxi-

Mo°bid1ty
per 100
Cases
l:16W
1:229
1:100
1:804
1:804
l:l(K>4
1:50
l:8W

1:25

Approxi-

MOTtality

■'c^aslT
1-1604

Bronchopneumonia.:

Bronchitis

Exacerbation of tuberculosis

Pleurisy

1:1604
0

0
0

0

Totals

1-320

Table 2 depicts the great preponderance of infarction cases. The
freedom of this group from fatalities will be discussed later. It is clear,
however, that in these cases death is due either to a large pulmonary
embolism (a separate class) or to septic lesions. Here the confusion in
the picture may place the case in the pneumonia or bronchitis class.

The infrequency of true lobar pneumonia is striking and further
emphasizes the general tendency in the more accurate recent studies
wherein lobar pneumonia is shown gradually to be disappearing as a
postoperative lesion. This is especially true when the clinical findings
have been verified by roentgen-ray studies. The reason for this lies
in the mechanism of its production, i. e., embolism rather than infection.

Lobar Pneumonia. — There was only one case this year in which the
complication was of the lobar pneumonia type. It is only fair to state
that without the roentgen-ray studies we should have misplaced even this
one case among the bronchopneumonias.

History. — Patient, SO years of age, was operated on for an indirect left
inguinal hernia under procain. Local anesthesia was used since the past his-
tory and physical examination indicated an already damaged lung. Three years
previously the patient had suffered from a severe attack of pneumonia and
ever since had been much troubled with asthma and frequent attacks of pleurisy.
Moreover, physical examination of the lungs demonstrated relative dulness in
the right upper lobe and many sonorous and musical rales throughout the
entire right chest. Within two days following operation tliere was evident
difficulty in breathing and some productive coughing. This increased and the

CUTLER-HUST— POSTOPERATIVE LUXG LESIOXS 459

temperature and respiratory rate began to climb. By the fourth day post-
operative there was a demonstrable area of "consolidation in the right lower
chest. Roentgen-ray studies on the eleventh day showed that all three lobes
were uniformly involved, another roentgen-ray examination made on the
eighteenth day showed beginning resolution in the upper and lower lobes but
no change in the middle lobe. Shortly following this the urinary output
diminished and the patient's general condition rapidly deteriorated. He died
on the thirtieth day after operation.

This case is presented in some detail because it is both typical and
atypical ; typical in that a pre-existing focus of pulmonary disease flared
up into a serious and fatal complication ; atypical in that the lesion was
lobar rather than lobular. Indeed, were it not for the excellent roent-
gen-ray studies, we had sufficient other clinical evidence to designate it
a bronchopneumonia. The explanation of the fatality must lie in
some connection with the previous attack of pneumonia, the patient's
individual susceptibility and the possibility that he still harbored a
malignant organism. One may suppose, however, that in the absence
of inhalation irritants, emboli from the field of operation may have
been the means of stirring into activity an already prepared focus of
disease.

Bronchopneumonia. — We have placed seven cases in this group.
All cases showed multiple areas of consolidation at some period follow-
ing operation. There was one fatality. The anesthesia was gas-oxygen
in four cases and ether in three. In two of the gas-oxygen cases, a
little ether was given for added relaxation while opening the abdomen.
In the straight ether cases, the open mask method was used twice and
the Connell machine with an intranasal pharyngeal tube once. The
operations were : laparotomy, six times, three times for appendicitis and
once each for gastric ulcer, salpingitis and hydronephrosis, and
craniotomy for extracerebellar acoustic neuroma once. The latter case
ended fatally. The well known relative predominance of such com-
plications with abdominal operations is thus exemplified.

The cranial case was of unusual interest to us in that we have
rarely observed pulmonary complications in such cases, except when
there has been ninth nerve involvement, which apparently was not
present in this case. However, following a long, tedious, suboccipital
operation this nerve may, to some extent, have been injured. The patient
was 47, a good risk and did well following operation until the seventh
day when without previous intimation there was a sudden rise in pulse,
temperature and respiration. Bilateral basal bronchopneumonia prog-
ressed steadily until death on the twenty-fourth day postoperative. The
sudden onset, although without pain, suggests the possibility of embolic
origin of the complication. The large vessels encountered in such
explorations are an obvious source of such thrombi.

460 ARCHIVES OF ISTERXAL MEDICI SE

The other six cases have both sepsis and abdominal section in com-
mon, a well recognized dangerous combination.^^ The three appendix
cases were all acute either with abscess or perforation, the case of
salpingitis was equally acute, a large pyosalpinx having ruptured and
given the signs of a general peritonitis, and the gastric ulcer case with a
sleeve resection for a pyloric lesion certainly holds infectious possibili-
ties. The lungs were clear before operation in all but one case.- The
complications were confined to the bases in all cases, bilateral in one
case, left side twice and at the right base three times. The time of
onset varied from immediately following operation to the end of the
first week. It should be said, however, that in all septic cases the
recognition of a concomitant complication is always delayed and difficult
because of the already existent picture of disease. In two cases only was
the onset of the complication striking from a study of the chart. One
of these began on the second and one on the eighth day after operation.
The majority of these patients had a productive cough due, apparently,
to a mild associated bronchitis. The presence of sepsis makes it diffi-
cult to explain the underlying etiology. It does not seem reasonable,
however, to suppose that aspiration has produced this untoward effect
only in such septic cases and the lobular rather than lobar distribution
of the lesions gives us encouragement to believe that minute septic
emboli may be at the bottom of the trouble in these cases. Or the
infection may have progressed by direct extension.

Bronchitis. — In this group are sixteen cases, and, as a whole, they
form a very distinctive picture. The whole field of surgery is covered,
both septic and aseptic ; si.x patients were operated on for hernia ; the
other operations, except for one each on the breast and thigh, were for
varying abdominal conditions. Septic lesions were present in more
than half the cases, and the patients were of all ages and both sexes.
Ether was given in eleven cases and gas-oxygen in five.

In eleven cases there was evidence of previously existent pulmonary
disease either in the past history or in the physical examination. This
usually was a chronic, intermittent, troublesome bronchitis. And this
appears to be of the greatest importance. There are undoubtedly many
cases among all pulmonary complications, like those in this group, in
which the development of complications can justly be charged to the

69. See ihe excellent papers of Bibergeil," Lawen" and Robb and Dittrick "
in which the high percentage of such complications following laparotomy is
demonstrated and discussed.

70. Bibergeil, E. : Ueber Lungen Komplicationen nach Bauch Operationen,
Arch. f. klin. Chir. 78:,339. 1905.

71. Lawen, A.: Ueber Lungenkomplikationen nach Bauchoperationen, Beitr.
z. klin. Chir. 50:501. 1906.

72. Robb and ■ Dittrick : Pulmonary Complications Following .'\bdominal
Operations, Surg., Gynec. & ()l>st. 3:51, 1906.

CUTLER-HU ST— POSTOPERATIVE LUXG LESIONS 461

anesthetic. In such cases the irritation of a pulmonary anesthesia is
sufficient to flare up a quiescent lesion. However, it must be under-
stood thoroughly that the existence of such a focus does not always
entail postoperative disease. By making roentgenograms in a large
series of cases before and after operation, irrespective of signs and
symptoms, we have been able to prove to ourselves that many times
such foci are present before operation and yet undergo no demonstrable
change after inhalation anesthesia. Just what the criteria are that
determine whether a case shall become active is not clear, but the fact
that in three of these sixteen cases the sudden and late onset of the
disease pictured the symptomatology of sniall emboli may mean that
in a certain proportion of these lesions, clinically recognized as
bronchitis, the flare-up is due to minute emboli from the operative field
that land in an already prepared area. In one of the cases classified
here as bronchitis an uneventful convalescence occurred until the eighth
day when a careful physical examination, done to explain a rise in tem-
perature, revealed a few rales at the base of the right lung. Such a
case might possibly have been better classified under pulmonary
infarction.

The majority of cases in this group are typical of postoperative
bronchitis, showing an immediate respiratory distress with a good deal
of cyanosis, frothy sputum and widespread rales. The upright position
and warm vapor inhalation proves of much benefit. In cases in which
any cardiac damage is suspected digitalis may be of assistance.

Exacerbation of Tuberculosis. — There are two cases in this year's
complications in which, subsequent to operation, an acute pulmonary
tuberculosis developed. One patient had acute appendicitis, the other
had tuberculous lymphadenitis. In the appendix case ether was, given
as the examination had revealed no pulmonary disease; in the second
case, both the physical signs and a chest roentgenogram gave warning
and gas-oxygen was used. Both patients developed almost immediately
following operation the signs of bronchopneumonia which were cor-
roborated by roentgen-ray studies which also confirmed the presence
of chronic lesions in the apices. Such a flare-up must be expected in
a certain percentage of patients with this condition. These cases are
presented here to emphasize the importance of pre-existent nontuber-
culous pulmonary disease as exemplified in the bronchitis group of
cases just discussed. Thus, a warning is sounded concerning casual
pre-operative pulmonary examination. In all such cases the anesthetic
of choice is local ; next choice, gas-oxygen.

Pleurisy. — Because of our adoption of infarction as a class of pul-
monary complication we find that the simple pleurisy class is very small.
However, it seemed wisest that the embolic lesions resulting in symp-

462 ARCHU-ES OF IXTERSAL MEDICIXE

toms that chiefly resemble pleurisy should be grouped with those larger
lesions that sometimes resemble pneumonia. W'e have, therefore, only
classified two cases this year as 'pleurisy. These are both of the serous
type and due to direct extension of infection from intra-abdominal
disease.^" One case occurred in a case of splenic abscess and one in a
case of high retrocecal appendix. At a late period during convalescence
respiratory pain led to a chest examination with the subsequent finding
of a straw colored fluid in both cases. In one organisms (strepto-
coccus) were present and the chest was repeatedly aspirated. Both
patients made good recoveries and were discharged well.

The etiology of such cases is manifestly direct extension through
the diaphragm. The work of Sabin *^ and Miller ^^ demonstrates the
avenues along which such extension may occur.

Empyema. — One case of postoperative empyema occurred. The
case was one of cholelithiasis (common-duct stone) in a man, aged 72.
Apparently because of insufficient local resistance an abscess developed
in the field of operation, gravitated down to the kidney and then spread
upward until a large subphrenic abscess formed. This finally ruptured
into the pleural cavity. The mechanism of the etiology in this case is
exactly that causing serous pleurisy, only more obvious and direct. It
needs no explanation.

Pulmonary Infarction. — W'e have placed thirty-two of the total
sixty-three complications in this group. The great relative size of this
group needs explanation for it is somewhat of a diversion from the
accepted classification of postoperative pulmonary disease. To all,
however, who are familiar with the difficulties of such classifications,
this group will be well understood.

In the past there have been many excellent papers on the embolic
origin of these complications. In most of such these lesions of embolic
origin, unless fatal, have been termed pneumonia. There have
remained, however, certain minor lesions, quite evidently of the same
origin, which have been designated pleurisy, and further, certain of the
ether classes of complication, as empyema and lung abscess, have been
known to be embolic in origin. Wc have long felt that the present
method of classification was unsatisfactory. Evidently, such lesions
are not a true pneumonia, and, furthermore, there were many border-
line cases which we found the greatest difficulty in classifying as either
pneumonia or pleurisy. Morever, as will be shown later, these cases
form a group with a rather typical clinical picture. The onset is usually
abrupt, the physical signs are characteristic, and the subsidence is sud-
den, except when the emboli have arisen in a septic field. As discussed
above under pneumonia, sepsis in the wound obscures the onset, the
typical picture and the subsidence of such lesions. When septic emboli

CUTLER-HUXT-POSTOPERATIVE LUNG LESIOXS 463

are present, the clinical picture may simulate pneumonia or lung abscess
may result, but when the clot is sterile, the resultant changes are char-
acteristic of minor pulmonary infarction.

There are no fatalities in these thirty-two cases. All ages and both
sexes are included. All but seven cases were laparotomies. The other
operations were one each for varicose veins, cancer of the tongue,
cancer of the breast, urethral stricture, tonsillectomy and two cranial
explorations. Anesthetic: procain, once; gas-oxygen, seven times;
ether, twenty-four times.

The clinical picture of all cases bears a close relationship. The
onset may occur from the second day to the third week, usually with
sudden pain on respiration, followed by expectoration in about one-half
the cases. The sputum is often blood tinged. Preceding the onset of
symptoms there is usually a rise in pulse, temperature and respirations,
and with the pain these may increase. Immediate auscultation of the
chest reveals one or more small areas filled with fine rales over which
there is some impairment of breath sounds and, if the focus is suffi-
ciently large, some change in fremitus. When pain is present a friction
rub may be the most distinct sign. However, it must be remembered
that although this is a characteristic sign when present it is so only
because it creates pain and thereby directs attention to that part. A
friction rub results only when the area of the infarct reaches the
periphery of a lobe, and it must be understood if we are to recognize
all these lesions, that some of the smaller thrombi do not cause suffi-
ciently large infarcts to do this.

We have long felt that infarction can occur early or even imme-
diately on recovery from operation. Cases have been observed in
which characteristic signs were present on recovery from ether with
chest pain, friction rub, bloody sputum and a demonstrable area of con-
solidation both by physical signs and the roentgen ray. In this year's
series the onset in twenty-one of the thirty-two cases was within the
first four days and in six cases it was within twenty-four hours.

Roentgen-ray studies should always be made and are often of the
greatest value. Invariably they will demonstrate small areas of con-
solidation which from time to time will take the form of a cone-shaped
shadow with its base out. Roentgen-ray studies, morever, should be
made immediately, since these lesions chiefly represent merely a change
in blood distribution and soon clear up. The adjustment to normal is
complete, as a rule, within six or seven days.

In order to demonstrate more completely our conception of these
lesions, the following cases with their charts and roentgen-ray studies
are presented.

464 ARCHIVES OF IXTERXAL MEDICINE

REPORT OF CASES

Case 1 (P. B. B. H. Surg. No. 13020).— J. B., aged 25, single.

Diagnosis. — Right indirect inguinal hernia.

Operation. — Aug. 25, 1920, procain, 1 per cent. ; Dr. Cutler. On the first
day after operation the patient cotuplained of severe pain in his back; his tem-
perature kept rising and there was evident respiratory difficuUy. Morphin and
digitahs were administered to control pain and coughing. Examination of the

7a

Peter Bent Brigham Hospital

SURGICAL SERVICE /3«««

Ovd

^ , , i ^ i i ■ , i - : i i

^

/ z 3 tf^ sr C •/ t y >i /, u /s/f 's-ri

-' --r^Y^J'-'-'-'jn^r. — rr

i fc'"" j

J±|j±:;-T---' - ^ f-- :z

"•p?-

IHj,^ • ^ •, .-:-!^jJ_

"■f **

?rWA. J\ H — r

■^' ^V^lrV.-/-VHH-r

is-l

:5

_._LJ_i_uJJ_|_UJ iJ^ 1 1 r ;

i !"'! i" M i '■ 1 i M

mjj Ij n_ 1 ! ^ !

z":-

J J_| L_IJ j_|_!_i_l_i^5i?«f^|-U!-^|—

'iw

ijl i 1 i 1 _j j i i 1 '

\

jiy^j 1- 1 ^ ij_^' J j_^i i

[loot

!n ;■

^ i !• j^y\j* .lj 1 Mi

m u —

^ ip!pj**^:^k. —

Ml 1

1 i M II M M i M i

Fi-. 1.— I'aso 1. Clinical char

chest showed small areas in both lungs in which there was slight dulness,
increased voice sounds and a few rales with .some increase in tactile fremitus.
Within ten days these signs had all disappeared. Unfortunately no roentgen-ray
studies were made until the twelfth day postoperative but even as late as that
a subsiding bronchopneumonic process in the right lower lobe was demonstrable.
Discharged well (Figs. 1 and 2).

CUTLER-HUXT—POSTOPERATIl-E LUXG LESIOXS 465

This case was unusually se\ere. It is presented because of its
immediate occurrence following operation. That emboli can actually
take place during operation we fee! satisfied and have known a patient
to come out of the anesthetic with chest pain, a demonstrable friction
rub and spitting up blood tinged sputum. The rough use of retractors
surely could endanger a patient to this extent.

ClK-st ph.

Case 2 (P. B. B. H. Surg. No. 13387).— .-X. H. McM., aged 33. married.

Diagnosis. — V'arix femoral vein.

O/'cro/ion.— Exploration, Oct. 16, 1920; ether; Dr. Homans. Convalescence
normal until the eighth day after operation when the patient complained of
a slight cough, and there was a beginning rise in pulse, temperature and res-
pirations. Examination of the chest revealed a small patch of impaired reso-
nance, bronchovesicular breathing and rales in the back on the left side below
the angle of the scapula. Within ten days all these signs had subsided, but
a roentgen-ray examination made on the seventeenth day after operation showed
a spotty consolidation in the left upper and lower lobes. Discharged well
(Figs. 3 and 4).

466 ARCHIVES OF IXTERXAL MEDICIXE

In this case, a sudden lesion appeared in a healthy young man eight
days after a simple incision over the left femoral vein.

Case 3 (P. B. B. H. Surg. No. 13544).— D. D. B., aged 43, married.
Diagnosis. — Duodenal ulcer.

Peter bent Brigham Hospital

SURGICAL SERVrCE /JJV/.

111

2if-s'C'7i']

-I ^

z:\

.%

t '^v

v^^m

\ '^

Explor: of- inff/n»l r€j,cri-

\%**«***,%%v

in-

^-1.

3. — Case 2. Clinical chart.

Operation. — Cauterization of ulcer; posterior gastro-enterostoniy, Nov. 22,
1920; gas-o.xygen-ether; Dr. Cheever. Convalescence undisturbed until the
eighth day after operation when the patient complained of pain in the right
chest on breathing. Examination showed a friction rub, increased breath
sounds and a few rales at the level of the angle of the right scapula. The
temperature, respiration and pulse rate began to climb and all physical signs
became more marked. There was no sputum. The leukocyte count rose to

Ci'TLER-HUXT—POSTOPERATIlE LiWG LESIOXS

467

15.000. On the eleventh day after operation a roentgen-ray examination showed
a definite pneumonic process involving the middle and upper lobes on the right
side. The entire picture cleared up in ten days (Figs. 5 and 6).

Case 4 (P. B. B. H. Surg. No. 12932).— E. A. H.. aged 48. married.

Diagnosis. — Pyonephrosis in right kidney.

Operation. — Right nephrectomy, Aug. 13, 1920 ; gas-oxygen ; Dr. Quinby.
Convalescence uninterrupted until the fifth day postoperative when the patient
complained of pain in the right lower chest. Examination of the lungs was

negative although there was a slight rise in pulse rate, temperature and res-
pirations. Two days later auscultation revealed rather distant breath sounds
and a questionable rub. There also was some limitation in the diaphragmatic
excursion and diminished resonance. .\ roentgen-ray examination on the
seventeenth day after operation revealed a resolving pneumonic patch involv-
ing the lower right lobe. From this time on improvement was rapid and the
patient was free from symptoms and signs when discharged thirty days after
operation (Figs. 7 and 8).

The explanation of why kidney operations sometimes result in ptil-
monary complications on the .same side is not clear, except when there
is direct extension of an infectious lesion, as when empyema occurs. Is

468 ARCHIVES Of IXTERXAL MEDICI XE

it possible that the manipulations on that side damage the lung some-
what so as to render it more susceptible to infarction? (Jr is it merely
because pulmonary lesions usually involve the lower lobes? We have
seen such complications in the opposite lung from the site cf operation.

Peter Bent Brigham Hospital

SURGICAL SERVICE 135"^/^
J. 3, ait ^,3. , WARO DATC>V<"

z d ¥- s- i 7 S 9 '1

''0 /f It /f 20 II iX ZJ^*

r

\-- .

,!#%

Uluodtnal OU

Cliiii

ill.

C.\.SE 5 (P. B. B. H. Surg. No. 13363). -A. T. S.. aged 23. single.

Diagnosis. — Acute appendicitis.

0/>c)-a/io«.— .\ppendicectomy ; no drainage, Oct. 18, 1920, ether, Dr. Jameson.
Convalescence uninterrupted and unusually clear until the seventh day after
operation when tlie patient complained of sudden severe pain in the right
side of the chest and within a few hours he expectorated some bright blood-
stained sputum. Tlie temperature rose somewhat and respirations climbed
markedly. E.xaniination of the chest revealed a small patch behind the angle

CUTLER-HU XT— POSTOPERATIVE LLWG LESIOXS 469

of the scapula in which there were medium moist rales. Xo other signs were
demonstrable and the whole picture subsided in three days. A roentgenogram
made on the eleventh day after operation showed an increase in the shadow at
the hilum on the right side and a little fluid at the base of the right lung.
The patient was discharged well on the fifteenth day (Figs. 9 and 10).

This case presents a more ideal picture of this group as a whole
than the preceding ones. All the cases, however, present different
p.spects, and include operations under the chief anesthetics.

Pulmonar\ Embolism. — There are two fatal cases of pulmonary
embolism in this year's series. A man, aged 67, operated on for
prostatic hypertrophy, died on the third day after operation ; a woman,
aged 55, operated on for ulcer of the stomach, died on the twenty-third
day, three days after getting out of bed. A necropsy was held in both

470 ARCHIVES OF IXTERXAL MEDICIXE

cases, and the emboli were demonstrable in the pulmonary arteries.
There is no need for any discussion concerning this frightful com-
plication. We would, however, like to call to mind that granted that
such large emboli occur, the mechanism for the manufacture of smaller
ones must be the same and probably they occur with far greater
frequency than in tlie large ones.

PETER BENT BRIGHAM HOSPITAL
SURGICAL SOnriCE /if JO.

. , V _.

'*^>sg^

iMSE

Fig. 7. — Case 4. Clinical chart.

irge i:
lused

DISCUSSION

The material presented would aj)pear to indicate that ;
portion of postoperative pulmonary complications are
embolism frnni the ojierative field. This opinion, which we ha\e pre-
viously a'.tem|)tL'(l to substantiate "' is in agreement willi a large propor-
tion of the studies in this field.'-' It is, however, by no means generally
accepted and curiously enough not by the anesthetists whom it
attempts to defend ! Preexistent pulmonary disease appears to be the

77. Co

Lilt introduc'.ion and review nf the literature cited.

CCTLER-HLWT—POSTGPERATirE LUXC LESIOXS 471

next most important factor. Such lesions are usually brought to activity
through the irritation of an inhalation anesthetic and result in Ijron-
chitis or pneumonia. It is only in this type of case that the anesthesia
has been proven at fault. The presence of sepsis in the operative field
and the location of the operation are often important secondary factors.
They do not, however, offer an explanation as to the mechanism by

which the lung is injured. The further secondary factors of chilling,
acidosis and seasonal incidence bear a varying relation in each case.

The arguments favoring embolism are: (1) the typical clinical pic-
ture, with sudden onset, focal signs, and, unless sepsis is present, rapid
subsidence; (2) the fact that these complications occur frequently with
local anesthesia : and ( 3 ) that they occur in a definite proportion
according to anatomic divisions, these divisions being those kept in

472 ARCHIVES OF IXTERXAL MEDICIXE

greatest mobility and giving easy access by blood and lymphatic chan-
nels to the lung."'* In favor of inhalation irritation is (1) the evidence
that aspiration into the lung of the mouth contents does occur during
anesthesia ; ( 2 ) that Group IV pneumococcus, a common mouth inhab-

Peter Bent Brigham Hospital

SURGICAL SERVICE liiCS.
„„E /}-TS. a.cr. ZS WARD DATT OcfSUr.

itant, is most frc(|ucutly tdund in thoc cases and (3) the many reports,
including our own, lh;it in certain c;ises with a pre-existing pulmon.-iry
disease inhalation anesthesia seems to have caused a tlare-up of ^uch a
lesion.

74. Mandl, F. ; Postoperative Lung Complicati
34:214 (April 28) 1921.

Ci:TLER-HLXT— POSTOPERATIVE LiWG LESIOXS 473

Although under anesthesia the aspiration to the lung of mouth con-
tent undoubtedly occurs, there is no great proof that it effects any great
pathology. All surgeons will recall execrable anesthesias with blue,
coughing, vomiting patients. And yet, in how many such cases did
pulmonary complication develop? If this were the chief factor, why
is it that with the increasing perfection of the technic of anesthesia
such complications continue to be found? And what is the explanation

Fit;. 10.— Case

Chest plat

for those cases occurring with local anesthesia? That Cirou]) I\' pncu-
mococcus is found in the sputum in such cases is no proof that a true
pneumonia has occurred for it has been shown that many normal
mouths harbor this organism. Morever, the enormous volume of
experimental pneumonia work does not reveal much success with
inhalation infection except with doses so large that the conditions are
not comparable to what might actually occur. Blake and Cecil's work
on monkeys stands out almost alone, for the successful production of

474

ARCHirES OF IXTERXAL MEDICIXE

pneumonia with small doses. Those who would blame inhalation
anesthesia must bring more evidence to substantiate their hypothesis,

Of the sixty-three cases with complications reported in this series,
only thirteen gave evidence of a pre-existing pulmonary lesion. More-
over, control roentgen-ray studies demonstrated that many persons
with such lesions take inhalation anesthesia without any bad effects,
whereas in more than 50 per cent, of the reported cases a typical pic-
ture of mild pulmonary infarction is presented.

1

20

^

!!

!!

19

f

IB

\

17

\

16

\

15

14

12

\

IE

11

10

9

6

f

7

/

A

6

/

i

5

i

4

\

3

\

2

h

1

Fig. II. — Incidence of postoperative piilmc

decade

Against the influence inhalation ancstln'sia may have in the jiroduc-
tion of these lesions are the facts, ( 1 i that they nccur equally when the
anesthesia is in the most skillful hands; (2) that they occur with local
anesthesia and, (3) that they occur in a greater relation to the mobility
of the part than can be explained on any irritation hypothesis. With
irritation, tiie presence of Croup IV pneumococcus and a pre-existing
pulmonary lesion one luight expect a lobar distribution were this the
true mechanism of these complications whereas lobar pneitmonia
I^roved to be a rare lesidu in this study.

CUTLER-HUXT— POSTOPERATIVE LiWG LESIOXS

The part that the predisposing, accessory and secondary factors
play is not clear. Chilling can be ruled out in a well ordered clinic.'^
The production of a real acidosis is still under dispute. If it does exist
we might expect some lowering of the resistance to infection.""' That
the majority of the complications have occurred during the middle
decade when individual resistance is at its best is not in keeping with
this theory.

Ills^ll^llSIa

35

/I

ao o.^ \ o.i6)4

f X ^

-. tv5

iz 9 1 J k

le T i ^ ^^

^^ 4 "

X4 i t

13 ±

12

11 U -t-

" U ft

1 □ >'^

^ 1 41 i

JL _! J

^ 5 A ^11 t

5 _A J. ^^jQ^ /

, ^%r %Xt^

t&-^^ 2 v5

1 5^ ^^^ i^

^

3=3. !3^

Fig. 12. — Seasonal variations in postoperative pulmonary complications:
open circle, number of operations per month (the number at the side should
be multiplied by 10) ; solid circle, number of pulmonary complications per
month; circle and dot. percentage of pulmonary complications per month.

In a previous study "* the material presented demonstrated the great
frequency of such complications with abdominal operations and the

75. Mandl" reports that during the winter 1919-1920 in Hochenegg's clmic
in Vienna no coal was available and that the percentage of postoperative com-
plications increased markedly. Such complications, however, do occur in the
most luxuriously appointed clinics.

76. Hamburger, H. J.: Researches on Phagocytosis, Brit. M. J. 1:37, 1916.

476 ARCHirES OF IXTERXAL MEDICIXE

cases this year bear out the general comparison. The incidence in this
restricted field has been so striking that there have been many excellent
studies concerning it. Table 3 demonstrates the number of complica-
tions in relation to the field of operation.

TABLE 3. — Complications in Rel.atiox to Field of Oper.\tion

Pulmo- Pulmo- Exacerba-

Lobar Broncho- nary nary tion of

Pneu- Pneu- inlarc- Bron- Pleu- Empy- Embo- tubercu-

Operatjve Field monja monia tion cliitis risy ema lism losis Totals

Cranium 1 2 .. .. .. .. 3

Breast .. 11 .. .. .. .. 2

Kidney ] 2 2 .. .. .. .. 5

Upper abdomen 1 4 5 1 1 1 ..13

Lower abdomen:

1. General 3 3 .. 1 . 1 S

2. Inguinal hernia.... 1 .. 6 4 .. .. .. .. 11

3. Pelvis 1 7 2 .. .. .. .. 10

4. Bladder (prostate) .. .. 1 .. .. 1 .. 2

Perineum .. 1 .. .. .. .. .. 1

.Scrotum .. .. 1 .. .. .. -. 1

Lower extremity .. .. 1 .. .. 1

Varicose veins .. 3 .. .. 3

Mouth .. 2 .. 2

Neck .. .. .. .. .... 11

This table demonstrates again that the preponderance of such com-
plications follow abdominal operations. In this series forty-three of
the sixty-three cases followed celiotomy (68 per cent.). Casually, it
might not appear that infarction would be more common with abdom-
inal operations and yet there are certainly several reasons favoring this.
The anatomic studies of Sabin have demonstrated the free lymphatic
drainage. But what of the blood supply? A large part of the intra-
abdominal vascular tree drains to the portal system and must be
excluded. The generally accepted higher incidence of .such complica-
tions with epigastric lesions led us to believe that it was the incision in
the abdominal wall itself which was most important, and the fact that
this year two patients on whom an exploratory laparotomy was ])er-
formed developed typical infarcts would seem to uphold this. The
greater mobility of the parts due to respiratory movements is the direct
cause for setting free the emboli, and the use of retractors, often used
with great force in laparotomies, is an added danger. Furthermore, it
has appeared that when great gentleness in traction was exerted and
where perfect hemostasis and accurate closure was practised the
incidence has been lessened. It is, however, quite possible, as we have
previously indicated, that by the transfer of particles through the larger
lymphatics jnilmonary changes resulting in the signs of infarction
occur. Mandl '* reports that pulmonary complications in 1,379 lapa-
rotomies reached 14.5 per cent., whereas 1.585 operations elsewhere
yielded 8.5 per cent, complications. These figures are higher than our
own but are relatively the same.

CUTLER-HUXT— POSTOPERATIVE LUXG LESIOXS 477

PETER BENT BRIGHAM HOSPITAL

ANAESTHESIA CHART /\

,..- Ward Age _ Date of OperaOon

..^lood Pressure .

Method of AdmiuistraiioQ .

ff rl I I I I I =

m

» IM

m - ;. -

m • : -

M IM

no

" IN

U IIS

u Its

!S M

U 7«

U H

CondjUon on leading operating r

Fig. 13. — A, front of chart used to record progress of anesthesia.

478 ARCHIVES OF IXTERXAL MEDICI XE

In the material presented this year there are no cases in which we
feel that aspiration or irritation due to the anesthetic was the chief
causal element, except in those cases in which a pre-existing pulmonary
lesion had been demonstrated. Thus, in the bronchitis and tuberculosis
group there is no doubt that the inhalation anesthetic is to be blamed in
many of the cases, but certainly not in all, for these complications also
occurred with local anesthesia. And, moreover, a routine series of pre-
operative and postoperative chest plates demonstrated the roentgen-
ray evidence of pre-existing disease in which sometimes no change
occurred after operation under inhalation anesthesia.

POST OPERATIVE OBSERVATION:
Rpcta) temp oii urri\al id rtcoverj rcM

Fig. 14.— B, reverse of anesthesia chart.

Through the successive years during which we have been interested
in this subject the roentgen ray has steadily proved an increasingly
valuable and reliable adjunct to accurate diagnosis. The opportunities
to put it to more extensive use during 1920 have not only enabled us
to make more accurate diagnoses and to pick up doubtful cases but have
given us a greater insight into pulmonary conditions and, in particular,
the frequency, type and reaction to inhalation anesthesia of such pre-
existing foci of pulmonary disease. These studies have proved to our
satisfaction that without such controls the classification and study of
these complications is difficult and incomplete.

A study of the distribution of the complications resulting from
infarction, embolism and pneumonia, shows twenty-five situated in the
base of the right lung, eleven in the base of the left lung and six
bilateral. If the lesions were irritative and infectious, it is inuisual

CUTLER-HCXT—POSTOPERATirE LUXG LESIOXS 479

that involvement of the upper lobes never occurred. Welch states that
pulmonary infarction is most common in the lower lobes, more fre-
quently right than left.

Figure 11 demonstrates the incidence of pulmonary complications
by decades. Figure 12 shows the seasonal variations and Table 4
shows the types of complications in relation to the various anesthetics.
There is little to be said about any of these findings. Obviously the
majority of complications occur in the fourth decade as the majority of
patients are that age. It should be remembered that pneumonias of
either type are more frequent in the first two decades and the fifth
decade (Osier and Macrae"). The seasonal chart shows no striking
segregation in relation to climatic conditions. Table 4 shows that about
one half as many cases received gas-oxygen as ether. This alone, how-
ever, is not convincing since obviously those with pre-existing pul-

T.ABLE 4. — .-^XESTHETics Used ix V.^rious Complications

Gas-Oxygen Procain

Complications

Lobar pneumonia

Bronchopneumonia

Bronchitis

Exacerbation of tuberculosis....

Pleurisy

Empyema

Ether
0

11
1
0

24

Gas-Oxygen

and Ether
0
1
0
0
0
0

Pulmonary embolism

0

Totals

41

1

monary signs were usually given gas-oxygen. That only two complica-
tions occurred with procain is exceptional although not far from pro-
portionate to the total use of this anesthetic in this clinic and the fact
that local anesthesia is rarely used in abdominal operations excludes
this added predisposing factor in the local anesthesia cases reported.
Of the total 1,604 operations, 875 were performed under ether, 546
under gas-oxygen, 179 under jjrocain, 2 under chloroform and in two
cases no anesthesia was used. The morbidity percentage is 4.6 per cent,
for ether, 3.5 per cent, for gas-oxygen and 1.1 per cent, for procain.

The difficulties of assembling accurate data have been somewhat
mitigated by the use of the chart shown in Figures 13 and 14 which is
reproduced in the hope that the few clinics which ha\e not as yet
adopted some such form may be led to use one. We have found the
data contained on this chart of the greatest value — (1) in ensuring a
careful anesthesia, and (2) for all kinds of follow-up work and case
studies. It has the great value of providing the anesthetists with sufti-
cient work so that their attention is never side-tracked to the operation
or to other activities in their pro.ximity.

77. Osier, W., and Macrae, T. : Loliar Pneumonia, Mod. Med. 1:202, 1913.

480 ARCHIVES OF I.XTERXAL MEDICIXE

As a further postoperative protection all patients are kept on the
operating room floor, in separate recovery rooms and under constant
observation until fully conscious. This does away with the possible
danger which may come from transporting unconscious patients
through draughty corridors and from the more active risk of reaching
a short-handed ward where they cannot be under constant observation
during the period of recovery.

The condition known as massive collapse of the lung first described
by Sir John Bradford " has recently through the papers of Scrimger '"
been brought into the category of postoperative pulmonary complica-
tions. In Scrimger's last paper he reports seven such cases in 540
operations. The routine anesthetic of chloroform-ether is used. His
patients soon after operation developed sudden respiratory distress,
pain and cough. Roentgen-ray studies revealed marked displacement
of the heart to the affected side. Recovery usually occurred within
twenty-four hours. The etiology of this condition is not known, pos-
sibly it is due to inhibition of the diaphragm or to vagal influences
causing contraction of the muscular elements of the lung. Mortimer '"
places great stress on respiratory movements and states that the lower
lobes may collapse when movements of the diaphragm are restricted
either by pain or bandages. This condition deserves greater study and
necessitates control roentgenograms. It has not been observed in this
clinic.*'

SUMM.\RV AND CONCLUSIONS

Postoperative pulmonary complications are due in the majority of
cases to embolism from the operative field. The result is pulmonary
infarction or fatal pulmonary embolism. The latter is a rapidly fatal,
well recognized clinical picture. The former is caused by the trans-
mission to the lung of a small thrombus or many thrombi with resultant
characteristic clinical picture. This consists of sudden pain, expectora-
tion, often blood tinged, the elevation of respiration and pulse rates and
temperature, and the signs of a focal consolidation often overlaid liy
a friction rub. These patches are demonstrable by the roentgen ray,
provided such studies are made at once. There usually follows rapid
defervescence and the subsidence of all symptoms within a week.

78. Bradford, Sir J. R. : Massive Collapse of the Lung, O.xford Medicine,
New York, and Oxford University Press 2:127. 1920.

79. Scrimger, F. A. C. : Postoperative Massive Collap.se of the Lung, Surg.,
Gynec. & 01)st. 32:486, 1921.

80. Mortimer. J. D.: Med. Press & Circ. 108:505, 1919,

81. Schult^c, E. C. : A Report of Twenty-Seven Cases of Pneumonia Fol-
lowing the Inhalation of Ether and Chloroform. Med. it Surg. Rep. Preshy-
terian Hospital. N. Y. (Jan.) 1898. p. 311.

CUTLER-HUST— POSTOPERATIVE LUXG LESIOXS 481

The causes of infarction are (1) trauma; (2) the mobility of the
pari; (3) sepsis. The prognosis in this group is excellent.

Irritation and aspiration due to inhalation anesthesia may be the
cause of a small percentage of postoperative pulmonary complications.
These usually fall into the bronchitis or pneumonia groups. Inhalation
anesthesia rarely results in such lesions unless there is present a pre-
existing pulmonary disease such as a chronic bronchitis, incipient tuber-
culosis or a definite tendency towards pulmonary disease.

The fact that these complications occur with local anesthesia, with
inhalation anesthesia in the most expert hands and in a definite relation
to the mobility of the operative field is taken as evidence against the
importance of the irritation of inhalation anesthesia in the production
of these lesions.

Anesthetists and anesthesia should not bear the blame for these
complications.

There is as yet no proof or evidence that chilling or acidosis plays
any appreciable role in this field.

A reduction in the number of cases that result in- these complica-
tions can be had by (1) a reduction in trauma at operation; (2) accu-
rate hemostasis; (3) the careful control of sepsis and (4) the use of
great caution in operating upon patients who have demonstrable pul-
monary disease. A high fluid intake and all general precautions giving
assistance to the circulatory apparatus will be of definite value once
such complications are established.

We are greatly indebted to Dr. Lawrence Reynolds, roentgenologist of the
Peter Bent Brigham Hospital. fi)r his excellent plates, fluoroscopic studies and
cooperation.

THE PATHOLOGY OF CIRRHOSIS OF THE LIVER

AX HISTORIC-PATHOLOGIC STUDY

FREDERICK EPPLEX, M.D.

SPOKANE, WASH.

Cirrhosis of the liver is a chronic, recurring, probably focal, possibly
diffuse degeneration or necrosis of the parenchymatous cells of the
liver, modified by concurrent and intercurrent periods of regenerative
proliferation of the parenchyma, with connective tissue replacement
of destroyed areas. Acceptance of this definition and any concept that
cirrhosis of the liver is a disease entity are incompatible. The granu-
lar hver, with its associated splenic uimor and portal stasis, is as
surely a consequence of some preceding disease as the cyanotic indura-
tion of the liver and lungs and edema are the consequences of chronic
valvular heart disease. That disease is the degeneration or necrosis
of the liver cells, a process, the' clinical manifestations of which are
as yet unknown.

The question arises : Are we concerned with a single disease result-
ing in this picture, or are there many entirely different processes which
eventuate in the same ultimate anatomic picture ? Further, and perhaps
more important, is the question whether the primary disease has its
seat in the liver, or whether that organ is involved only secondarily
or in association with some other organ, as for example, the spleen.
The varying clinical pictures under which cirrhosis may manifest itself,
jiarticularly with reference to its onset, and still more, its etiology,
makes it seem reasonable to assume that the end-result called cirrhosis
of the liver can be produced by a variety of primary clinical processes.
Accepting its association with other diseases, the second question may
be reasonably assumed to be correctly answered when it is stated that
it may be primary in the liver or may be secondary to (or associated
with) disease in some other organ. Of course, this statement might,
with equal accuracy, be made the other way around, namely, that this
pathologic picture may exist alone or it may produce lesions in other
organs. Perhaps, both are true. For example, it might be secondary
to changes in the spleen in Banti's disease, while the brain lesion in,
Wilson's disease might be dependent on the liver pathology.

ETIOLOGY

The intimate relationship of the etiology and pathogenesis of cir-
rhosis is sufficient reason for a brief resume of the usual etiologic
factors concerned in Laennec's cirrhosis, especially since this relation-
ship is here particularly interesting and apparently insoluble.

EPPLEX—UrER CIRRHOSIS 483

The term "alcoholic cirrhosis"' applied so universally to the hob-
nail liver characteristic of Laennec's cirrhosis is not altogether justi-
fiable. Alcohol is undoubtedly closely associated etiologically with the
great majority of cases; and that alcohol alone can produce the disease
has apparently been proved experimentally by many workers. Grover ^
published the most recent work on experimental alcoholic cirrhosis,
and has made an admirable collection of the literature. He concludes
that alcohol given to experimental animals over a long period of time
can produce a degeneration of liver cells followed by connective tissue
proliferation. He concedes that other processes may be associated, but,
if so, their nature is so obscure that their modus operandi is not known.
Saltykow - produced cirrhosis of the liver by injecting alcohol into the
ear vein of rabbits. Fahr ^ and Schafer * were successful in similar
experiments. Kyrle and Schopper ° produced parenchymatous changes
and fatty degeneration in all cases of repeated introduction of alcohol
by intravenous, subcutaneous and gastric routes. In seven cases there
was necrosis of portions of the lobule and replacement fibrosis; in
fourteen cases there was round-cell infiltration in the portal fields, and
in seven others, there was connective tissue proliferation with bile
duct formation. The livers of three rabbits showed typical Laennec's
cirrhosis.

Xot all workers, however, were able to produce cirrhosis in experi-
mental animals. Pogenpohl," Klopstock,' von Baumgarten * and
Bischoff" were notably unsuccessful. All agreed, however, that alco-
hol was productive of extensive fatty degeneration.

Lancereaux," who studied the etiology of cirrhosis of the liver for
many years, concluded that cirrhosis could be produced by substances
used as preservatives and to give certain wines an especially "dry"
flavor. He had particular reference to potassium bisulphate with which
salt he was able to produce Laennec's cirrhosis by feeding it to animals.

MacCallum. in his last textbook (1916), is very emphatic that it
has not been demonstrated experimentally that alcohol produces cir-
rhosis, n a large number of chronic alcohol addicts be studied, only
a very few will be found to suffer from cirrhosis. He states that all
sorts of degenerations may take place, both in experimental alcoholism

1. Grover: Arch. Int. Med. 17:193 (Feb.) 1916.

2. Saltykow: Yerhandl. d. deutsch. patholog. Gesellsch. 15:228, 1910;
Zentralbl. f. allg. Path. u. path. Anat. 22: 1910.

3. Fahr: Virchows .\rch. f. path. Anat. 217:.397, 1911.

4. Schafer: Virchows .^rch. f. path. Anat. 215: 191.3.

5. Kyrle and Schopper: Virchows Arch. f. path. Anat. 215:.359, 1913.

6. Pogenpohl: Virchows Arch. f. path. Anat. 116:466, 1909.

7. Klopstock: Berl. klin. Wchnschr. 47:1532, 1910.

8. Von Baumgarten: Verhandl. d. Deutsch. path. Gesell.. 1907.

9. Bischoff: Ztschr. f. Exper. Path. u. Thcrap. 11:445, 1912.

10. Lanccreaux: Bull, dc I'Acad. de med. Par. 74:15, 1910.

4S4 ARCHirES OF IXTERXAL MEDICIXE

in animals and overindulgence in man. Since MacCallum has been
one of the particularly deep students of cirrhosis, his statements are
of especial value. He calls attention to the work of Longcope, who has
produced changes very closely resembling cirrhosis of the liver by the
injection of proteins and by other experiments suggesting some form of
protein sensitization or intoxication as the best explanation. Mac-
Callum's ultimate conclusions are, that what we see as cirrhosis at
the necropsy table is scar tissue and efforts at regeneration and that
there are many processes which can lead to them.

These observations greatly accentuate the difficulty of establishing a
clear-cut pathogenesis, particularly if one tries to accept the theory that
cirrhosis of the liver is' essentially secondary to some toxic process in
the portal zone.

There are, however, so many etiologic factors other than alcoholism
operative that this is only the beginning of confusion. Warthin recently,
before the Portland Academy of ^ledicine. expressed the belief that
some day it would or might be proved that "cirrhosis was the result
of the joint action of syphilis and alcohol. At that time he stated that
he had demonstrated Spirochaetae pallidae in two such livers examined.
Syphilis as an etiologic factor was the subject of a recent paper by
Symmers."

Malaria is immediately suspected in malarial districts. This may
be a relic of the days of pure empiric medicine ; the enlarged spleen
of both diseases furnishing the empiric resemblance. However, malaria
has not yet been entirely swept from the paragraphs on etiology, even
in relatively recent articles.

The outstanding objection to alcoholism as the only etiologic factor
is that we see so many cases in which there is not the slightest suspicion
of alcoholic abuse. Indeed, the patient from whom the liver presented
was taken, was a total abstainer, and so it is in many other case
histories. It will be interesting to note whether the present prohibition
law will cause this disease to diminish, or whether, in the perversity
of things, the illicit brews and di.stillations, at present so frequently
heard of, will actually continue to be a factor. Perhaps, their vicious
nature and impurities will lead to a greater morbidity with reference
to cirrhosis among drinkers than ever before.

Recently Miller,'- in discussing this question, pulilished figures
which show a definite decrease in the admission of cases of cirrhosis
of the liver to Cook County Hospital since prohibition went into eft"ect
July 1, 1919. He feels, however, that some of tlie reduction in the
number of cases admitted to the hospital may have been due to the
influenza epidemic raging at that time, during which the hospital was

11. Symmers: Intcrnat. Clin. 1:58. 1917.

12. Miller: J. A. M. A. 76:1646 (June 1<5) 1921.

EPPLEX— LIVER CIRRHOSIS 485

closed to all but acute cases. More time must elapse and illicit vending
be better controlled before the effect of prohibition can be estimated
accurately.

It is possible that a particular type or degree of gastro-enteriiis
must be produced by alcoholism before cirrhosis can develop. On this
basis we can also explain those cases in which alcohol was not indulged
in, on the assumption that an identical or very similar gastro-enteritis
is produced by other etiologic factors, for example, the custom of
pepper and curry eating in India. Cirrhosis then follows as it does
after that of alcoholism. It does not, however, explain those cases of
experimental cirrhosis in which alcohol was administered subcu-
taneously and intravenously.

As in all chronic degenerative diseases, infection is being spoken
of more and more in recent literature as a possible etiologic factor
in cirrhosis, and there is much to commend this hypothesis. In search-
ing the literature we found this etiology mentioned by Raoul Gaston '=*
as early as 1893. He believed that he had established an infectious
origin for some cases which followed prolonged infectious diseases.
His interpretation was that an inflammatory infiltration had occurred
about the radicals of the portal vein, and thus precipitated the disease.
It is to be noted that at that time the disease was believed to be primary
in the connective tissue.

The etiology of no chronic degenerative disease would be complete
unless some mention be made of chronic metabolic or digestive toxi-
coses. They are visionary, indeed, and yet possess an alluring appeal
in so many diseases, and especially in connection with cirrhosis, that
they cannot be dismissed entirely, theoretical though they be at present.
Particularly appealing is the assumption that the poison responsible for
this disease may be produced in the spleen.

Tabulate the clinical types of Laennec's cirrhosis and the all-
comprising single etiologic factor at once disappears.

1. Laennec's Cirrhosis in Topers. — These are so predominant in
numbers that alcohol becomes overwhelmingly the greatest single
factor.

2. Laennec's Cirrhosis, Where Alcoholism Clearly Was Not a
Factor. — The etiology of these cases is entirely unknown. They are not
so rare as is generally assumed, and are most common in women and
children.

3. Banti's Disease. — By definition Banti described this as a certain
type of anaemia, plus Laennec's cirrhosis and splenomegaly in which
the etiology must be unknown. If the victim of this disease is a
chronic alcoholic Banti's disease may not be used as a designation.

13. Gaston, Raoul : These de Par., Du Foie Infectieux, Paris. 1893.

486 ARCHIVES OF IXTERXAL MEDICIXE

4. Diabcte Bronze; Cirrlwse Pigmentaire. — Clearly this is a definite
clinical entity. It is perhaps an event in the course of some other
disease, but for that matter the same may be said of Banti's disease.

5. Progressive Lenticular Degeneration (commonly called Wilson's
disease'^). Here we have a rare familial nervous disease occurring
during the years of adolescence in individuals who are not addicted to
alcoholic excess, frequently not even to occasional indulgence. The
etiology is entirely unknown.

Cirrhosis is occasionally found by surgeons unrelated to the condi-
tion which called for the operation and apparently of no immediate
clinical importance,^^ So, too, it is often accidentally found at necropsy,
apparently not the cause of death nor even contributory. However,
some caution should be exercised in brushing aside a cirrhosis of the
liver as an "accidental finding" at necropsy. This is exactly what
occurred in the first case of progressive lenticular degeneration found
at necropsy at the National Hospital for the Paralyzed and Epileptic
in London, only to be resurrected from the archives years later by
Wilson, who then properly classified it and placed it in his series
reported in Brain in 1912. Howard and Royce ^"^ narrowly missed
making the same mistake in their case. It is just this association with
other apparently more important anatomic findings that makes such
"accidentally" found cirrhoses interesting and at the same time
extremely puzzling.

P.\THOGENESIS

The older theories located the origin of the disease in the connective
tissue stroma, a chronic interstitial hepatitis. Assuming that the disease
was more or less acute in its earlier stages, and that all acute inflamma-
tory processes are associated with edema and hyperemia, the swollen
liver of early cases was thus readily explained. As the disease
progressed it became more chronic, hence more productive and the
cellular elements of the connective tissue proliferated, sooner or later
insinuating themselves between the liver cells, in the lobule, gradually
snaring them off. As the new tissue acquired age it began to contract,
reducing the size of the liver simultaneously by contraction and by
pinching the liver cells to death, lessened parenchyma meaning lesser
volume. In younger individuals the snaring process was likely to take
larger bites of tissue, while in older individuals small areas would be
surrounded, so the theory .stated. Hence we had multilobular and
unilobular cirrhoses, both terms now very little used, the former being
reserved by some writers for the end-result of syphilitic hepatitis in
children.

14. Wilson: Brain 34:295. 1912.

15. Mayo: J. A. M. A. 70:136 (May 11) 1918.

16. Howard and Royce: Arch. Int. Med. 24:497 (Nov. 15) 1919.

EPPLEX—LIIER CIRRHOSIS 487

This theory of the pathogenesis of the disease did not take into
consideration the incongruity of assuming a special connective tissue
vulnerability in the liver as compared with that of other parenchymatous
organs. No other organ was known to suffer an inflammatorj' disease
in its supporting structures with secondary destruction of the paren-
chyma. Indeed, it is a well known law that a tissue is vulnerable in
direct proportion to its specialization.

Furthermore, pathologists of that day did not understand the regen-
erative powers of parenchymatous organs and particularly those of
the liver. Ponfick " was the first pathologist to experiment on the
regenerative power of the liver. He found that as much as four-fifths
of the liver of dogs could be removed with subsequent regeneration to
normal size. Considering that the liver normally has an enormous
surplus of tissue as a margin of safety, its regenerative power after
experimental extirpations is but another manifestation of its marvel-
ously large margin of safety.

In 1894, Marchand ^* reported a case of acute yellow atrophy
terminating in "multiple nodular hyperplasia." The patient, a woman,
aged 28, was brought to the hospital moribund, therefore an accurate
history was not available. It was ascertained, however, that she had
had "catarrhal jaundice" for several months, during part of which
time she had a swollen abdomen and legs, and that in the last few days
she had become suddenly and seriously ill. The skin was yellowish
and there were some small ecchymoses in the conjunctiva.

At the necropsy the liver was found greatly reduced in size, with
a hobnail surface, the eminences ranging in size from that of a pea to
a cherry. The elevated areas were pale reddish yellow in color, the
depressions dark red. On cut section areas of yellowish liver tissue
were seen with a reddish structure intervening and predominating, the
former elevated, the latter depressed. Acinus markings were not to
be seen in the nodules. Marchand gives a critical analysis of the
pathologic anatomy and histologj' of this liver, reaching the conclusion
that the process could not be a cirrhosis, because of the nature of the
dark red structure between the nodules of liver cells, a technical study
that must be read in the original to be appreciated. His conclusion
was that he had before him "a liver presenting changes which can leave
no doubt that we are dealing with the residua or results of a so-called
acute yellow atrophy." He mentions a similar case published by Klebs,
which "showed anatomically in extenso the picture of a typical red
atrophy of the liver whereas the clinical course of the disease had
been that of a chronic, possibly recurrent type of the disease."

17. Ponfick, quoted by Kretz : Verhandl. d. Deutsch. path. Gesellsch. 8

18. Marchand: Beitr. z. path. Anat. u. z. allg. Path. 17: 1894.

488 ARCHIVES OF IXTERXAL MEDICIXE

Marchand's concluding paragraph freely translated, states that
we have

the very unusual example of a yellow atrophic liver which has reached an
unusually high degree of regeneration, or better still, reproduction. It is not
unlikely that under more favorable circumstances a greater development of
regenerative changes could have occurred ; never, however, could complete
regeneration of structure be attained. If we could imagine the regeneration
I have described still further advanced, we would then have the picture of
a large lobule cirrhosis of the liver such as we see in childhood.

Meder," his assistant, published six or seven cases very closely
resembling Marchand's. MacCallum "" reports similar changes of a
reproductive nature occurring after acute yellow atrophy. -"•■'

Kretz -' was the first to coordinate these facts, namely, the regen-
erative changes in the liver, as ascertained by Ponfick experimentally
and by Marchand pathologically, and the close resemblance of Mar-
chand's and similar cases of the end stages of acute yellow atrophy, on
the one hand, with fully developed cirrhosis of the liver on the other.
He enunciated the present day conception of the pathogenesis of cir-
rhosis of the liver, for which MacCallum,^^ the greatest American
student of cirrhosis, gives him full credit. Yet Ackerman ^' was the
first to express the belief that the disease was primarily a parenchy-
luatous process.

The primary process in cirrhosis of the liver is a parenchymatous
change as in actue yellow atrophy, but not so rapid; slowly the paren-
chyma is destroyed — a little today, a little more tomorrow — inumerable
minimal attacks, continuing from onset to termination, with periods of
quiescence intervening. While the degeneration and necrosis of cells
is progressing, regenerative mitoses are occurring concurrently, at first
only among the liver cells, but later the biliary ducts, genetically closely
related to liver cells, partake in the efifort to replace destroyed liver
cells by metaplasia, perhaps after the regenerative powers of the liver
cells have become exhausted or cannot maintain the pace set by the
advancing disease. Periods of complete quiescence may result in resti-
tution to nearly normal liver sufTiciency. Alternating periods of progress

19. Meder: Beitr. z. path. AwM. u. z. al!g. Path. 17: 189-4.

20. MacCallum: Johns Hopkins Hosp. Rep. 10: 1903.

20°. Since this paper was submitted for publication, I found a liver of this
type at necropsy in the body of a young man, aged 22. Briefly, the clinical his-
tory was that he had had jaundice for five years, which was associated with
ascites and edema, marked weakness and frequent bowel movements. The
mode of onset could not be determined definitely but was more or less acute.
The clinical diagnosis had been chronic pancreatitis. At necropsy a nodular,
hyperplastic liver was found. The immediate cause of death was a terminal
peritonitis. The case will be reported in full later.

21. Kretz: Wien. klin. Wchnschr. 13: 1900; u. Ergb. d. allg. Path. 8: 1904.

22. MacCallum: J. A. M. A. 43:480 (Sept. 3) 1904.

23. Ackerman : Virchows Arch. f. path. Anat. 115.

EPPLEX— LIVER CIRRHOSIS 489

and .quiescence maj' continue for yeats until finally the patient succumbs
or, perhaps, in rare cases the disease may become permanently quiescent
while the liver is still capable of meeting the metabolic needs of the
individual.

So it is that the parenchyma dies and is replaced by connective
tissue, new parenchyma replaces the old and is again destroyed to be
replaced by other new parenchyma; constant repetition of destruction
And regeneration resulting in a completely rebuilt liver. In advanced
cases not the slightest vestige of the original tissue remains; even the
circulatory system has been rebuilt.

If we are ever permitted to use comparisons and analogies, it seems
justifiable here. So we may think of cirrhosis of the liver as the result
of minute attacks of something closely resembling acute yellow atrophy,
at least in the nature of its action. Acute yellow atrophy is no more a
definite clinical entity than is Laennec's cirrhosis; both may be pro-
duced by a variety of causes. One should think of the two diseases
as being analogous, one acute, the other chronic, differing only slightly
in the nature of the pathologic process, but very markedly in degree.
One is always acute and nearly always fatal ; the other is chronic and
likewise nearly always fatal. Occasionally the patient lives long enough
after an attack of acute yellow atrophy to produce regenerative
phenomena of sufficient degree to make the liver closely resemble that
of the chronic process, in which the patient always lives long enough
to e.xcite regeneration.

The origin of the new or regenerated cells appearing in the nodules
is manifestly from the surviving liver cells assisted by the bile ducts.
The objection sometimes raised that it would be impossible for cells
to regenerate in the presence of a circulating noxious agent capable
of destroying mature cells is scarcely valid. There are many known
instances of such occurrences. The further objection to this belief,
that we do not see partially destroyed cells, is not true. We do see
them, rarely it is true, but this is a very chronic disease ; cellular death
is hastened but not to a degree to be a ])redominating feature easily
found in any given slide. The same may be said of another objection
sometimes made, that we do not see mitoses. We do see them
occasionally, especially early. MacCallum -'- answers this argument by
referring to the diflferences in acute and chronic nephritis. Mitoses are
seen in acute nephritis relatively frequently. Regenerative changes
are perfectly evident macroscopically and are generally accepted in
chronic nephritis but they are not sufficiently in evidence to be dis-
covered microscopically. They are found only in acute cases or acute
exacerbations.

This comparison of acute yellow atrophy and cirrhosis at once
brings two diseases of obscure origin into close relationship clinically ;

490 ARCHIVES OF IXTERSAL MEDICI XE

indeed, their behavior is often quite similar, except in the matter of
time. It serves the further purpose of making ihe pathogenesis of the
chronic form of the two diseases, cirrhosis, quite clear.

MIXUTE .^X.XTOMV OF LIVER

As a prerequisite for an understanding of the architecture of the
cirrhotic liver, a clear conception of its normal architecture is essential.

Early in intra-uterine life the liver consists ot two dense meshworks
of capillaries derived from the portal vein and hepatic artery converg-
ing into the hepatic vein, within which are embedded the liver cells.
Essentially the liver is a radiation of two sets of capillaries arising
from a common point, the portal fissure, converging into a second
common point, the hepatic vein. Since the liver is usually regarded
embryologically as an epithelial bud arising from the primitive foregut,
therefore essentially an epithelial structure, an adequate conception of
its complicated vascular architecture cannot be obtained. Indeed, it
is just this that leads to the usual misconception of the anatomy of the
lobule or acinus, described as it is, as a rounded or polyhedral mass
of radiating cells with a central vein, surrounded by a connective tissue
capsule. There are actually no such glandular lobules in the liver each
separate and distinct from its neighbor. The liver is essentially a
vascular skeleton in the interstices of which is a mass of cells, just like
the flesh on a skeleton. Hence, all parts of the liver tissue are not
only in contiguity but in direct continuity with all other parts without
complete connective tissue separation or segregation into lobules.

The efferent blood flows into the ultimate hepatic vein through
its various collaterals, these branches collecting the blood from different
parts of the liver by a system of "continental divides" which makes the
separation into lobules or acini as we must understand them. These
"continental divides" between acini are theoretically hard and fast
divisions but histologically and physiologically they are neither visible
nor demonstrable, in fact, they may vary with physiologic necessities.
At various points this parenchymatous mass or theoretical lobule is
penetrated by branches of the portal vein and hepatic artery, clothed
by a thin coat of connective tissue in which are also enclosed the bile
ducts. It is the multitudinous, almost ubiquitous, ramifications of
these vessels that give the appearance of well defined rounded lobules
.separated one from the other by connective tissue, since they tend to
course along the "continental divides." It is an artefact pure and
simple produced by the necessity of making thin cross .sections for
minute study. What is .seen are cross sections of branches of the
larger real acinus or lobule, which may or may not contain a branch
of the central vein. These cross section lobules are not isolated
anatomically one from the other, and still less so physiologically.

EPPLES— LIVER CIRRHOSIS 491

Physiologically, the passage across the "continental divide" from one
lobule to the other is not marked by any boundarj' whatever. That
which is often seen and usually interpreted as the boundary between
lobules is really a "tunnel" or path within which the portal vein and
hepatic artery traverse the lobule. Such paths are, however, some-
times really located between lobules because it is a vantage point for
the distribution of blood or cellection of bile.

In order to have a simple comparison so that the structure of the
liver, as a whole, may clearly be visualized in a brief description, it
may be regarded as consisting of a vascular skeleton, for the first half
of which the lower half of a pine tree is used to represent the hepatic
vein and its branches, the pine needles representing the ultimate or
so-called central veins of the lobules. If the parenchyma of the liver
can be imagined to have been poured into the lower half of the tree.
like batter into a mold, filling out all spaces in the interstices between
the branches, the best concept of the gross, and at the same time
histologic, relationship of the parenchyma to the hepatic vein would
be obtained. If one could imagine this artificial, half-finished liver
to have been completed by introducing at some point on the upper
surface another lower half pine tree so inserted into the original mass
that its branches fit and dove-tail between the branches of the original
tree, the second representing the portal vein, a reasonably understand-
able concept of the relations of the hepatic vein, parenchyma and
portal vein will be gained. It would, at the same time, permit one to
accept the usual concept of the liver lobule or acinus and at the same
time shatter it by showing that no lobule is entirely independent and
separate from its neighbor. If one could still further imagine these
two trees fitted one into the other so accurately that not only the
branches would fit between each other, but the needles on the branches
of one fit between the needles on the branches of the other, one
would have a conception of the intimate relationship of the ulti-
mate branches of portal and hepatic veins.

In order to understand the relationship which the hepatic artery
and biliary ducts bear to the whole, one needs to return to the moment
when the second half tree was introduced into the half finished liver
and imagine a dense net-work of vessels having been wound about
the branches of the second half tree, one red to represent the hepatic
artery, the other green to represent the biliary ducts.

Obviously a description of this sort is purely artificial and does not
consider the embryologic facts, yet it is far more nearly correct than
the description given in courses on histologic anatomy, which also
clearly ignore the embryology. In fact, most slides for histologic study
are derived from the liver of the pig, because it is peculiarly rich in

492 ARCHU'ES OF I.XTfiRX.IL MEDICIXE

connective tissue. This makes the so-called lobule or acinus appear
especially distinct, particularly for teaching purposes, and erroneous
teacliings at that.

The plan of building up an acinus about a single unbranched, cen-
tral (efferent) vein in this manner and using it as an index of a
typical lobule, surrounding it with columns of liver cells and then
completing the lobule as it were, by pasting the afiferent portal vein and
hepatic artery on the finished lobule without showing its relationship
to its neighbors is irrational and leads to false notions of simplicity.
It would seem far better were we taught from the very outset how
complicated the structure and relationship of the lobule really are.
And yet the behavior of the liver in certain circulatory disturbances,
particularly high grades of passive congestion, make it appear that
these artificial structures or lobules functionate or behave more or less
as structures of some independence.

Mall injected livers through the hepatic vein and after corrosion
demonstrated that it branches very freely, that there is no such struc-
ture as a separate terminal central vein for each lobule, but rather that
the peripheral or terminal branches anastomose freely. Reconstruction
of the lobules about these veins produces an exceedingly complicated
outline, not at all rounded or polyhedral as is usually taught. Such
areas susceptible of apparent demonstration in slides are artefacts, as
previously stated, produced by making cross sections of small branches
of the lobule. The real lobule is a considerable mass of liver tissue
enveloping a radical of the hepatic vein and its branches, penetrated and
"tunnelled" by Glisson's capsule with contained vessels and ducts, so
that afferent blood is introduced into it at many points. Such a lobule,
if torn from its site, would be shaped roughly like an irregular pine
tree, or, better still, a misshapen pine tree. It may have some branches
or processes extending far afield, it may be bent on itself, or show
marked twists or bends. Most important of all it would not "shell out"
of a connective tissue capsule, because it has none. Its margins or outer
surface would be ragged for there is no sharp demarcation between
lobules, therefore, it would be torn from its surroundings. Moreover,
its communications with neighboring lobules would be shown by the
mouths of gaping veins which had. when in situ, communicated or
inosculated with those of nearby lobules. The "water-sheds" between
areas of the same lobule tributory to one or another branch of its
central vein are imperce])tible and unmarked, indeed they may shift
depending on secretory and circulatory exigencies. Exactly the same
may be said of the greater "continental divides" between distinct
lobules.

While this, in its essence, is not fundamentally different from the
usual idea, it is of value when the finer pathologic anatomy of the

EPPLEX—LIIER CIRRHOSIS 493

liver is under consideration, especially when an attempt is made to
explain the long strands of passive congestion in the cyanotic liver or
the long connective tissue strands of cirrhose cardiaque.

PATHOLOGY

The cirrhotic liver as it is seen at the necropsy presents certain
changes, usually a diminution, occasionally an increase, in size, hobnail
surface, and increased consistency and stiffness.

Examining the cut surface, three types of tissue, in varying relative
quantities are found: (a) normal liver tissue; (b) nodules of liver
tissue, lighter in color, varying in size from that of a millet seed to
that of a large bean or occasionally a walnut, and (c) connective tissue.

The entire surface has a tawny to brown color, varying with age
and degree, which led Laennec to give it the name cirrhosis. This color
is produced by a histologic finding of considerable importance to be
described later. No particular naked eye changes can be noted in
the smaller blood vessels, except that the small dark area in the center
of the acinus is entirely absent in the nodules, that is to say. the usual
index of the acinus, the central vein, is absent.

Closer study of these nodules shows that they vary in size, are of
paler color than normal liver tissue, and that they are likely, on cut
section, to be elevated slightly above the level of surrounding structures,
be it normal liver or connective tissue. They are not always rounded
or oval but may at times show several rounded eminences as though a
cluster had been cross sectioned. Examination of these nodules for the
position of the eft'erent lobular (central) vein shows a very wide
variation from the normal. It is no longer central but eccentric, often
j)eripheral and occasionally actually beside the nodule, gathering its
blood from the body of the nodule by exceedingly fine capillaries,
whose walls correspond to Kupfer's cells. This can have but one
meaning, namely, asymmetric destructive processes and, as a corollary,
asymmetric regenerative processes. The nodules of lighter colored
liver tissue, now universally accepted as regenerated liver cells, .show
a total absence of the acinal markings. The smooth continuity of
liver tissue has disappeared; the nodules are isolated one from the
other by strands of connective tissue, although clusters are sometimes
seen. These clusters are usually regarded as being produced by bud-
ding from a parent mass rather than by a confluence of several groups
or masses. Sometimes the nodules take up most of the cut surface;
at other times almost the whole organ is composed of connective tissue
with very small occasional nodules of liver tissue, often stained a
deep green (bile stasis).

Further study of the blood vessels in any cross section shows that
the larger vessels are closer together than normal. This corresponds
to the general reduction in the mass of the liver.

494 ARCHIVES OF I.XTf.RXAL MEDICINE

The color of the cut surface is variable depending on the age of
the patient as well as of the disease. The older both are, the darker
the brown color. If there is much fatty degeneration or infiltration,
there is a distinct yellow color, especially is this true of the nodules.
Extensive hemosiderosis is the cause of the rusty or tawny color that
gave the disease its name. Occasionally in the presence .of jaundice,
the liver is green in color.

The connective tissue is tough, present m varying quantities,
depending on the degree of the disease, and in color varying with that
of the liver in general. Where the connective tissue has shrunken
considerably, it may be very white and shiny, but when so shrunken it
is particularly likely to be associated with jaundice and to be dark
green. The capsule, in general, follows the connective tissue in
appearance.

The areas of normal liver tissue, if any remain, differ very little
from the tissue of a healthy liver. It shows the same brownish liver
color, with .soft texture, and if there be sufficient passive congestion,
which is nearly always true, the acinus markings are clearly visible.

If attention be directed to the histologic picture, one finds an
extremely varied type of cell as well as a very unfamiliar general
structure. The Uver cells may show the following variations (though
not all in any given slide) : (a) Normal cells in the familiar radial
arrangement; (b) normally arranged cells in various stages of retro-
gressive changes chiefly fatty degeneration, also parenchymatous
degeneration and possibly actual necroses. These are original liver
cells in process of destruction; (c) among the above normally arranged
cells may be found large, very pale cells with unusually clear proto-
plasm, occurring singly or in groups, sometimes among the degenerat-
ing cells, sometimes wholly surrounded by connective tissue. Since
there is no evidence of circulatory obstruction such as engorgment of
capillaries, these cells could not have been snared off' by contracting
.strands of connective tissue. Their clear, pale protoplasm with larger
cell bodies at once proclaims them as newly formed cells, regenerated
cells, if you please; (d) very large groups of the .same type of cell,
arranged more or less in parallel columns but entirely without radial
arrangement or central vein, with an eccentrically placed vein.
These represent nodules of regenerated cells with a new type of
vascularization; (e) greatly increased numbers of bile ducts in the
connective tissue surrounding the nodules; (f) among the cells of (c)
and (d) may be found cells in various .stages of degeneration.

It was with a definite purpose in mind that I described the normal
liver structure in terms of its circulatory elements. I will consider
more carefully the circulation of the cirrhotic liver.

EPPLEX— LIVER CIRRHOSIS 495

During injection experiments on cirrhotic livers carried out by
Kretz,-^ or under his direction, the following observations were made.
If the injection mass was introduced through the portal vein it would
appear in the hepatic vein at a time when large areas of hepatic
tissue were still uninjected. These uninjected areas proved to be the
larger nodules of newly formed liver cells as well as the small groups
of isolated liver cells previously mentioned. If the hepatic artery
was injected by a celloidin mass of different color it would be dis-
covered that it was exactly these areas that would be injected. It was
also noted that the hepatic artery was hypertrophic in such cases.

But one conclusion can be drawn from these findings, the regen-
erative process is directly under the nutritional influence of the hyper-
trophic hepatic artery. The portal vein continues to supply the original
liver tissue, which has escaped the destructive influences of the disease
and possibly some of the newly formed or regenerated masses of
cells slightly or indifferently. Since the new groups of cells are drained
chiefly by newly formed eccentric "central veins," a new communication
between the hepatic artery and the portal vein is established via
branches of the hepatic vein. Under the normal circulatory relation-
ship between hepatic artery and portal vein, the latter is not embarrassed
by the greater pressure of the hepatic artery, but with the opening of
new channels through the regenerated nodules the anastomosis is much
more extensive and free. This, plus the hypertrophy of the hepatic
artery, leads to definite embarrassment of the portal circulation. It
is like making an anastomosis much greater in size than the usual
capillar}' transition between an artery and a vein, at the same time that
the artery in question has had an opportunity to hypertrophy. Venous
stasis and edema must result. This is probably the real reason for
portal stasis and ascites, certainly when they appear before connective
tissue production has taken place in the liver. The fact that injections
of the portal vein in cirrhotic livers fail to fill large areas of liver
tissue, newly formed to be sure, but liver tissue nevertheless, demon-
strates the presence of an internal collateral circulation that is never
mentioned when the collaterals of cirrhosis of the liver are described.

It has been shown by Ponfick that large masses of liver tissue may
be resected in dogs only to have a regeneration take place adequate for
the physiologic needs of the animal, plus a large margin of safety.
The question arises whether this internal collateral circulation or
anastomosis of the portal vein around rather than through the newly
formed nodules in cirrhosis of the liver, is not the leak which is
responsible for the hepatic insufficiency in this disease. To put it
another way, is it not because the portal blood is carried past the
nodules of regenerated liver cells rather than through them that insuf-

24. Krctz: Yerhantll. d. deutsch. path. Gesell. 8.

496 ARCHIl'ES OF IXTERXAL MEDICI XE

ficiency results? The liver is doubly insufficient, first by reason of its
parenchymatous loss, but also because the portal blood fails to reach
the parenchyma in adequate quantities by reason of both internal
and external collateral circulation. In spite of the parenchyma loss
had the circulation been rebuilt properly, relative hepatic sufficiency
would have prevailed much longer than is usually the case.

HISTOP.-\THOLOGV

Liver. — The microscopic picture is essentially one of disorder and
disarrangement. In a well marked case there will be seen extensive
areas of connective tissue arranged in strands or streets irregidarly
disposed, crossing and recrossing each other. In the meshes will be
seen masses of liver cells in most unusual and disorderly arrangements.
The cells are disposed more or less in parallel columns, but there is no
evidence of systematic, much less systematic radial, arrangement.
Efferent veins will not be found in these masses, or, if so, they will
be eccentrically placed or in still other cases be found beside the cells.
The portal vein will not present the normal close approach, but will be
distant in the connective tissue strands in which it probably passes
to. the side of the nodules, rather than to or through them. The hepatic
artery is seen in the connective tissue strands in a loose relationship
to the portal vein. That it supplies the nodules with their afferent
blood, as maintained in Kretz, cannot be demonstrated histologically.
It is an assumption based entirely on his injection experiments. The
capillaries in the nodules lying as they must between the columns of
cells, have become intricate networks or labyrinths from which the
efferent blood has great difficulty in escaping.

Occasionally, quite normally arranged liver lobules, with radiating
cell columns — central veins and portal vein in normal contact — may
be found, residuums of the original liver structure. Degeneration or
necrosis, if seen at all, is most likely to be found in these lobules, though
lesser retrogressive changes, especially fatty degeneration, are often
seen in the cells of the regenerated nodules.

In the connective tissue strands are seen inumerable freely branch-
ing ducts, lined with cuboidal cells with deeply staining nuclei. These
are newly formed bile ducts. They probably possess a double function.
They are making unusual efforts by proliferation and branching to
search out and assume contact with the new nodules for purposes of
bile drainage, and, in addition they are probably concerned in the for-
mation, in small part at least, of new liver cells, for they are genetically
closely related to them. MacCuilum has demonstrated this transforma-
tion several times in hi.stologic slides by showing their transition and
union with the liver cells in nodules or groups of newly formed cells.

EPPLE\— LIVER CIRRHOSIS 497

The connective tissue strands are irregular in size and vascularity.
The variations in size are dependent on the amount of destroyed liver
tissue represented, that is to say, a large thick strand represents the
collapsed skeleton of a large area of destroyed liver, a small strand less.
While the increase in the bile ducts is, in part, absolute, some of the
increase is relative, the collapse bringing original ducts closer together.
The same is true of the increase in blood vessels — some are newly
formed, others have collapsed with the skeleton into closer proximity.
Often one can see indications of the lobules (their number, size, etc.)
that have been destroyed in an area of connective tissue; it is almost
as though the parenchyma had dropped out and the skeleton collapsed.
The connective tissue is increased relatively as well as absolutely.
Many round cells and wandering cells of all types are found in these
areas.

Hemosiderin deposits in cirrhotic livers are extremely frequent ;
when diligently searched for. are found in more than half the cases. In
studying old slides in our possession, derived from cases of cirrhoses of
varying degrees, it was present in all of the more advanced cases. It is
true that in some cases special stains may be necessary, yet when search-
ing carefully, minute deposits will be found either in the connective tis-
sue or the liver cells themselves. It is because it is not so overwhelmingly
present as to be immediately striking that this finding is not more often
described. Just why this should be found in a degenerative disease of
a parenchymatous organ has been the subject of much controversy.
It was to be expected during the early days of the literature of cir-
rhosis that it would be associated with and regarded as dependent on
hemorrhages in the gastro-intestinal tract, absorption being through
the portal system with subsequent deposit in the liver. Kretz -^ was the
first to set up the hypothesis that this pigment was liberated by destruc-
tion of red blood corpuscles in the liver, probably by the action of a
circulatory toxin, possibly a known chemical such as alcohol. Bleich-
roeder -" has written very exten.sively on this phase of the pathology
of cirrhosis.

To forestall any misapprehension we are not, when making these
statements, thinking of cases of so-called bronze diabetes but of portal
cirrhosis.

Spleen. — The spleen is very much enlarged, often it is much larger
than the liver. There has been much controversy over the cause of
this enlargement. Most of us were taught, and it is still believed by
some, that this enlargement is due to the passive congestion as.sociated
with cirrhosis of the liver. There is, however, much accumulated
evidence that there is a distinct pathologic process in the spleen typical

25. Kretz: Beitr. z. klin. Mc.l. u. Tlierap.. 1896.

26. Bleichroeder: \'irchows .\rcli. i. path. .Xnat. 177.

498 ARCHJl'ES OF IXTERXAL MEDICINE

of cirrhosis of the Hver. There are, for example, many cases of
cirrhosis in which the spleen is greatly enlarged long before there is
ascites, that it to say, long before passive congestion has occurred.
Leichtenstern -' was the first to call attention to these cases, indeed,
he regarded them as pre-cirrhotic and called them by the name "Pre-
cirrhotic Splenic Tumor." However, he is scarcely justified in the
conclusion implied by the name, that the spleen is diseased before the
liver, for cirrhosis is an insidious disease, it is always quite advanced
before a diagnosis is made; in fact, the clinical manifestations of its
early stages are entirely unknown. The splenic enlargement is always
considerable, it is much greater than that of cardiac decompensation.
It is not nearly as hard, either to palpation of the intact organ or when
its cut surface is examined. The spleen of passive congestion gives
the impression of being stuiTed full to bursting, yet on the cut surface
the pulp does not pout, nor can it be scraped off, both evidences of
great increase in connective tissue. The spleen of cirrhosis is slightly
softer, more like that of a subacute infection, its pulp usually pouts on
the cut surface and can be scraped oti' readily. Its color is lighter than
either the normal or the cardiac spleen, often having a slightly grayish
tinge as though milky water had been poured over it, or it may have
a rusty, tawny tinge. Bleichroeder found a considerable difference
in the specific gravity, the spleen of cirrhosis on an average having
a specific gravity of 1.059 against 1.044 for passive congestion.

Histologically, the .spleen of cirrhosis shows a slight congestion
as compared with the normal ; presumably this is passive, but it does
not approach that- of the spleen of chronic cardiac decompensation.
Furthermore, there is a great increase in the various types of wandering
cells and lymphocytes. \'arious writers have descriljed circumvascular
proliferations which at times penetrate the blood vessels, one of them
(Bleichroeder '^), attributes to these a special pathogenetic significance,
believing that they may become dislocated and swept from the spleen
to the liver, there producing either focal necroses or setting up the
inflammatory process that eventuates in the destruction of normal
liver tissue.

Finally, there is a great deposit of hemosiderin and various other
iron pigments, chiefly in the connective tissue of the spleen but also
in the pulp and in the various wandering cells. The source of this is
not yet clearly understood, but the as.sumption that it arises from
extensive destruction of red blood corpuscles is as justifiable here as it
is for the origin of the same deposits in the liver. Often it gives the
spleen a slightly rusty or tawny color.

27. Leichtenstern, quoted by Naunyn : X'erhandl. d. Dcutsch. p.ith. Gescll. 8.

EPPLEX—LIJ-ER CIRRHOSIS 499

ASSOCIATED PATHOLOGY

Embarrassment of the portal circulation leads to various distur-
bances. First among these is ascites. This is often of extreme degree.

Passive congestion leads to a catarrhal pseudo-inflammation of the
gastro-intestinal mucosa. The mucosa of the stomach is intensely
injected, edematous and at times shows submucous extravasates. In
well marked cases these may be very great and lead to distressing
hematemesis. The mucosa is covered by an exceedingly tenacious
mucus, often containing exfoliated epithelium and red and white blood
corpuscles. The mucosa and serosa of the stomach and sometimes of
the bowel are often tinged a faint brown by hemosiderin deposits, a
finding attributed by some to the results of passive congestion and
extravasates. by others to processes identical with those causing similar
deposits in the liver and spleen. These are never seen in the stomach
in the passive congestion of cardiac decompensation.^*

Collateral venous hypertrophies or varicosities reach extreme
degrees and represent efforts of the portal blood to reach the right heart
by routes other than the normal. They are most marked in the anasto-
mosis of the portal circulation with the systemic veins at the lower
end of the esophagus and between the portal and systemic veins in the
lower rectum via the hemorrhoidal veins and via the veins of the
abdomen, the so-called caput medusae. It might be well again to
mention the collateral circulation around the nodules of regenerated
cells within the liver itself.

Obscure Indefinite Findings. — There is always a very low grade of
secondary anemia present, the nature of which is not well understood,
whether it is dependent on the same noxious agent that produces
cirrhosis itself or on associated digestive and consequent nutritional
disturbances is not yet settled. Possibly, it may be regarded as an
abortive anemia of one of the types often associated with cirrhosis.

The heart usually shows slight myocardial changes of a degenerative
nature. Whether this is dependent on the toxins producing the disease
or on the toxins produced by the disease is not clear.

Most cirrhotic patients are slightly jaundiced, not the clear-cut
yellow color seen in biliary obstruction, but a pale, almost imperceptible
color, which seems to give an undertone of yellow to all light colored
surface tissues, particlarly the sclerae and the skin of the covered
parts of the body. The color is very much like that of low grade
sepsis. The urine contains large amounts of urobilin and urobilinogen.
The conclusion that this color is due to urobilin seems justifiable. True
jaundice occurs in some cases, especially when contraction of the
connective tissue has embarrassed biliary circulation.

28. Wagner: .^rch. f. klin. Med. 34.

500 ARCHIVES OF IXTERXAL MEDICIXE

Occasional!}- peculiar hemorrhages or ecchymoses are seen in the
conjunctiva, the retina and the skin, especially where the texture is
soft. These and the clinically noted attacks of epistaxis are as yet not
satisfactorily explained. They remind one of the dyscrasias of certain
blood diseases.

There are several notations in the literature of transformation of
the normal into red bone marrow, especially in the femur.-' Bleich-
roeder -" maintains that this is true in the majority of cases, he having
found changes in the upper end of the femur in twelve or thirteen
cases examined for this change.

In certain instances, particularly those in which there is a very large
spleen, there are very marked sclerotic changes in the splenic artery
and vein and also in the mesenteric ^eins. Just how this is brought
about is not very clear. It is most often present in Banti's syndrome
and could be used as a particular argument against Banti's contentions.
If the toxin producing cirrhosis originates in the spleen alone, why does
it produce sclerotic and hyalin degeneration of identical nature in the
mesenteric veins? Just this finding is particularly suggestive of an
enteric origin for the poison though of course it does not explain why
the splenic artery and vein should be similarly involved.

Tuberculosis, or rather, isolated occasional tubercles in the peri-
toneum, and the cirrhotic liver itself are frequent, apparently accidental
findings. They are usually explained as having developed by virtue of
the generally lowered resistance of the patient, and are believed to be
terminal or subterminal events. Others see in them a close relationship
etiologically. This phase of the subject has been given much study
but without definite conclusions. Some writers have made a separate
classification of these and called them tuberculous cirrhoses, a step
scarcely warranted at present.

REL.\TION TO OTHER DISEASES

Banti ^^ described what he believed to be a syndrome with a definite
pathologic complex, worthy of separate classification. He set up
certain requirements that make his definition exceedingly complex,
indeed it is often regarded as impossible. There must be no known
etiology for the disease; there must be enlargement of the spleen
preceding the cirrhosis and anemia, which must be of a secondary type
with a low leukocyte count, but relatively a lymphocytosis; and the
disease must be separable into three stages: (a) splenomegaly; (b)
anemia; (c) cirrhosis of the liver with a.scites.

Banti himself does not always write in the same vein. More
recently ''' he maintains for the disease a characteristic pathology, par-

29. Zypkin: Virchovvs Arcli. f. path. .-Xii

30. Banti: Sperimentalc. 1894. j). 407.

31. Folia Haemal. 1:11. 1910.

EPPLEX-LIIER CIRRHOSIS' 501

ticularly in tlie spleen. He speaks of it as a "fibroadenie," which is
essentially a fibrosis of the malpihgian follicles situated around the
splenic artery which itself shows hyalin changes, and an extensive
fibrosis of the splenic reticulum. All of this he believes to be of
noninflammatory origin, because he has not found fibroblasts present.
There is hyalin degeneration of the capsule of the spleen, and cirrhosis
of the liver difTering in no respect from the usual picture, except in the
splenic changes mentioned. Blood pigments were not found in the
spleen or liver in his cases, hence he assumes that there was no blood
destruction. The red bone marrow of secondary anemia is uniformly
found. He assumes that there is an infectious agent producing the
splenic changes ("fibroadenie") which, in turn, cause the spleen to
elaborate a poison producing the liver cirrhosis. There is no known
analogy for this in all pathogenesis, it is manufactured to fit the case,
a reasoning from efTect to cause.

Banti admits that the diagnosis cannot be completed until the liver
is seen by the pathologist (or surgeon) and that the pathologist cannot
make the diagnosis without the clinical data, namely the history of
three consecutive stages, splenomegaly, anemia and cirrhosis. This is
merely another way of saying that the disease is essentially undiag-
nosticable.

In a general way, the literature may be divided into two classes
with reference to the acceptance or rejection of Banti's disease as a
clinical entity: (1) That of authors who accept it, usually clinicians,
with the notable exceptio.rt of Naunyn,^' and (2) that of authors who
reject it. generally speaking, pathologists, though they usually manifest
a more scientific conservatism and may be said to assume an attitude
of skepticism.

While I have never seen a case classifiable as Banti's disease, it
seems to me that there is no definite reason why it should at present
be accepted as a clinical entity. There is too much controversy con-
cerning it, and Banti is neither definite nor sure enough of himself.
Finally, if we accept Banti's classification, there is no such disease
as splenic anemia, for he does not permit even this to escape him,
maintaining that it is Banti's disease in the second stage before cirrhosis
has had time to develop.

Banti's contention that a poison is developed in the spleen has been
supported in a scientific manner but once, and never since confirmed.
Umber '^ made metabolic studies on a case and found that there was
disintegration of the blood albumins on a toxemic basis. After extirpa-
tion of the spleen this ceased. This is possibly analogous to some of
the experiments of Longcope, interpreted by him as anaphylactic
phenomena. This is not to be construed as an efifort to state that toxins

32. Umber: Ztschr. f. khn. Mt.l. 55: 1905.

502 ARCHIVES OF IXTERXAL MEDICI. \E

cannot originate in the spleen, it has not been proved in cirrhosis. On
the other hand, it seems proved in both splenic anemia and in hemolytic
jaundice by the results obtained by splenectomy.

Mayo ''^ arguing on purely clinical grounds, experience, if you
please, states that extirpation of the spleen in cirrhosis, when the spleen
is very large, is beneficial by reducing the volume of portal blood so
that the liver may again "carry on." He suggests, however, that it
may prevent "those irritants" ordinarily filtered out in the spleen from
reaching the liver as, for example, in splenic anemia. He offers no
proof of their existence. Again the alluring theory, that abnormal
splenic metabolic products are the cause, cannot be resisted by one
usually so matter of fact that theories have no place in his articles.

Every patient with cirrhosis of the liver is anemic or shows periods
of anemia, of the secondary type. Every case shows great variations in
its course, not only as to symptomatology but as to duration of the
disease. It would seem more rational then to regard Banti's syndrome
as one of the various clinical pictures under which cirrhosis may present
itself. Moschcowitz^* assumes this attitude with great emphasis, probably
because the two cases presented in his article showing Banti's syndrome
clinically, failed to show any evidences whatsoever of cirrhosis at
necropsy.

It strikes me that Banti has surrounded his definition with so many
conditions and modifications, some reasonable and some — e. g. his
contention that the etiology must be unknown — so unreasonable as to
make the disease purely imaginary. Dropping it will still leave a diag-
nostic pigeon-hole for all cases. Chronic cirrhosis with high grade
anemia and splenomegaly will describe Banti's typical cases. Splenic
anemia will cover those cases claimed by him to be in the second stage
before cirrhosis has had time to develop.

Diabctc Bronze. — Diabete bronze, or cirrhosis pigmentaire, is a
disease of obscure origin characterized by a cirrhosis of the liver like
that of portal cirrhosis with extensive visceral and cutaneous hemosid-
erosis, and diabetes mellitus.

Various theories of its pathogenesis are offered, the chief of which,
ascribe the primary role variously to the hepatic cirrhosis, to the
diabetes and to the hemosiderosis. Another somewhat differing etiology
is offered in the form of an hypothetical toxin the cause of all the other
pathologic changes. Under this characterization it is spoken of as
hemochromatosis.

It is an extremely rare disease, the literature offering only seventy-
five cases for study. In a general way they may be divided into two

33. Mayo: Ann. Surg. 68:183 (Aug.^ 1918.

34. Moschcowitz : J. A. M. A. 69:1045 (Sept. 29) 1917.

EPPLEX—LllER CIRRHOSIS 503

groups, one, those in which the cirrhosis and pigmentation represent
the less severe form, the other — probably an advanced stage — in which
diabetes mellitus has been added to the foregoing changes.

The pathology of the cirrhosis differs but little from that of ordinary
portal cirrhosis; the liver is usually larger than normal, death occurring
before atrophy can occur. Ascites is rare. There are no distinctive
features in the associated diabetes; it develops in the great majority
of cases and is the cause of most of the deaths. It may be due to
extensive hemosiderotic fibrosis of the pancreas.

Usually a moderate degree of secondary anemia is present.

IVilson's Disease. — The association of cirrhosis of the liver with a
definite disease of the nervous system described by Wilson in 1912 is
a most remarkable combination, an incongruity pathologically. It is
a progressive degenerative disease located in the lenticular nuclei occur-
ring in young adults, familial in occurrence, yet not hereditary. Its
chief manifestations are various motor phenomena of e.xtra pyramidal
origin with some mental symptoms. Atrophic cirrhosis of the liver is
constantly found. Aside from its remarkable association with a nervous
disease, it is noteworthy that it occurs in youth, is familial, that alcohol
is not implicated in its etiology, that ascites and other evidences of
portal stasis are absent, and that clinically the cirrhosis is not demon-
strable, even the small size of the liver has only rarely been
demonstrated clinically. Wilson characterizes the changes in the liver
as a multilobular or mixed cirrhosis.

CL.\SSIFIC.\TIONS OF CIRRHOSIS

So very much has been written in efforts to prove that all cirrhoses
are essentially the same process that we will venture a few remarks on
this subject.

Mayo^^ is the most arbitrary of all writers and brusquely divides
them into portal and biliary cirrhoses. The former is what is generally
known as Laennec's cirrhosis, but he also includes all forms of hyper-
trophic cirrhoses, whether of alcoholic or other origin. He does not
accept the view that the enlarged livers of certain types of cirrhosis,
often containing considerable fat, later become atrophic. He therefore
speaks of atrophic portal cirrhosis which he regards as the characteristic
response to concentrated spirits, such as gin, and to pepper excess, and
to poisons carried to the liver from the spleen. On the other hand,
he speaks of hypertrophic portal cirrhosis which is the characteristic
response to excesses of beer, and is associated with fat deposits. Biliary
cirrhosis is dependent on infection of the biliary ducts, and is char-
acterized by a large liver and a very large spleen. Mayo's classification
could be summarized thus: (a) Portal cirrhosis, (a) atrophic type
(Laennec); (b) hypertrophic type, {b) biliary cirrhosis, with an
enlarged liver.

504 ARCHirES OF IXTERXAL MEDICIXE

In his opinion, there is no pathologic or clinical basis for a separate
classification of Hanoi's cirrhosis. He believes that the disease so
designated is either an obstructive biliary cirrhosis or an hemolytic
icterus, in which there is work hypertrophy of the liver. Since hemo-
lytic icterus is very frequently associated with gallbladder pathology
(more than 60 per cent, of the cases), he believes that there are
combinations of the two. that is to say. a work hypertrophy due to
hemolytic jaundice plus biliary cirrhosis of infectious origin.

Pathologists still adhere to a much more elaborate scheme : ( 1 )
Laennec's cirrhosis (atrophic), including multilobular cirrhosis, nature
not clear, probably syphilitic. (2) Hanot's cirrhosis (primary biliary
hypertrophic cirrhosis). (3) Obstructive biliary cirrhosis. (4) Hepar
lobatum. (5) Cirrhose cardiaque.

A brief description of the nonportal cirrhoses and other diseases
sometimes confused with Laennec's type of cirrhosis will serve to
clarify some of the difficulties in classifying them.

Hanoi's Cirrhosis. — In 1876 Hanot presented for clinical and patho-
logic study a type of disease, la cirrhose hypertrophique avec ictere
chronique. He based his studies on four cases of his own and about
a dozen cases collected from the literature. The disease is characterized
by a chronic intermittently fiebrile course, with severe jaundice but
without clay-colored stools or ascites, with a very large, smooth liver
and a very large spleen. Histologically, there is a striking intra-
acinous development of connective, tissue. In later publications he
broadens his views considerably, so that it becomes somewhat difficult
to know just what he includes. Still later writings are even less clear,
and his original ideal type becomes confused. Perhaps he was led
afield by other" French writers who described various "forms" of
Hanot's cirrhosis, some accepting as a standard the presence of ititra-
acinous development of connective tissue, others described a capillary
cholangitis as the essential pathologic standard. As will be seen later,
they .were doubtless describing what is now called obstructive biliary
cirrhosis. In>answering these writers in the course of a long series of
polemical articles, he allowed himself to become confused until his
articles lost much of their clearness, and, like Banti, he did not seem
to know just what constituted a clear-cut type, an ideal of the disease
known by his name.

Today a somewhat broadened or modified view is taken as to what
constitutes Hanot's cirrhosis. There are two distinct tyi^es or forms:
First, those cases of cirrhosis in which there is a degenerative process
in the liver as.sociated with a toxemic jaundice. This descri])tion is
strikingly like that of Mayo, who uses it to deny Hanot's cirrhosis
a separate place in pathology, but terms it hemolytic jaundice witlr

EPPLEX— LIVER CIRRHOSIS 505

work hypertrophy. However, these cases are said by Kretz '^ to
develop an ascites if they Hve long enough, because of changes in the
texture of the liver with contraction or atrophy, therefore, they do not
conform to Hanoi's requirements of an absent ascites, and are not his
ideal type. These show no inflammation of the finer bile ducts but
there is a very fine intralobular almost intercellular distribution of
connective tissue breaking up the lobules into very small groups of
cells. Second, those cases of jaundiced hypertrophic cirrhosis in -which
there is an intense capillary cholangitis, a type brought forward by
Heineke ^'^ as the ideal type of Hanot. The connective tissue about the
biliary capillaries proliferates and secondarily enters the acini insinuat-
ing itself between the cells. It shows no tendency to contract and cause
."trophy of the liver, the feature which differentiates it from obstructive
biliary cirrhosis. Its etiologj- is clearly infectious. In both types
the liver is large, hard and smooth, often weighing as much as .S,000
gm. The spleen is also large and hard, much greater in size than in
atrophic cirrhosis. Doubtless many cases of Laennec's cirrhosis with
enlarged liver are erroneously described as being of Hanot's type.

Obstructive Biliary Cirrhosis. — Obstructive biliary cirrhosis is an
exceedingly rare type. The liver is small, very hard, like leather, and
is very dark green in color because of its etiology, chronic obstructive
jaimdice. Probably, one reason this disease is so rare is because
modem surgery does not permit an obstructive jaundice to exist long
enough to produce the consequent changes unless it be due to an
exceedingly slowly progressing carcinoma of the gallbladder. Histo-
logically, there is still normal acinal structure with intense bile stasis
and consequent destruction of liver cells. These are destroyed in part
by the biliary stasis and back pressure, but, perhaps, still more by the
consequent chronic infection which is always associated, either sooner
or later. A thick connective tissue mantle is formed about the biliary
ducts and capillaries, associated with the histologic phenomena of
inflammation.

Hcpar Lobatmn. — While often classified with the cirrhoses, hepar
lobatum does not properly belong here. It is essentially normal liver
tissue traversed by long, deep scars communicating with each other and
dividing the liver into adventitious lobes. They are the remains of
gunimalous proces.ses, and are often characteri.stically localized in the
left lobe of the liver. Sometimes it is a mere connective tissue or
scar-like membrane.

Multilobular Cirrhosis. — Multilobular cirrhosis is not clearly classi-
fied. The term is reserved by some writers for the end-result of
gummato.us hepatitis in children. The liver is small, roughly resem-

35. Kretz: Verhandl. d. Deutsch. path. Gesell 9.

36. Heineke: Beitr. z. path. Anat. u. z. allg. Path. 22.

506 ARCHIVES OF IXTERXAL MEDICIXE

bling that of Laennec's cirrhosis but not quite so tough. There are
relatively large pseudo-acini produced by connective tissue subdivision
of normal liver tissue in which true acini are present. Histologically,
the connective tissue strands are broad, but differ from those of -
Laennec's cirrhosis in that there are no evidences of inflammatory
infiltration, it is clearly a quiescent scar tissue. Biliary ducts show
no evidences of sprouting and there is little or no parenchymatous
regeneration. Normal acini clearly differentiate it from atrophic
cirrhosis.

Cirrhose Cardiaquc.—C\rr\\osc cardiaque is not a cirrhosis at all ;
it is a high degree of passive congestion in which the parenchymatous
cells destroyed by back pressure of the blood have been replaced by
connective tissue. It has been credited with a specific infectious
etiology because it is most typically found in cases in which the
cardiac difficulty was acquired in early youth especially in concretio
pericardii cum corde.

RESUME OF CLASSIFICATION

The following would represent a comparison of the orthodox
classification of pathologists with the abrupt of Mayo, indicating
the equivalents.

Pathologists' Classification Mayo's Classification

1 . Laennec's cirrhosis

2. Biliary cirrhosis 2. Hemolytic jaundice, with work

(a) Primary biliary Hanot \. Portal cirrhosis.
L Toxemic without hypertrophy.

biliary capillary in-
fection

n. Biliary capillary "j

cholangitis o. Biliary Cirrhosis.

(b) Biliary obstructive....]

3. Multilobular cirrhosis, probably an end-result of syphilis in
childhood.

4. Hepar lobatum — post syphilitic scars.

5. Cirrhose cardiaque.

In closing, I cannot resist the temptation to give e.xpression to a
thought, which is growing into a conviction, that when dealing with
cirrhosis of the liver we must regard it as closely allied to the diseases
of the blood, for the following reasons. Banti's disease is an accepted
blood disease, associated with cirrhosis of the portal type. Splenic
anemia if accepted as a clinical entity always suggests liver changes
to the clinician, a suggestion born of an involuntary association of the
two, based possibly on Banti's writings. Diabete bronze is a typical

EPPLEN— LIVER CIRRHOSIS 507

cirrhosis and diabetes plus a marked disturbance in the metaboHsm of
iron and a moderate anemia. From this to the hemosiderosis of ordi-
nary atrophic cirrhosis is not so far a cry as to be unheard, especially if
the latter be associated with an anemia. The jaundice present in most
cases of cirrhosis is urobilin jaundice, probably the same substance
which causes the yellow color in pernicious anemia. Some writers who
have made special search have found red bone marrow of the type
seen in the high grades of anemias. While it is trui? that this has not
been found by many clinicians, it is also equally true that very few
have searched for it. Finally, when we begin studying hypertrophic
cirrhosis we find such practical men as Mayo abandoning the term
hypertrophic cirrhosis and visualizing it as a work hypertrophy in
hemolytic jaundice, a disease which is unquestionably classifiable as
a disease of the blood.

THE ANTIDIURETIC EFFECT OF PITUITARY EXTRACT

APPLIED INTRANASALLY IN A CASE OF

DIABETES INSIPIDUS *

HERRMAXX L. BLUMGART, M.D.

BOSTON
INTRODUCTION

The combination in diabetes insipidus of insatiable thirst and
polyuria interfering with sleep and all the ordinary activities of life
naturally has stimulated numerous workers in the past ten years to
devise some method whereby the lives of the sufferers of this disease
could be made more tolerable. In 1913, Von der Velden ^ and Farini ^
demonstrated that the subcutaneous injection of pituitary extract
checked both the polyuria and the polydipsia. This observation has been
abundantly confirmed. Two features of hypophyseal therapy render
it still highly unsatisfactory. First, the effect is transitory, and, second,
hypodermic injection has always been essential and is inconvenient and
difficult for continued use of patients.

Even though the effect is transitory, however, if pituitary extract
could be introduced into the body in more frequent doses and in a less
inconvenient manner, great comfort would naturally result. Absorption
by wa}- of the gastro-intestinal tract has been tried by many observers.
Failure has atteftded practically all attempts to diminish the urinary
output by dried extract given by mouth. Motzfeld^ controlled the
diuresis in one case by feeding fresh posterior lobe of the ox. Christie
and Stewart * were, however, unable to confirm this result in their case.
Christian,'" in a case studied by him, found that pituitary extract intro-
duced in suppositories, by colonic irrigation, and in gumdrops which
were allowed to dissolve slowly in the mouth, failed to exercise any
antidiuretic effect.

In the case under observation only 0.005 c.c. of pituitrin "O" *
subcutaneously was necessary to effect a marked diminutinn in the

* From the Medical Clinic of the Peter Bent Brigham Hospital

1. Von der Velden. R. : Berl. kiln. Wchnschr. 50:2083, 191.^

2. Farini. A., and Ceccaroni, B, : Gazz. d. osp., Milano, 34:879, 1913; Clin,
med. ital., Milano 52:497. 1913.

3. Motzfeld: Endocrinologj- 2:112, 1918.

4. Christie and Stewart: Arch. Int. Med. 20:10 (July) 1917.

5. Christian: Med. Clin. N. America 3:849 (Jan.) 1920.

6. Pituitrin "O" is the aqueous extract of the posterior lobe of the pituitary
prepared by Parke, Davis & Company for obstetric use and is one-half the
strength of pituitrin ''S" prepared for surgical use.

- BLUMGART— PITUITARY EXTRACT 509

urinarj- output. Since so small an amount was effective and since cer-
tain methods had not been attempted, it seemed worth while to under-
take the following study.

REPORT OF CASE

G. C. (Xo. 31,569). a schooll)oy. ased 16. entered the Peter Bent Brigham
Hospital Nov. 28. 1921, complaining of thirst and frequency of urination.

Family History. — Father, mother, two hrothers and tive sisters are living
and well.

Past History. — Xegative, save for diphtheria at 3.

Present Illness.— PiUent felt perfectly well until three months before admis-
sion when he noted the rather sudden onset of polydipsia and polyuria which
gradually increased.

Physical Examination. — Patient was a poorly developed and poorly nourished
boy, with slight diffuse brownish pigmentation of the skin over the entire body.
Ophthalmoscopic examination was negative. .\ complete general and neurologic
examination, including perimetry, showed no other abnormalities.

Clinical Pathology.— Blood: Hemoglobin was 85 per cent. (Sahli) ; red
blood cells, 4,832,000; white blood cells, 11.300. A stained smear showed definite
achromia and slight anisocytosis. The blood Wassermann reaction on two
occasions was strongly positive. Blood sugar : 7,1 mg. per hundred c.c. Blood
urea nitrogen, 10 mg. per hundred c.c.

The blood Wassermann reaction of tlie patient's father was very weakly
positive.

Urine: Clear, pale, acid, without sediment; specific gravity, from 1.000 to
1.003; no albumin or sugar. Daily urine output varied from 6 to 9 liters.

Spinal Fluid: Clear and colorless; pressure, 155 mm. water; contained six
cells per c. mm. Globulin reaction was slightly positive. The Wassermann
reaction was weakly positive in 2 c.c. and in 1 c.c.

RoENTGEN-R.^v Ex.^MiN.^TiON : Stereoscopic plates of the skull showed no
evidence of any abnormality.

METHOD

The procedure employed was as follows : The patient was given a
diet containing a fixed amount of protein and salt. This precaution
was taken because it has been shown conclusively that the chlorid
and nitrogen intake influences the urinary output." Between meals the
patient was encouraged to take as much liquid as he desired. He was
not allowed to eat between meals nor was he allowed lemonade, coffee
or other beverages which might introduce confusing factors. The
water conteflt of the food, while not accurately determined, was kept
approximately uniform. Under this regimen, the patient's fluid intake
was 3,000 c.c. from 7 a. ni. to 7 p. m. and about 3,000 c.c. during the
night. In order to establish a standard curve of excretion, the 5.000 c.c.
intake was distributed evenly throughout the day, 200 c.c. being given
every half hour from 7 a. m. to 7 p. m. During the same period, the
urine was collected every hour. Under these conditions the fluid intake
and output were relatively uniform. .Single do.ses of pituitary extract
subcutaneously, intranasally, by rectum and by mouth, and of histamin

7. Oehmc and Oehme: Dcutsch. .\rcli. f. klin .\Ic<I. 127:261, 1914.

510 ARCHIVES OF IXTERXAL MEDICIXE

subcutaneously and intranasally were administered. The intake being
constant, the effect sought was a delayed excretion rather than a
diminution in the total urinary output. This procedure, in excluding
subjective factors and ensuring a more uniform rate of excretion
makes effects of lesser magnitude discernible.

The methods which gave any indication of influencing the urinary
output were then tested by placing the patient on unlimited fluids and
noting the effect over twenty-four hours.

0BSERV.-\TIONS

Effect of Pituitary Extract and Histamin. — The patient was put on
a fixed intake of 200 c.c. every half hour with instructions to void every
hour. On this regimen it was found that no antidiuretic effect was

X

X

x

500 Jll^ dL

±^^± : 's.

x' 1 1 T J. ^s_

^ ' I \ I i ^'

5 400 ^ L^B t^^

$ E

^ X -4i

« 1 ^r

"^300 : [ 5

^ i /

^ J X ^

J ^

<u 200 I //

c ^'''' Vv jT

i; t^ ^ UL

^ \ ^ g^_2

fOO J-,Xv ^3t

jZj^

X -^afe

-j- a-u'2i

0 ± +X

Chart 1. — The comparative antidiuretic effect of subcutaneous injection of
pituitrin "O" and intranasal spray of pituitrin "O" with fluid intake of 200 c.c.
every half hour. Solid line represents the effect of subcutaneous injection of
0.005 c.c. of pituitrin "O," dashes indicate the effect of 0.5 c.c. pituitrin "O"
sprayed intranasally at time noted by arrow.

produced by: (1)2 c.c. pituitrin "O" retained in the moutl\ for ten
minutes and then swallowed; (2) from 4 to 8 c.c. pituitrin "O" with
20 c.c. tap water introduced by rectum; (3) histamin, 1 c.c. of 1 : 10,000
solution injected subcutaneously; (4) histamin, 0.1 c.c. sprayed intra-
nasally; (5) tap water 1 c.c. sprayed intranasally. On the other hand,
a marked antidiuretic effect was produced by: (1) from 0.005 c.c. to
0.5 c.c. pituitrin "O" injected subcutaneously; (2) from 0.5 to 5 c.c.
pituitrin "O" sprayed intranasally (Chart 1 ; Table 1).

BLUMGART— PITUITARY EXTRACT

TABLE 1.— Effect of Administering Pituitary Extract and Histamin by
Various Routes

Trine per Hour

Alter Administration

Method
Xo Antidiuretic Effect:

2 CC. pituitrin "O"' by mouth

4C.C pituitrin "O" by rectum

I CC. histamin. 1:10.000. subcutaneously

Marked Antidiuretic Effect:

0.25 CC pituitrin "O"' subcutaneously

1.0 CC pituitrin "O" subcutaneously

0.8 CC oral pituitary extract, intranasally..

5 CC oral pituitary estract, intranasally. . .

In order to determine the efficacy of these measures in reducing the
twenty-four hour intake and output, the patient was encouraged to
drink all the water necessary for comfort. The fluid intake and output
were carefully measured. The same diet was continued. One and
five-tenths c.c. of oral pituitary extract was sprayed intranasally every
three hours on one day (Column 2, Chart 2), and every four hours

CC 1

Hvr^ Hi

<rr c c

sooo

n flooD

6or)n

!_.

n

(inno

4000

Anon

P.OOO

■

1

Ir.

2000

0

In

in

1

Il

n

/
1

0 / 0
IT

10 10

m m

/ 0 10
Y W

; 0 I

W 2

0

w:

CTiart 2.— Comparative effect of various measures on twenty-four hourly
fluid intake and output. I. Indicate.s fluid intake; O, urinary output. I.
Effect of lumbar puncture. II. Intranasal spray of 1.5 c.c. oral pituitary extract
every three hours. III. Intranasal spray of 1.5 c.c. oral pituitary extract every
four hours. IV. Phenyl salicylate coated tablets, 1.3 gm. V. Posterior lobe
pituitary pills (Burroughs Wellcome Co.) 0.13 gm. every four hours. VI. Swab
soaked in 1 c.c. pituitrin "O" inserted in one nostril and changed every four
hours. VII. Subcutaneous injection of 0.5 c.c. pituitrin "O" every six hours.
VIII. Histamin, 1 c.c. 1 : 10.000 solution ( Parke, Davis & Company) injected
subcutaneously every six hours.

512 ARCHITES OF IXTERXAL MEDICIXE

on the next day (Column 3), with marked decrease in urinary output.
A cotton plug soaked with 1 c.c. of pituitrin "O" was introduced into
one nostril at four hour intenals with essentially the same result
(Column 6). Pituitary extract was then withheld and after the
patient had returned to his usual intake and output level, phenyl
salicylate coated tablets (Column 4), and posterior lobe pituitary tab-
lets (Burroughs Wellcome Co.). (Column 5), were administered by
mouth.

The results are graphically represented in Chart 2.

Changes in Blood Concentration. — An attempt was made to deter-
mine what changes, if any, occurred in the concentration of the blood
serum before and after intranasal administration of pituitary extract.
The patient was allowed an unlimited intake of fluid. Two specimens
of blood were taken at hourly intervals, after which 4 c.c. of pituitrin
"O" was sprayed intranasally. Samples of blood were taken at the
end of one-half hour, one hour, two hours and three hours. The protein '
concentration of the blood was determined by the refractometer. That
this is both exceedingly sensitive and reliable as an index of blood
concentration has been demonstrated by Reiss.'

T.-\BLE 2. — Ch.\nges in Blood Concentr.\tion Before .vxn .\fter Intranasal
Administration of Pituitary Extract

Metaholism. — -The possible eitect of pituitrin on the metabolic
rate was also investigated. The patient's basal metabolism was there-
fore determined on four occasions by the Tissot method, two deter-
minations which checked within 4 per cent, being made each time.
The normal standards of Aub and DuBois for calories per square
meter of body surface (height-weight formula) per hour were used,
and the results expressed in per cent, of normal. The determinations
were made after a fourteen hours' fast, on two occasions after the
administration of pituitary extract and on two other occasions without
any medication.

The results were as follows: Dec. l.\ 1921, — 4; December 31,
— 26 ; Jan. 9, 1922, — 20 ; January 13, — 23. The relatively high result
of the first determination was probably due to the patient's restlessness,
associated with thirst and a desire to void. The second and third
determinations were made immediately following pituitary extract

8. Reiss: -ALrcIi. f. exper. Path. u. Pharniakol. 51:18. 1903.

BLUMGART—PITfJTARV EXTRACT 513

sprays with the patient perfectly comfortable and quiet. The fourth
determination was done two days after pituitary extract was discon-
tinued, when his daily intake and outinit were 8,600 c.c. and 8,400 c.c,
respectively. Prior to the first three determinations the patient had
been deprived of water for approximately three hours, but on this
fourth occasion he was permitted small amounts of warmed water at
room temperature until one hour before the metabolic rate was deter-
mined. As a result, the patient was not restless and the metabolic
rate was determined under more satisfactory conditions.

The foregoing results do not indicate any marked pituitary extract
effect on the metabolic rate. Whether the lowered metabolism in this
case bears any relation to the sj-ndrome of diabetes insipidus, or whether
it is due entirely to the malnutrition of the patient.'' it is impossible
to say.

DISCUSSION

. Extract of the posterior lobe of the pituitary sprayed intranasally
checked both the polyuria and polydipsia as effectively as hypodermic
injections. .Vll administrations of dried or aqueous extracts by mouth
or rectum proved ineffectual, this being in accord with the results of
practically all previous observers.

Histamin, whether .sprayed, swallowed or injected subcutaneously,
failed to modify the thirst or polyuria. One c.c. of a 1 : 10,000 solution
injected subcutaneously did not cause any loxic symptoms. The prep-
aration used was shown by Dale's method to be physiologically active
in concentrations of 1 : 10,000,000.

Lumbar puncture, performed on two different occasions, did not
lower the water or urinary output as it did in the cases reported by
Herrick "' and Graham."

The effect of pituitary extract intranasally and subcutaneously was a
diminution of the water intake and urinary output with a correspond-
ing alleviation of thirst, a rise in the specific gravity of the urine and
a dilution of the blood. The dilution of the blood has also been noted
by Konschegg and Shuster '- and by Priestly.'^

Whether intranasal sprays will be as successful in other cases, is,
of course, impossible to state, for it is conceivable that the efficacy of
the method in this instance was due to the small amount of ])ituitary
extract that the patient required.

The exact mechanism underlying the nasopharyngeal absorption is
not clear. Certain facts are, however, of considerable interest in this

9. Benedict, Francis G., Miles, Victor R.. Rath, Pane, Smith, H., Monmouth:
Publication No. 280, Carnegie Institution of Washington, 1919, p. 694.

10. Herrick: Arch. Int. Med. 10:1 (}u\v) 1S12.

11. Graham: J. A. M. A. 69:1498 (Nov. 3) 1917.

12. Konschegg and Shuster; Deutsch. Med. Wchnschr. 51:1091. 1915.

13. Priestly, J. G. : J. Physiol. 55:305, 1921,

514 ARCHIVES OF IXTER.XAL MEDICIXE

connection. "That the lymphatics of the nasal mucosa are in almost
direct communication with the subarachnoid space has been clearly
demonstrated," " and clinically, in children, a surprisingly small patch
of inflammation in the nasopharynx excites convulsions, stupor and
other phenomena indicative of considerable cerebral irritation. ^^
Flexner ^^ has shown that after intraspinal inoculation of monkeys with
the Diplococcus Intracellular is, the organisms can be detected both free
and intracellularly in the nasopharynx ; and similarly, the virus of
poliomyelitis has been demonstrated in the nasal mucosa of monkeys
inoculated intraspinally.'' The successful inoculation of poliomyelitis
virus into the nasopharynx of monkeys, clearly demonstrating that the
nasopharynx may serve as a portal of entry, is also suggestive in this
connection.^*

The preceding evidence indicates that the nasopharynx constitutes an
important factor in certain diseased states, but whether absorption in
the present instance is accomplished by the blood stream, by the lym-
phatics, or by both channels, it is impossible to state.

CONCLUSIONS

1. In a case of diabetes insipidus under observation, extract of the
posterior lobe of the pituitary applied intranasally checked both the
polyuria and polydipsia as effectively as hypodermic injection.

2. Histamin, subcutaneously ; lumbar puncture, and pituitary extract
by mouth, by rectum, and by phenyl salicylate coated tablets, proved
meffectual.

Note. — After this paper had been written, three additional cases of diabetes
insipidus were studied. In each instance intranasal application of pituitary
extract was found to be fully as satisfactory as hypodermic injection in reduc-
ing the fluid intake and urinary output to an approximately normal level.

14. Peabody, Draper and Dochez : A Clinical Study of .Acute Poliomyelitis,
Monograph of the Rockefeller Institute for Medical Research, \o. 4, June 1,
1912, p. 12.

15. Schloss, O. : Personal communication.

16. Fle-xner, S.: J. A. M. A. 55:1105 (Sept. 24) 1910.

17. Flexner, S., and Lewis, P. A.: J. A. M. A. 54:535 (Feb. 12) 1910.

18. Landsteiner, K., and Levaditi, C. : Ann. dc Tlnst, Pasteur 24:833, 1910.

THE VITAL CAPACITY IN A GROUP OF
COLLEGE STUDENTS *

A. W. HEWLETT, M.D., and N. R. JACKSON, M.D.

SAN FRANCISCO

The lessened vital capacity in intrathoracic diseases and the recom-
mendation that vital capacity be used as a test of physical fitness have
renewed interest in the question of normal standards for vital capacity.
The vital capacity varies greatly even among healthy individuals, and
some of the factors which accompany these variations are known.
Among them are the sex, weight, height, size of the chest, age and
general physical fitness. The clinician desires a normal standard with
which he may compare the vital capacity of his patient. Would he
do better, for example, to compare it with the average for individuals
of the same height, of the same weight or of the same chest measure-
ments; or should he use some combination of these? Obviously, that
standard is best which shows the least variation among normal
individuals. The convenience of the standard also deserves some
consideration, for a convenient standard is more likely to be generally
used.

Hutchinson ' after examining about 3,000 men came to the conclu-
sion that the most reliable standard for estimating the vital capacity
of men was their standing height. He stated that on the average the
vital capacity increased 8 cubic inches of air for every inch increase of
height between the heights of 5 and 6 feet. .According to Hutchinson
the vital capacity increased also with the weight, but this occurred
only up to an average weight of about 155 pounds. Increases of
weight beyond this were not, on the average, accompanied by increased
vital capacity. According to Hutchinson also the vital capacity tended
to grow less after the age of about 33 years, although the chest
circumference showed a slight tendency to grow greater. Peabody and
Wentworth ^ grouped their normals into three classes according to
height. This method has an obvious disadvantage in the case of those
whose heights lie near the class borders, for they are compared with
a standard that is best suited for a dififerent height. Lundsgaard and
Van Slyke ^ compared the vital capacity with certain chest dimensions ;

• From the Department of Medicine. Leland Stanford Junior University.

1. Hutchinson. J.: On the Capacity of the Lungs, and on the Respiratory
Functions, with a View of Establishing a Precise and Easy Method of Detect-
ing Disease by the Spirometer, Med. Chir. Tr. Lond. 29: U9, 1846.

2. Peabody. F. W., and Wentworth, J. A.: Clinical Studies on Respiration,
Arch. Int. Med. 20:443 (Oct.) 1917.

3. Lundsgaard. C, and Van Slyke. D. R. : Relation Between Thorax Size
and Lung Volume, J. Exper. M. 27:65. 1918.

516 ARCHIVES OF IXTERXAL MEDICIXE

but West * found a poor correlation between these dimensions and the
vital capacity of his subjects. Dreyer = compared the vital capacity
with the weight, the stem height (measured from the top of the head
to the end of the sacrum) and the circumference of the chest. So
far as weight was concerned. Dreyer neglected this in individuals
whose weight did not correspond to their stem height; so that for
practical purposes his comparisons were based on the stem height
and the circumference of the chest. Finally West ■■ compared the vital
capacity with the surface area, as calculated from the height and
weight by the DuBois' formula.*"'

Our observations were made on 400 normal young men at Leland
Stanford Jr. University between the ages of 18 and 30. AH of these
were active and showed no evident signs of disease. In each case the
height was taken in bare feet and the weight was stripped weight. The
vital capacity was determined by a spirometer which gave accurate
readings. Each individual was first shown how the test was performed
and was then given three trials. The highest of the three readings
was recorded as his vital capacity. The body surface area was estimated
from the height and weight by using the diagram of DuBois and
DuBois. Unfortunately, neither the sitting height nor the stem
height according to Dreyer were taken. The individual observations
are shown in Table 1.

A general conception of the relation between the vital capacities
and weights of our students may be gained from Figure 1. The obser-
vations were grouped according to weights, and the average vital
capacity for each weight group was determined. These averages
increased with increasing weights but the rate of increase was relatively
rapid at low weights and relatively slow at high weights. Hutchinson
noted a similar change at the higher weights. In his statistics, however,
the change at higher weights was more marked. We are inclined to
attribute this diflference to the fact that Hutchinson's observations
included men of all ages. The excess fat so often accumulated as a
person grows older is probably accompanied by no corresponding
increase in vital capacity. Dreyer stated that the vital capacity is
proportional to the 0.72 power of the weight (W + o'^). Dreyer's
line representing this relationship is curved in the general direction
of our averages ; but the curve for the weights here under consideration
is hardly apprecialile. For the sake of comparison we have inserted

4. West, H. F. : Clinical Studies on Respiration; Comparison of Various
Standards for Normal \'ital Capacity of Lungs. Arch. Int. Med. 25:306 (March)
1920.

5. Dreyer, G. : The Assessment of Physical Fitness, New York, Paul B.
Hoeher, 1921.

6. DuBois. D.. and DuBois. E. F.: Clinical Caloiimetry. Fifth Paper. Th.-
Measurement of the Surface Area of Man. .\rch. Int. Med. 15:868 (June) 1913.

HEW LETT-J AC KSOX— VITAL CAPACITY
TABLE 1. — Vital Capacity of Group of College Students

Vital

Height

Weight

Surface

Capacity

Area

6,700

187.0

80.4

2.00

6,555

186.7

75.7

1.98

6,555

193.7

87.7

2.17

6,655

189.2

82.3

2.07

6.500

195.5

85.6

2.18

6,200

180.5

79.0

1.99

5,981

185.4

77.3

1.98

5.961

184.0

94.1

2.14

5.899

172.0

70.5

5,860

187.2

65.0

i:87

5.889

185.4

75.0

1.98

5,817

172.7

70.9

1.82

5,800

179.5

71.6

1.90

5,760

185.0

TT.O

-11

5,736

180.0

60.5

5.736

180.0

76.4

1:95

5,738

180.3

90.0

2.10

5,736

184.0

78.2

2.00

5,736

181.2

68.2

1.95

5,752

191.0

79.5

2.C8

5,736

180.3

72.7

1.90

5,736

187.9

65.0

1.88

5,700

174.5

72.4

1.87

5,-00

186.0

70.2

1.93

■5,637

184.0

76.8

1.96

5.654

185.0

68.7

1.92

5,572

185.4

79.5

2.03

5,572

185.4

82.3

2.05

5,572

188.1

78.2

2.06

5,572

191.0

94.0

2.24

5,672

190,5

92.2

2.20

5,572

185.3

78.2

2.03

5:572

182.8

75.0

1.96

5,571

173.0

71.4

1.83

5,505

184.0

63.6

1.85

5.500

179.0

64.6

1.82

5.450

177.7

.70.6

1.88

5.490

179.1

63.6

1.83

5.490

185.4

84.0

2.07

5.407

170.0

77.3

1.87

5.407

186.5

85.0

2.10

5.407

187.2

72.7

1.95

5,407

188.0

71.4

1.96

5,410

178.1

79.5

1.98

5,420

189.0

83.2

5,415

180.3

77.5

1:95

5.488

177.8

62.3

1.76

5.407

177.0

60.0

1.75

5.405

177.5

58.2

1.76

5.325

185.3

84.1

2,08 ■

5,325

181.8

84.1

2.0.)

5,320

178.0

60.5

1.76

5,325

185.3

70.0

1.93

5,325

175.2

65.9

1.79

5,338

181.0

70.5

1.89

5,300

180.0

64.1

1.83

5,340

170.0

70.3

1.81

5,325

180.3

68.0

1.87

5.300

181.0

80.8

2.01

5.300

168.0

63.3

1.72

sisoo

181.5

70.4

1.90

181.8

74.4

1.95

5!244

18i.4

72.7

1.94

5.244

175.9

64.5

1.79

5.244

179.0

63.0

1.83

6.244

173.9

68.6

1.84

6.244

173.0

64.0

1.76

6211

184.0

70.0

1.91

5;235

174.0

94.0

2.08

6,244

175.5

80.0

i.9r.

5.244

177.0

69.1

1.86

5,260

178.1

71.8

1.88

5.244

180.6

80.0

2.00

6,244

185.6

, 72.7

1.96

6.244

185.0

66.9

1.91

5,244

188.0

77.7

2.06

5244

174.0

77.7

1.94

6.244

■ 167.6

58.2

1.68

5,277

184.0

70.5

1.92

176.0

72.3

1.86

5:244

180.8

76.3

1.9.-,

Vital

Height

■Weight

Surface

Capacity

Area

5.277

181.1

93.2

5.14

5.244

173.3

63.6

1.77

5,200

180.0

67.2

1.66

5,200

175.5

61.5

1.76

5,200

175.3

65.0

1.78

5:200

174.5

64.4

1.78

5,161

187.0

77.3

2.01

5,114

178.0

71.4

1.88

5,163

182.8

78.2

2.01

5:i63

171.6

69.5

l.Sl

5,163

171.0

67.3

1.76

5,150

170.5

64.4

1.75

5.100

174.0

61.6

1.75

5,081

175.9

. 64.5

1.80

5,061

182.9

72.7

1.92

5,081

176.0

66.4

1.78

5,081

184.0

86.4

2.08

5:0T4

■174.0

72.7

1.68

5,065

172.0

65.0

1.77

5.031

183.0

70.0

1.90

5,081

70.5

1.90

5,073

172:0

78.1

1.92

5,081

174.0

79.5

1.94

5,081

187.0

69.1

1.94

5,081

180.6

82.3

2.01

5:081

176.0

78.6

1.95

5,061

183.0

84.1

2.06

5,077

180.0

70.9

1.89

5,031

179.0

81.0

1.98

5,000

179.5

65.2

6,000

175.0

78.5

1.94

3,000

173.0

64.3

1.78

5:000

183.0

75.2

1.97

5.000

170.8

64.1

1.75

4 916

17.5.5

70.5

1.85

167.6

65.0

1.74

4:916

180.3

67.3

1.84

4.916

172.0

67.0

1.77

4.916

184.0

64.5

1.85

4,916

173.0

63.6

1.77

4,916

170.0

71.8

1.82

4,916

180.0

65.9

1.84

4,916

172.5

70.0

1.83

4,949

185.6

66.8

1.87

4,916

172.5

69.0

1.S3

4,916

185.6

1.97

4,962

184.0

72:7

1.96

4,952

190.0

78.1

2.06

4,900

185.0

73.2

1.94

4,916

173.2

79.5

1.93

4,916

183.6

71.4

1.92

4,916

182.8

84.1

2.06

4,998

177.8

63.0

1.81

4.916

181.5

71.0

1.89

4,900

177.8

66.7

1.83

4,916

178.3

70.0

1.87

177.8

60.5

1.75

4:998

180.3

61.0

1.77

4,916

179.0

74.0

1.93

4,998

179.0

67.8

1.87

4,916

174.8

79.5

1.95

4,916

168.5

66.9

1.76

4,960

168.8

65.0

1.73

4,900

169.0

66.0

1.77

4,900

169.0

72.0

1.83

4,900

172.5

68.2

1.61

4,850

179.0

1.80

4,850

175.8

67:8

1.83

4,850

175.0

60.5

1.74

4.800

171.0

65.0

1.77

4,800

178.0

65.2

4.834

175.4

68.2

i:84

4,850

186.5

82.7

2.07

4,834

184.0

92.0

2.14

4,834

176.2

70.0

1.85

4.741

172.7

67.3

1.78

4,734

176.9

75.9

1.91

4,752

172.0

03.6

1.76

4,752

184.0

69.1

1.91

4,752

173.0

71,8

1.84

4,752

180.3

70.5

1.89

518 ARCHIVES OF IKTERNAL ^^EDIC1^^E

TABLE 1.— Vital Capacity of Group of College Students— (Co«(i««^(J)

4.752
4,734
4,752
4,788
4.700
4,752
4,752
4,734
4.752

4.752
4.746
4,752

4,670
4.600
4.670

4.580
4.588
4.572

4.506
4.597
4.500
4.506
4.500
4,506

4,539
4,506
4.606

17S.0
181.0
183.0
186.5
175.7
175.6

172.7
180.3
180.3
180.0
170.0

171.0
182.0
180.5
183.1
173.8
189.0
183.0

173.8
180.3
181.5
180.3
178.0
180,0
163.0

187.5
170.0
164.3

60.5
67.5
114.5

4,425

187.2

79.5

2.05

4.425

175.8

68.2

1.85

4.425

179.0

62.8

1.82

4.42S

172.0

60.5

1.71

4.420

172.7

67.3

1.77

4.416

170.0

71.8

1.82

4:425

184.0

72.7

1.94

4.425

174.0

78.0

1.94

4.425

170.0

79.5

1.92

4.425

180.3

90.9

2.10

4,425

168.2

55.0

1.63

4,416

175.0

84.5

2.00

4,425

177.4

68.7

1.86

4.457

170.0

66.9

4,440

178.0

78.2

1.97

4,409

166.5

56.0

4:463

169.9

58.6

i:69

4.425

169.0

69.1

1.69

4,440

172.5

72.3

1.83

4,450

173.8

57.6

1.69

4,400

161.5

59.5

1.63

4,400

173.5

64.8

1.78

4,400

177.8

71.6

1.89

4,400

183.0

70.8

1.92.

4,343

176.0

65.5

1.80

4,343

170.1

60.5

1.70

4,343

162.0

64.5

1.69

4,343

175.7

74.5

1.90

4,343

168.0

66.4

1.76

4,350

163.3

60.0

1.65

4,300

169.0

67.4

1.78

4,300

174.0

57.2

1.69

4,300

179.0

60.2

1.77

4,300

167.5

58.8

i.er

4,300

166.0

64.8

1.72

4,251

168.0

67.5

1.78

4,250

170.0

60.0

1.70

4,250

169.0

60.5

1.69

4,250

162.0

63.2

1.64

4,251

180.0

67.8

1.87

4,251

170.1

63.6

1.75

4,251

176.4

87.7

2.06

4.211

176.0

69.1

1.85

4,250

172.6

88.8

2.08

4.250

165.0

65.0

1.73

4.211

173.0

75.0

1.88

4:251

177.0

60.0

1.75

4.251

167.6

56.0

1.63

4.250

178.0

68.7

1.87

4,242

170.0

61.8

1.72

4,245

174.0

60.5

1.73

4,250

173.6

64.5

1.77

4:250

166.5

57.7

1.66

4,250

178.0

68.2

1.87

4,250

164.0

64.5 ■

1.72

4.250

172.5

62.2

1.74

4,210

180.2

55.5

1,71

4,200

175.6

65.8

1.81

4.200

167.5

51.4

1.57

.4,200

172.0

57.3

1.68

4,200

165.0

56.3

1.62

4,200

in.o

57.4

1.67

4:100

186.0

60.0

1.82

4,179

160.0

64.5

1.68

4,170

172.6

81.2

1.94

4170

177.8

67.3

1.85

4,180

173.5

62.7

1.77

4.180

181.5

70.5

1.89

4180

170.0

64.6

1.75

4:100

171.5

57.0

1.67

4,100

174.0

64.0

1.78

4,087

172.7

64.4

1.65

4,087

179.1

60.4

1.77

4:080

168.9

.54.5

1.63

4,067

180.3

61.8

1.78

4,087 •

172.0

60.5

1.72

4060

176.0

67.3.

1.83

4:087

176.0

62.7

4,087

183.0

75.6

1:97

4,070

162.0

60.0

1.84

4:087

172.6

67.3

1.68

HEWLETT-.! ACKSOX— VITAL CAPACITY

TABLE 1.— Vital Cap.

ACITY OF

Group of College STUDENTS^(CoHftnMfd)

Vital

Height

Weight

Surface

Vital

Height

Weight

Surface

Capacity

Area

Capacity

Area

4.0S7

171.3

53.2

1.60

3,750

160.0

70.5

1.73

4.095

176.4

62.3

1.75

3,769

162.5

57.5

1.59

1,015

173.4

71.4

1.84

3,730

84.5

1.90

4,060

170.1

80.4

1.92

3,769

16510

57.7

1.66

4,015

170.0

55.4

1.64

3,750

175.5

1.72

4.015

165.0

54.6

1.60

3,750

168.5

56!5

1.64

4,087

. 168.2

63.6

1.73

3.605

172.6

1.70

4,087

164.0

64.5

1.71

3,605

157.4

47!3

1.46

4,050

166.3

59.8

1.67

3.687

172.6

68.2

1.82

4,(K0

172.0

57.3

1.68

3,687

60.0

1.64

4,050

172.2

60.9

3,605

17718

57.0

1.74

4050

172.5

55.9

1.67

3,687

185.0

1.79

4.000

169.5

60.0

1.69

3,687

173.0

59!l

1.72

4.000

173.6

58.7

1.70

3,605

172.0

60.0

1.72

3.960

172.7

60.5

1.72

3,688

167.6

65.9

1.V5

3.982

168.3

60.4

1.68

3,675

171.0

66.9

1.77

3,960

176.0

63.6

1.80

3,625

170.0

55.9

1.64

3,933

180.0

77.3

1.96

3.687

179.0

62.2

1.77

3,933

173.0

58.2

1.70

3,687

175.2

3,900

180.3

74.5

1.94

3,605

171.0

m.9

i!ts

3,933

158.7

60.0

1.63

3,650

160.7

57.3

1.60

3,960

169.0

62.7

1.73

3,600

157.5

K.O

1.63

3.933

168.0

75.9

.1.86

3,500

162.8

56.0

3,933

170.5

59.1

1.70

3,400

166.9

3,950

169.0

58.8

1.68

3,441

180.3

68^2

i:78

3.910

174.5

72.0

1.87

3.441

166.0

65.4

1.74

3,900

160.6

53.4

1.55

3,450

167.0

65.0

?,900

167.0

62.5

1.70

3.425

187.0

59.1

3.851

182.8

67.8

1.90

3.441

ITO.l

60.0

IJO

3.842

172.6

65.3

1.77

3,472

170.1

75.9

1.82

3.851

167.6

56.8

1.64

3,359

169.5

58.2

1.69

3,801

176.6

64.5

1.79

3,277

176.0

61.0

1.75

3,802

179.0

63.2

1.81

3,290

168.0

52.3

1.57

3.769

172.0

67.5

1.78

3,277

160.8

63.6

1.65

3,769

179.0

60.4

1.76

3,277

164.3

52.7

1.55

3,769

168.9

70.7

1.81

3,150

167.0

66.9

1.77

3,789

170.0

58.2

1.68

3,115

165.0

63.6

1.69

3,769

179.3

58.6

1.76

3,048

175.5

60.0

1.74

in Figure 1 the straight line relationship between weight and vital
capacity as calculated from our data, Dreyer's line, and Hutchinson's
averages of vital capacity for different weight groups. The lower
averages in Hutchinson's statistics will be considered farther on.

Figure 2 indicates for our students the average vital capacities of
the different height groups. We have added the similar averages of
Hutchinson's cases, the straight line relationship between vital capacity
and height as calculated from our data and the straight line relationship
which Hutchinson proposed for his. Both in our cases and in those
of Hutchin-son the relationship between the height and the vital capacity
for height groups approached a straight line.

The statistical methods employed are based on two assumptions.
The first is that the relationship between the factors compared
approaches a straight line relationship. We have shown that this is
approximately the case so far as the relation between height and vital
capacity is concerned. The relation between weight and vital capacity
deviates somewhat from a straight line in the case of our students
and this deviation is quite marked in Hutchinson's observations. The
second assumption is that the data for each group of observations
approximates the so-called normal distribution curve. This is true of

520 ARCHIVES OF IXTERXAL MEDICIXE

r

"~

...

■

-1

,

5500

-

r-

<f

1

^

'

,-'

5000

1 i !■

^

--'

*r

^

^

>-

•

"'

•<

,

[^

--^

O

.

^

'

>

+500

^

^

^

•

)^

---

■

^

■-

,

^

Jd

,

J

)

\^n

o

|>

iSUU

Fig. 1. — The vital capacity and weight. The dots represent the average
vital capacities for different weight groups of Stanford students. The circles
represent similar averages for Hutchinson's cases. The upper straight line is
the calculated line for Stanford students. The line just below is Dreyer's line.
Its curve is hardly perceptible.

i

n

li

^\

\

a

c

n

I.

»

»

S

S

,

55M

•

^

-3

J

r-

-

■^

5 mo

,

-

^

J

.—

^

.

_^,^^

.-'

,

--

^

»

-■

^

•

-^

J

C-

d

^

^

^

d

—^

'

300C

'

■

Fig. 2. — The vital capacity and height. The dots represent the average
vital capacities for different height groups of Stanford students. The circles
represent similar averages for Hutchinson's cases. The upper straight line is
the calculated line for Stanford students. The lower straight line is that pro-
posed by Hutchinson.

HEIVLETT-JACKSOX— VITAL CAPACITY

521

biologic measurements in general, including height and weight. It is
also true of the vital capacity. Figure 3 shows the actual observations
in Stanford students and the calculated theoretical distribution curve.

^^'ith the above assumption it is possible by the mathematical
methods employed in statistical studies to determine the relation of
vital capacity to height, to weight and to their combination. For deter-
mining these relations in college students we have used our figures on
400 Stanford students, the figures of West on eighty-five Harvard
medical students and the figures of Schuster ' on 959 Oxford students.
For comparison. Hutchinson's data on 1.285 men as shown in his
Table D have also been used. We have calculated from these data
the standard deviations, the coefficients of correlation, and the formulas
which best express the straight line relationship between vital capacity

1 i 1 1 1 1 1 1 J

1 1 1 1 ' 1 i

X5 1 " "~7

: it -L

■^^ 1 1 / ^/ /

Ml 1 ! 1 / / / / /

S Hi ' ~r It

00 1 i ' ^ ' ^

V 1

.

'^177771'^

7 \- -

ns : : 'yy^yy^rf

' L LlllL ' I

-^(r

1177 777777

So : i„-.i^-.Lyty-^yyyy

iJIllr ^

777711 K

Ati^y M. " it

15 - zw_yy7jty7/7ty/

^^J-Z-T-i ^/-IJ-lT-ll

lllAlli A^

J^^ttJ^ I

118 8

^ cO ,1. ^

ii.il

Fig. 3. — The frequency curve for the vital capacities of Stanford students.
The shaded columns represent the number of observations between different
limits of vital capacity. The curve represents the theoretical distribution as
calculated from the standard deviation, the mean, and the total number of
observations.

and various other factors. The methods used are described in Yule's
"Introduction to the Theory of Statistics."

Briefly, it may be stated that the standard deviation expresses the dis-
persion of the data on either side of the average. It is expressed by the

here " is the standard deviation, x is the deviation of

any one point from the average, Sx' is the sum of the squares of these indi-
vidual deviations, and N is the number of observations.

7. Schuster. F,. : First Results from the Oxford Anthropometric La1)oratory,
Jiomctrika 8:40. 1911.

522

ARCHIVES OF IXTERXAL MEDICI XE

The coefficient of correlation, r. indicates how closely the observations of
any two sets of data on the same individuals, e. g. vital capacity and height,
range themselves about a straight line relationship. It is calculated from

the formula r = ~^^ where x and y represent the deviations of each point

Nffxffy
from the averages of x and y. respectively, 2xy the sum of the products of
these deviations, N the number of observations and <rx and "y the standard
deviations of the two groups of observations that are being compared, e. g.,
vital capacity and height. The more nearly r approaches unity the more
closely do the data range themselves on a straight line. i. e.. the more exact
is the linear correlation. A mathematical expression for the line of relation-
ship is given by the formula y = a-)-r^x where for example, y represents
the vital capacity, x the height, r the coefficient of correlation between height
and vital capacity, "^y the standard deviation of the vital capacity, ffx the
standard deviation of the height and a is a constant. The standard deviation
of the individual observations from this calculated line is represented by the
expression ffy\ 1 — r".

TABLE 2.— Results of Comp.\risox of Data from Three Groups of Students

Oslord

Harvard

Stanford

S. &H.

S. H. & 0.

Hutchinson

Number

959
68.52

85

64.50
173.6
4.651
1.776

653.5
0.1167

0.61
0.67
0.63
0.73

400

68.49
175.9

if

0.1332

0.50
0.49
0.55
0.37

485

1,444

itf
4.426

7.885
6.805
646.9

1,285

mXT:

li^AiT.',^:'::::::

''^l^

s,eo2

St^n^rd Deviations:

7.428
6.608
61S.2

8.796

Hrilht.:::::::::::::

6.637

Vital capacity

655.3
0.1S24

(446.3)*

Coefficients of Corre-

0.66
0.59
0.57

0.60
O.aS
0.53

0.60

Wt. : V.C

Ht. : V.C

S. .4. : V. C

<0.86r

0.59

Table 2 gives the results obtained by these statistical calculations.
In the group studied by Hutchinson, which comprised men of different
ages and different occupations, the average weight was the same as
the average weight of the college students, whereas the average height
was definitely less. On the average, then. Hutchinson's subjects were
somewhat shorter and relatively fatter than our students. The average
vital capacity differed considerably in tlie different groups. The Stanford
and Harvard medical students showed the highest average vital capac-
ity ; the Oxford average was about 7 per cent, lower and Hutchinson's
average group was more than 20 per cent, lower. This low average
in Hutchinson's group was probably due. in part, to the fact that he
studied men of all classes and all ages, whereas college students
represent a picked class both as to age and general physical fitness. It
seems unlikely, however, that the low average in Hutchinson's group
was due entirely to this cause. Difference in technic or in the instru-
ments used may have been partly responsible for his low figures.

HElfLETT-JACKSOX—riTAL CAPACITY 523

Fluctuations on either side of the average are indicated by the
standard deviations given in Table 2. These fluctuations were greater
among the Stanford students than among the Oxford or the Harvard
medical students. In all groups the fluctuations of vital capacity were
relatively greater than the fluctuations of either weight or height. Thus
for all students the standard deviation for vital capacity amounted to
14.6 per cent, of the mean, the standard deviation for weight amounted
to 11.5 per cent, of the mean, and the standard deviation for height
amounted to 3.9 per cent, of the mean. It is evident, therefore, both
from the different averages and from the standard deviations that the
vital capacity of normal individuals shows a rather wide range of
fluctuation.

Table 2 shows also that among college students the correlation
between weight and vital capacity is approximately the same as the
correlation between height and vital capacity. Hutchinson did not
record the individual vital capacities of his subjects and it is not possible
from his data properly to calculate correlation coefficients for vital
capacity and either weight or height. We have, however, made such
calculations from the average vital capacities given in his Table D.
From these it is evident, as he stated and as is shown in our Figures
1 and 2, that for his cases the correlation of vital capacity with height
is far better than the correlation with weight. This difference between
college men and men at large with respect to the correlation of weight
and vital capacity may be explained on the assumption that the latter
g^roup includes some fat men whose excess weight is accompanied by
no increase or perhaps by a decrease of vital capacity. It seems to us,
tKerefore, that weight is not a reliable index of vital capacity unless
one can exclude those with excess fat.

Among college students, where excess fat is not common, there
is a closer correlation between vital capacity and a combination of
weight and height than between vital capacity and either height or
weight alone. West propo.sed that the body surface as calculated from
the weight and. height be used as an index of vital capacity and it may
be seen from Table 2 that the correlation with body surface is somewhat
better than the correlation with either height or the weight alone. A
combination of the linear relationships of vital capacity to weight and
to height was calculated. This correlation will be discussed later.

By applying the methods of statistical study to the above data,
formulas may be obtained by means of which one may calculate for
college students the probable vital capacity either from the height, from
the surface area, or from the linear combination of height and weight.
The first and last formulas are based on the combined Stanford, Oxford
and Harvard medical statistics. The second is based on the correlation

524 ARCHIVES OF IXTERXAL MEDICINE

coefficient between vital capacity and surface area of the Stanford and
the Harvard medical groups and on the average for all students. The
formulas obtained are as follows:
VC=50 Ht— 4.400
VC==2,900 S A— 1,000
VC=27 \Vt+31.5 Ht— 3,000

In these formulas vital capacity is expressed in cubic centimeters,
height in centimeters, surface area in square meters and weight in
kilograms.

The standard deviations of the observed vital capacities of students
from vital capacities which have been calculated from the above
formulas are as follows :

Standard deviation of vital capacities from height formula,
548.6 c.c.

Standard deviation of vital capacities from surface area formula,
529.1 c.c.

Standard deviations of vital capacities from weight and height
formula, 521.1 c.c.

T.ABLE 3. — Showing Number of Students Out of Each Hundred Who
M.w Be E.xpected to H.\ve a Vit.\l C-\p.acity Which Falls Below the
Following Percentages of Their Calcul.\ted Vital Capacity, When
the Differe.nt Formulas Are Used

Falling Below

90%

Height formula 21.0

•Surface formula *-0.1

Height-Weight fonnula 19.8

It will be seen that in tlie case of college students the fluctuations
from a formula based on both height and weight are less than the
fluctuations from a formula based on height alone. Between the formula
based on the calculated body surface area, and the formula based on
linear relationships between vital capacity and height and weight, there
is no significant difference. This conclusion, of course, -applies only to
college students. The significance of these standard deviations can
better be appreciated if one compares how many out of each inindrcd
college students will have a vital ca])acity that falls below any assumed
percentage of the calculated vital capacity. Table 3 gives these figures.

Of every hundred college students appro.ximately twenty-one
will have a vital capacity less than 90 per cent, of that calculated by
the height formula, and approximately twenty will have a vital capacity
less than 90 per cent, of that calculated by the body surface or by the
height-weight formula. On the other hand, only about two will have
a vital capacity less than 75 i)er cent, of the calculated anmunt. We

HEWLETT-JACKSON— ]-ITAL CAPACITY 525

have compared the calculated fluctuations among Stanford students with
those actually observed and find a close agreement between the two.

DISCUSSION-

. College students constitute an excellent group for determining
normal vital capacity standards. They are young, intelligent and
healthy. In this test intelligence is a factor because the test demands
a maximum efifort and statistics might easily be impaired if some
subjects did not understand the instructions or did not make the
necessary effort.

By applying statistical methods of study to the figures obtained
from college students we have determined which gave the best index
of vital capacity — the height, the weight or a combination of these;
we have obtained formulas by which one may predict the average vital
capacity for students from their heights or a combination of these
with their weights ; and we have defined the probable fluctuations
from these calculated averages. Unfortunately, we have not the
figures necessary for comparing the vital capacity with the sitting or
with Dreyer's stem height, but we hope later to report on the value
of this standard as judged by college statistics.

Most standards for vital capacity are based on the assumption that
there is a simple ratio between the vital capacity and some body
measurement or some power of this measurement. Thus the ratio to
the height, to the surface area or to some power of the weight or stem
length is given. Statistical formulas usually introduce an additional
constant which is added to or subtracted from one side of the equation.
This distinguishes our formulas from those that have been proposed.

Our formulas are strictly applicable only to college students. How
far may they applied to a larger field? We have no figures which
answer this question but it is evident from the work of others that
they are quite inapplicable to females in general,* as well as boys.'*
Furthermore, our standard is a high one for men at large, partly because
college students represent a picked class and, partly, because
the vital capacity tends to lessen as persons grow older. Excess fat
is a disturbing factor in any formula which is based on weight. For
this reason we have given no weight formula and we suspect that
for men in general a formula based on height and weight combined may
be no better than one based on height alone.

Vital capacity varies considerably even in such a selected group
as college students and even when it is compared with the height
or with a combination of height and weight. The amount of these

8. Emerson. P. W., and Cue, H. : \'ital Capacity of the Lungs of Cliildren.
Am. J. Dis. Child. 22:202 (Aug.) 1921.

526 ARCH WES OF INTERNAL MEDICINE

variations is shown in Table 3. What constitutes an abnormal reduc-
tion of the vital capacity? The answer to this question depends on
whether we are comparing individuals or group averages with the
standard. The difference between different groups of. normal indi-
viduals is illustrated by the deviations of the Oxford average and
the Harvard medical average from our standard for college students.
Making allowances for height, the former is 2.5 per cent, below the
average and the latter 8.7 per cent, above the average. To be signifi-
cant, a group average should probably differ from the college standard
by not less than 10 per cent. With respect to individual observations,
our study shows that of every hundred college students about two
have a vital capacity that is less than 75 per cent, of the standard.
Assuming that an occasional student has some unrecognized disease,
we may conclude that for men in general a reduction below 70 per
cent, of the standard is almost always abnormal. Studies on heart
patients and on patients with pulmonary tuberculosis indicate that in
these diseases the vital capacity frequently falls below this figure.

CONCLUSIONS

1. Among college students the correlation of vital capacity with
height and the correlation of vital capacity with weight are approxi-
mately equal. The correlation of vital capacity with a combination of
weight and height is a little better than the correlation with either
separately.

2. Formulas are given which express the average vital capacity of
college students for different heights and combinations of height and
weight.

3. From these formulas there is a very considerable fluctuation
even among college students. The fluctuations for men in general,
and the deviations in disease nnist necessarily be still greater.

THE LENGTH OF LIFE OF TRANSFUSED ERYTHRO-
CYTES L\ PATIENTS WITH PRIMARY AND
SECONDARY ANEMIA *

JOSEPH T. WEARN. M.D.,
SYL\IA WARREN and 0LI\IA AMES

BOSTON

Introduction. — During recent years, and especially since the intro-
duction of the sodium citrate method, blood transfusion has become one
of the most common forms of treatment of the various types of
primary and secondary anemias. Thus far, however, little is known
as to the length of life of the transfused erythrocytes in patients with
primary and secondary anemia, and until the recent work of Ashby ^
no observations had been made on this subject. This investigator
studied the length of life of transfused red corpuscles in pernicious
anemia and found the average to be about three months. Previous
to this time various observers had claimed the length of life of the
red corpuscles to be from fourteen to fifty-two days.- Information
on this point seemed desirable in that it might demonstrate the practical
value of transfusion, serve as an aid in deciding what the proper
intervals between transfusions should be, and because of its relation to
the debated problem as to whether or not there is an increase of
hemolysins in the blood serum of individuals with primary and sec-
ondary anemias.

It is generally known that the transfusion of blood is a safe pro-
cedure when the donor's cells are compatible with the plasma of the
recipient. The fact that the donor's plasma may agglutinate and
hemolyze the cells of the recipient is negligible because the donor's
plasma is diluted so rapidly as it enters the recipient's blood stream
that agglutination and hemolysis are impossible. It follows then that
Group IV blood (Moss classification) may be transfused without ill
effects into persons whose bloods fall into Groups I, II and HI. In
a similar manner a recipient in Group I may be transfused with bloods
of Groups II, III and R".

* From the Medical Service of the Peter Bent Brigham Hospital.

* This paper is Xo. 20 of a series of studies on the physiology and
pathology of the blood from the Harvard Medical School and allied hospitals,
a part of the expense of which has been defrayed from a grant from the
Proctor Fund of the Harvard Medical School for the study of chronic diseases.

1. Ashby: J. Exper. M. 29:267, 1919.

2. Ashby: M. Clin N. America, November, 1919: J. Exper. M. 29:267, 1919.
Ward and Muller ; Von Ott ; Hunter, W. : Quoted by .Ashby. 7. Exper. M.
29:207, 1919.

528 ARCHIVES OF IXTERXAL MEDICIXE

Technic. — Ashby has devised an ingenious method of following the
life of the transfused red blood cells by means of group agglutina-
tion.=* Blood is taken from the recipient's finger and mixed with
citrated blood serum which will agglutinate the cells of the recipient
but not the cells which have been transfused. The unagglutinated
cells (i. e., those that have been transfused) may then be counted. Iht
technic of the method is as follows (assuming a Group II recipient
and a Group IV donor) : The blood of the recipient is taken in a
leukocyte counting pipet up to the 0.5 mark, and is diluted up to the
11 mark with the agglutinating fluid, which is made up of Group IX
serum and a 4.4 per cent, solution of sodium citrate in the proportion of
1 : 20. The pipet is shaken and the mixture expelled into a small test
tube in which it is incubated at 2>7 C. for forty minutes, with thorough
shakings every ten minutes. It is then left in the icebox over night.
Just before counting, the mixture is shaken and a drop placed in the
blood counting chamber and, as directed by Ashby, "160 small squares
in each of the two chambers are counted, tlie average of the counts
taken and multiplied by 110O/2 to give the number of unagglutinated
or transfused corpuscles per cubic millimeter of blood." Controls on
the activity of the agglutinating serum are made by using the technic
described alx)ve on blood of an untransfused individual in Group II.
Theoretically, all the Group II cells should be agglutinated by the Group
\\ serum ; practically, however, there remain unagglutinated on an
average of from 20,000 to 50,000 cells per c.mm. Serum which will
agglutinate the blood of a normal person in Group II, leaving only
50.000 cells per c.mm. unagglutinated may be considered to be active.
When the unagglutinated cell counts in transfused individuals remain
higher than these control counts, it may be assumed that the number
of unagglutinated corpuscles in excess of the control count represent the
number of unagglutinated donor's cells present in the circulation. Cases
have been studied by us according to Ashby 's method and technic. All
counts have been made in duplicate, and the figures in the accompanying
tables represent the average. In the cases reported, controls on the
serum were made as described above, and, except in two instances, the
same agglutinating serum was used throughout for all patients. Control
counts of the new and old serums were slinwn to l)e practically identical
before the changes were made.

Observations on Patients ivitli Pernicious Anemia. — Four patients
with pernicious anemia, whose bloods were in Group II, were trans-
fu,sed with citrated blood from donors in Group I\'. Counts were
then made of the transfused or unagglutinated cells in the bloods of
these recipients, and repeated many times until the total transfused
cells dropped to the level of the control counts. These patients pre-
sented the clear cut and classical signs and symptoms of jirimary

3. Ashby: J. Kxper. M. 34:147. 192

WEARN-IVARREX-AMES—LIFE OF ERYTHROCYTES 529

anemia, but among them were representatives of the different stages
of the disease. Patient A was in the sixth year of tlie disease, was
bedridden and had been observed in a remission during his stay in
the hospital the year before, but in a series of counts taken before
his transfusion his erj-throcytes were found to be steadily decreasing.
Patient C, in the third year of the disease, was also bedridden and
his red blood cell counts showed the same general curve as the counts
of Patient A. In contrast. Patients B and D were in the earlier stages
of the disease, had been able to do some work and were ambulatory
at the time of their transfusion.

The observations on Patient D are of especial interest in that the
donor for transfusion was another patient with pernicious anemia
(Group IV). When the donor was bled, 300 c.c. were withdrawn and
he was immediately transfused with 800 c.c. of normal blood. The
volume of cells in the 300 c.c. of pernicious anemia blood being very
small, the unagglutinated cell count in this recipient never rose above
97,000 per c.mm. Frequent counts were made in this case, and the
control counts which were made on untrans fused normals in Group II
and also on the blood of an untransfused patient with pernicious anemia
in Group II. agreed closely.

REPORT CF CASES
Case 1. — Patient A (jMedical No. 13211) was a man. aged 41, belonging to
Group II. His symptoms began four years before his entrance to the hospital.
They were weakness, pallor, distress after eating and numbness of the fingers
and toes. Two years previously, while under observation in the hospital, he
had a typical remission with marked improvement of all his symptoms except
the numbness of the extremities. Eight months later he again began to grow
steadily weaker. In the month previous to his present entrance to the hos-
pital his red cell count dropped from 1,332,000 to 676,000. He was then
transfused with 575 c,c. citrated blood of Group IV. Blood findings before
transfusion were: white cell count, 2.500 per c.mm.: hemoglobin, 30 per cent.:
platelets, 98,000 per c.mm.; reticulated cells. 0.5 per cent. The blood picture
was typical of pernicious anemia (Table 1, Fig. 1).

TABLE 1. — Blood Picture of P.\tient A

Nov 16 i»:o

1«

Dec 3 iq^O

1«

D« 1 1920

ra

Dec 8 1920

28

Dec 13 1920

.SO

Jan n 1921

an

Jan 20 1921

73

n

Jan 31 1921

7K

Feb 1 1921

79

Feb 2, 1921

Ki

. Feb 8 1921

»5

Feb 18 1921

101

Feb 24 I<K1

106

Mnrrli 1 1921

n«

March 10 1921

ri*

March 14 1921

12«

March 22 1921

Cnagglutlnated or

Rpd Blood Cells

Donor's Cells

1.152,000

Transfused 575 c.c. cit

ratid blood. Group IV

1,024,000

530.000

1,936,000

.'J3S,700

1.800,000

691,900

1,552,000

553,850

1.216,000

464,200

848.000

349.800

1,344.000

300,300

811,436

148,500

884,268

228.2.50

Transfused 600 c.c. cit

rated blood. Group II

1,160,000

229.900

1,160.000

1,088,000

904,000

.•501,400

1.328,000

3.il.450

1.744,000

258,500

1,312,000

122,650

1,094,808

i22'?f!>

530 ARCHIVES OF IXTERXAL MEDICIXE

Case 2. — Patient B (Medical No. 15159), a woman aged 46, also belonged
to Group II. Her illness began one year before admission to the hospital.
The symptoms were gradually increasing weakness and pallor, several attacks
of diarrhea and slight numbness of the fingers. She had been able to do a
little work around the house. Her blood report before transfusion was as
follows: Red cell count, 3,456.000 per c.mm. ; reticulated cells, 0.9 per cent.;
white cell count. 4,700 per c.mm.; platelets, 151,000 per c.mm. The blood smears
showed the typical findings of pernicious anemia. She was transfused with
400 c.c. of citrated blood from a Group IV donor (Table 2, Fig. 2).

TABLE 2. — Blood Findings of Patient B

No. of Unagglutinated or

Days Date Bed Blood Cells Donor's Oils

1 Jan. 9,1921 3,456,000

1 Jan. 9,1921 Transfused 400 c.c. citrated blood. Group IV

3 Jan. 11, 1921 4,152,000 4M.V.W

13 Jan.21,1921 4,648,000 381,160

19 Jan.27,1921 4,338,000 282,150

24 Feb. 1,1921 4,256,000 247.500

31 Feb. 8,1921 3,592,000 307,450

38 Feb. 15, 1921 5,446,000*

51 Feb.28,1921 5,338,000 259,600

55 March 4,1921 3,688,000 347,600

75 March24,1921 4,368,000 216,150

90 April 8,1921 4,214,000 110,000

98 Aprill6,1921 4,320,000 106,150

111 April 29, 1921 4,566,000 35,200

• The sudden rise i
by a sudden increase
total red cell count.

Case 3. — Patient C (Medical No. 14705) w^as a woman, aged 55, belonging
to Group II. Two years before entrance to the hospital, a gradually increas-
ing weakness, pallor and gastric distress marked the onset of her illness. Then
she developed numbness of the hands and feet. Her blood findings before
transfusion were: hemoglobin, 45 per cent.; red cell count, 2,356,000 per c.mm.;
reticulated cells. 1.5 per cent.; white cell count, 4,100 per c.mm. The stained
smear was typical .of pernicious anemia. She was transfused with 600 c.c.
citrated blood from a Group IV donor and later had several transfusions of
Group II blood (Table 3, Fig. 3).

T.^BLE 3. — Blood Findings of P.\tient C

No. of Unagglutinated or

Days Date Red Blood Cells Donor's Cells

1 Jan.18.1921 l.-268,000

1 Jan, 18, 1921 Transfused 250 cc. citrated blood, Group IV

3 Jan.21,1921 1,028,000 238,150

7 Jan. 25, 1921 Transfused 600 c.c. citrated blood. Group II

10 Jan.28,1921 2,356,000 165,000

14 Feb. 1.1921 2,016,000 161,;0O

16 Feb. 3,1921 Transfused 550 c.c. citrated blood. Group II

17 Feb. 4.1921 2,776,000 189,7.-0

43 Marchl9,1921 1,360,000 124.850

43 March 19, 1921 Transfused 600 c.c. citrated blood, Group IT

71 April 16, 1921 1,544,000 57,750

C.\SE 4. — Patient D (Medical No. 15642), a man aged 50, belonging to
Group II, noticed symptoms of weakness and pallor three months before admis-
sion to the hospital. One month later he had a sore tongue, distress after
eating and numbness of the fingers. He had been able to work until shortly
before he entered the hospital. His blood findings at entrance were as fol-
lows: hemoglobin, 18 per cent.; red cell count, 1.224.000 per c.min.; reticulated
cells, 0.9 per cent.; white cell count, 5,600 per c.mm., and platelets, 128,000
per c.mm. The sirear was characteristic of pernicious anemia. He was trans-
fused with .300 c.c. of (Iroup IV blood from another patient who had a typical
pernicious anemia (Table 4, Fig. 4).

Fig. 1. — Blood findings of Patient A. In this and the accompanying charts
the figures on the left border represent millions in the total red blood cell
counts, and hundreds of thousands in the unagglutinated or transfused red
blood cell counts.

Fig. 3. — Blood findings of Patient C.

Fig. 4. — Blood findings of Patient D.

ARCHIVES OF INTERNAL MEDICINE
TABLE 4. — Blood Findings of Patient D

Unagglutintited or
Date Red Blood Cells Donor's Cells

March 19. 1921 1.224.000

March 19.1921 Transfused 300 c.c. citrated blood, Group IV (P. A.'

March 22, 1921 1,401.000

March25,1921 1,200.000 86,350

March 28, 1921 1,448,000

March 29. 1921 1.520,000 89,6.50

March 30. 1921 1.928.0<K» 87,liJO

March 31, 1921 1,884,000 86,900

April 1,1921 1,701,000 94.600

April 2,1921 1.618.0OO 80.850

April 4,1921 1.680.000 78,650

April 5,1921 1.688,000 84.150

April 6,1921 1,568,000 94,050

April 7,1921 2,144.000 90,200

April 11, 1921 2,064.000 78,100

Aprill3,1921 2,120,000 80,850

Aprill9,1921 2,560,000 96,800

April24,1921 2,688,000 84,700

April 28, 1921 2,804.000

May 5.1921 3,832,000 82,500

May26,1921 2,600,00* 75,900

Jime9,1921 2,528,000 74,250

July 6, 1921 1,600.000 47,850

July 8,1921 1,872,000 54,833

Observations on Patients zvitli Secondary Anemia- — Four cases
(E, F, G and H) of secondary anemia, accompanying an advanced
nephritis, were studied in the same way. Before transfusion, the
patients showed progressive anemias as proved by decreasing red cell
counts. Patients F and H were bleeding constantly from the kidneys
and their urines showed gross blood. Patient E had a very small
amount of blood in the urine, averaging from five to six red blood
corpuscles per low power field when examined microscopically, while
Patient G showed no evidence of hematuria or bleeding elsewhere.

Case S.— Patient E (Medical No. 15126). a man. aged iZ, in Group II, had
had chronic nephritis for about six months before admission to the hospital.
On entry his systolic blood pressure was 190; diastolic. 110. Phenolsulphone-
phthalein excretion was 42 per cent., blood urea nitrogen. 20 mg. per hundred
c.c. of blood. His urine showed a few red blood cells constantly, averaging
about five or six per low power microscopic field. The blood findings were:
hemoglobin, 60 per cent.; red cell count, 4,060,000 per c.mni. ; reticulated
cells. 0.9 per cent.; white cell count. 9.650 per c.mm. The red corpuscles were
practically normal in size and shape but showed definite achromia. He was
transfused with 600 c.c. citrated Iilood from a Ooup IV donor (Table 5. Fig. 5).

TABLE 5.— Blood Findings of Patient E

No. ol
Days

1

Date

Jan. 1.5. 1921

Jan. 15. 1921

Red Blood Cells

1.060.000

Tran.sfuseiieooe.c. lit

Unagglutlnated or
Donor's Cells

rMt.d hlood. Group IV

52
54
70

March 1,1921

March 11, 1921

March 2.-!, 1921

March 29, 1921

4.992.000

4.793.00O

4.-560.000

4.656,000

244,7.50
3.32.750
239,800
162,2.50

April .5,1921

4,008,000

143,550

92

April 20, 1921

3,368,000

94.600

3.420 000

119

May 17,1921

3.248.000

34.650

UEARX-n-ARREX-AMES—UFE Of ERYTHROCYTES 533

Case 6.— Patient F (Medical Xo. 15371), a man. aged 23. in Group 11. had
had influenza followed by acute nephritis two years before his entrance to the
hospital. He has passed bloody urine constantly since that time and through-
out his stay in the hospital the urine showed gross blood. The phenolsul-
phonephthalein excretion varied from 18 to 35 per cent.; blood urea nitrogen
was from 26 to 61 mg. per hundred c.c. on a low protein diet. Blood findings
before transfusion were as follows: hemoglobin. 65 per cent.; red cell count,
4.512.000 per c.mm. ; reticulated cells, 2 per cent.; white cell count, 11,200 per
c.mm. The smear showed slight achromia of the red cells but was otherwise
essentially normal in appearance. He was transfused with 550 c.c. citrated
blood from a Group IV donor (Table 6; Fig. 6).

T.^BLE 6. — Blood Fixdixgs of Patient F

Date Red Blood Cells Donor's Cells

.March 2S. 1921 4,572,000

April 6. 1921 Transfused 560 c.c. citrated blood. Group IV

April 6. 1921 3,664,000 248,060

April 8,1921 3,9^4,000 275,560

Aprilll,1921 4,040,000 259,600

Aprill3,1921 4,160,000 296,450

Aprill9,1921 4,736,000 294,800

April26,1921 3,072,000 208,450

May 2,1921 3,216,000 202,950

May 5,1921 3,552,000 242,000

Mav 12, 1921 3,416,000 189,200

Mayl9,1921 3,208,000 143,005

May22.1921 4,288,000 176,550

June 4,1921 2.81.6,000 156,760

June27,1921 3,828,000 83,600

July 1,1921 3,600,000 68,750

July 13,1921 3,506,448 42,185

July 19,1921 4,016,000 33,000

Case 7. — Patient G (Medical No. 14976), a man. aged 33, in Group II. had
had nephritis for about eight months. Physical examination revealed ascites
and considerable edema of the genitals and lower extremities. The blood pres-
sure was : systolic. 128 : diastolic. 96. Phenolsulphonephthalein excretion was
13 per cent., and blood urea nitrogen was 25 mg. per hundred c.c. The urinary
sediment contained many hyalin. granular and waxy casts but no red blood
cells. Before transfusion the red count was 3,736.000 per c.mm.; reticulated
cells. 0.9 per cent.: hemoglobin. 45 per cent. The stained smear showed prac-
tically normal red cells with slight achromia. He was transfused with 250 c.c.
citrated blood (containing 10 000,000 red corpuscles per c.mm.) from a donor
in Group I\' (Table 7, Fig. 7).

TABLE 7.— Bwon Findings of Patient G

No. of

Days Date

1 March 28 1921

Xi May 2,1921

45 May 22 1921

57 May 24 1921

.'•9 May 26. 1921

67 June 3 1921

74 June 10 1921

79 .Tune 15 1921

a5 June 21, 1921

94 June 30 1921

106 July 11, 1921

112 July 18 1921

142 August 17, 1921

150 August 25 1921

1.V, AueuFt31 1921

Unagglutlnated or

Red Blood Cells

•i 764 000

2 896 00O

an«fiised 2.50 c.c. (

MtratPd blood. Group TV

■-.064,000

4 576 000

261,2.50

4fi88 0W)

2f0.500

5 024 000

265,650

163,350

4 344 000

179,30^

1712 COO

130,660

4 456,000

69.300

4,f00 000

49,560

Fig. S. — Blood findings of Patient E.

Fig. 6. — Blood findings of Patient F.

Fig. 7.— Blood findings of Patient G. .

Fig. 8. — Blood findings of Patient H.

U-EARX-jyARREX-AMES—LIFE OF ERYTHROCYTES 535

Case 8.— Patient H (Medical No. 15653). a boy. aged 16, in Group II had
nephritis, of two months' duration, which followed a respiratory infection. He
was not edematous and his blood pressure w as : systolic. 148 ; diastolic, 90. The
urinary sediment contained hyalin and finely granular brown casts and many
red blood cells so that the blood was visible grossly from time to time. Phenol-
sulphonephthalein excretion was 50 per cent., and the blood urea nitrogen was
12 mg. per hundred c.c. His red cell count before transfusion was 3,360,000
per c.mm. ; reticulated cells. 2.2 per cent.; hemoglobin, 52 per cent.; white cell
count. 6.900 per c.mm. The stained smear showed slight achromia of the red
corpuscles. He was transfused with 250 c.c. citrated blood (containing
10.000.000 red corpuscles per c.mm.) from a Group IV donor (Table 8. Fig. 8).

T.\BLE 8.— Blood Findinxs of P.^tient H

No. of Unagglutinated or

Days Date Bed Blood CeUs Donor's Oils

1 Aprin6.1921 4.144,000

17 May 2.19-21 3.672.000

37 May22.1921 3.360,000

39 May 24, 1921 Transfused 250 c.c. citrated blood. Group IV

41 May26.1921 4,497,000 451.5.tO

49 June 3,1921 4,560,000 281,150

56 June 10, 1921 4,344.000 306.350

61 JunelS, 1921 4.408.000 288,150

66 June 20, 1921 4,456,000 228,060

75 June29,1921 5,000,000 162,250

87 July 11, 1921 5.256,000 141,350

94 July 18,1921 4,904,000 123,760

103 July 27, 1921 5,144,000 88,000

111 August 4,1921 5,344,000 79,750

119 August 12, 1921 4,952,000 62.700

124 August 17, 1921 5,616,000 68,200

150 Auguse 25, 1921 4,592,000 58.850

156 August 31, 1921 5,208.000 78.100

164 Sept. 8,1921 4,688,000 48,950

DISCUSSION

Unfortunately it has not been possible in this investigation to study
the length of life of cells transfused into normal individuals, but, in
the light of the present findings, further studies on normals should be
made. Ashby observed only one normal case until the transfused
corpu.scles disappeared from the circulation and found that they lived
foi: thirty-nine days, while in her other two cases, which were followed
incompletely, the cells lived about thirty-two days. In a later study,
this author reports the length of life of transfused red blood cells in
eight patients "without blood disease" to be very variable, some living
as long as one hundred days, others disappearing in thirty days. The
protocols of these patients, however, show that some of them had cancer
or other malignant growths, one was in the tertiary stage of syphilis
and none can be regarded as normal. Indeed, one finds included in the
lis» patients with some of the most common and frequent causes of
secondary anemia, and while in the technical sense of the term they
may be "without blood disease" they show a decided disturbance of the
blood.

Ashby found the length of life of the transfused crj-throcytes in
pernicious anemia to be about three months and concluded that there
was no hemolytic toxin producing the anemia in this disease. The

536 ARCHIVES OF IXTERXAL MEDICIXE

results of this investigation show the length of life of the transfused
corpuscles in primary anemia, and in the one type of secondary anemia
studied, to be from seventy-one to one hundred and ten days, but it is
felt that no conclusion regarding the presence of a hemolysin is justi-
fiable, because these observations furnish no direct evidence on this
point, and also because of the lack of accurate information as to the
duration of life of red corpuscles transfused into normal persons.
Furthermore, no evidence could be found to support the claim, made
by Ashby, that "on the whole, blood destruction is quiescent in per-
nicious anemia." The results here show merely that the length of life
of transfused erythrocytes is greater in patients with primary anemia
and with secondary anemia due to nephritis, than in the one normal
case on record ; and this is only known to be true .when the donor and
recipient are of unlike groups. Whether or not the same would hold
true when the donor and the recipient are in the same group it is not
possible to say. In addition, these observations do not throw any
light on what is happening to the patient's own cells during this period,
and there is no reason to believe that the rate of destruction of the
transfused cells is any indication of the rate of destruction of the
patient's own cells, for it is possible that the transfused erythrocytes,
belonging to a group foreign to the patient, are not acted on in the
same manner as the patient's own cells.

It has not been proved conclusively that the transfused blood cells
function during their stay in the circulation, but the fact that many
of the patients show some clinical improvement after transfusion sug-
gets that this is the case. Moreover, the observations of several investi-
gators have shown that any foreign material, such as manganese or
carbon particles and foreign blood cells, when injected into the blood
stream of an animal are removed quickly.'' This being the case, if the
transfused cells were not living and functioning as normal red blood
corpuscles one would expect them to be removed from the blood
stream.

Several rises in the counts of transfused red cells were noted just
before their final disappearance (Figs. 1 and 5). As the control counts
on these days did not increase, and there was no variation in technic,
dififerential counts of the number of microcytes were made to see if a
breaking up or fragmentation of cells might account for these rises,
but no evidence to support this theory was found. In view of the
coincident rise of the total red blood cell count, it is not unlikely that
changes in the blood volume account for the rise. One other possible
explanation of this occurrence is that following transfusion there may
be a definite improvement in the circulation with a resulting rise in

4. Mcjiinkin: J. Exper. M. 21:59, 1918. Lund, Shaw and Drinker: J. Exper.
M. 33:231, 1921. Drinker and .Shaw: J. Exper. M. 33:77, 1921.

IVEARX-n-ARREX-AMES-LlFE OF ERYTHROCYTES 537

the number of circulating cells in the periphery. This seems plausible
in Case. 1.

The most striking result of these observations on the life of the
transfused cells in the primary and secondary anemias is their sur-
prisingly long stay in the circulation of the recipient. The longest
period before the disappearance of the last cells was 113 days and the
shortest fifty-nine days, the average for all these observed being about
eighty-three days. It must be remembered that in addition to this
period of life as transfused erythrocytes in a foreign circulation, that
some of these corpuscles were functioning as adult cells in the donor
before they were transferred, but of the length of time that they had
been in the adult stage nothing is known, nor is there any accurate
knowledge of the period of time required for a red blood cell to pass
from its immature nucleated stage to its adult nonnucleated stage.
These considerations, together with the findings of this investigation,
indicate that the life of the human red blood cell is rrfuch longer than
has been believed to be the case. Whether the duration of life and
the stages of development of these cells would have been the same
in ths circulation of the individual from whom they originally came it
is not possible to say.

It is also of interest that the life of the transfused red cells in
both secondary and primary anemias was of the same duration. It
seems almost probable that the transfused cells are adult red corpuscles
of varying ages, and this, if true, would account for their steady
gradual disappearance from the circulation of the recipient, also for
the fact that some of the cells begin to disappear almost immediately
after transfusion. This explanation is also compatible with the idea
that new red corpuscles are being constantly supplied to the circulation.
There were no sudden drops in the transfused cell count during these
observations, but this may be due to the time intervals between counts,
and in this connection it will be noted that two of the patients with
pernicious anemia were women, both of whom had ceased to men-
.struate, so that no loss of transfused cells can be accounted for by
that route, as noted by Ashby. In Case 6, in which there was con-
stantly a large amount of blood in the urine, an attempt was made to
determine the number of transfused corpuscles that were being lost in
this way, but this was unsuccessful as the total unagglutinated count
of the red corpuscles in the urine after centrifugalization was less than
the control counts of the agglutinating serum on normal Group II
blood.

.Another point that is clearly brought out in this study is that due
to the long life of the transfused red cells one may expect to tide
patients over the acute stages of primary and secondary anemias by
purely mechanical means. The improvement after some transfusions.

538 ARCHIVES OF IXTERXAL MEDICI XE

except when a real remission begins, might be explained by the increase
in oxygen carriers. This improvement, which is generally of about
two or three months' duration, is probably governed by the fact that
some of the transfused corpuscles seem to function between sixty and
ninety days.

CONCLUSIONS

Red blood corpuscles from donors in Group IV transfused into
patients in Group II with pernicious anemia and anemia secondary to
nephritis, remained in the circulation longer than has been generally
believed to be the case. The last of the transfused red blood cells
disappeared from the circulation in from fifty-nine to 113 days, with
an average of eighty-three days.

No difference was noted in a series of observations in the duration
of the stay of the transtused red blood cofpuscles in the circulation
between patients with primary anemia and secondary anemia (due to
nephritis).

In a single observation red blood corpuscles from a patient with
pernicious anemia transfused into another patient with pernicious
anemia, behaved as did the corpuscles from normal donors.

BLOOD PRESSURE AND PULSE RATE LE\'ELS

F[RST paper: the levels under bas.\l and daytime conditions*

T. ADDIS

SAN FRANCISCO

This study is primarily concemed with the question of normal blood
pressures and pulse rates under varying conditions. The observations
on patients cover only a restricted field and are introduced, in the main,
as illustrations of the deductions which may be drawn from comparison
with the normal data. The work falls naturally into two parts, one,
the measurement of the level of pressure and pulse rate under fixed
conditions, with which this paper is concerned ; the other, the measure-
ment and interpretation of the changes in pressure and pulse rate
induced by alteration of the conditions, which are dealt with in a
succeeding paper. Most of the obsen'ations were made on soldiers at
Camp Lewis, Wash. For the opportunity to do this work I am indebted
to Dr. Kerr, who was in charge of the medical division of the Base
Hospital, to Dr. Northington and Dr. Fulton who were successively in
command, and to Lieutenant-Colonel Gibner who was camp surgeon.

The level of pressure and pulse rate under what we have called
"basal" and "daytime" conditions is the subject of this paper. In work
on metabolism the word basal is used to indicate that the measure-
ments have been made in the early morning before food has been taken
and before the subject has done any muscular work. In this paper it
has the same significance. The observations were made in the early
morning before the subjects had. risen from bed. In most cases they
were first awakened by the application of the arm band. By daytime
observations are meant those taken at various times during the day
after the subjects had risen from bed and had taken food. In all cases
the readings were made while the men were lying down. Those who
had recently done any strenuous muscular work were excluded. Most
of the work was done on Sundays when bad weather had kept the men
relatively inactive in their barracks.

The incentive to collect data on the normal basal blood pressure and
pulse rate was derived from a difiiculty in diagnosis in a group of
patients who presented signs and symptoms resembling those seen in
hyperthyroidism. While recruits from the first and second drafts were
arriving at Camp Lewis, men were seen every day who combined an
enlargement of the thyroid gland with tachycardia, tremor and evident

■ From the Medical Department of Stanford University Medical School.

540 ARCHIVES OF IXTERXAL MEDICINE

signs of nervous instability. It was recognized from the first that only
a small percentage were likely to have exophthalmic goiter for true
exophthalmos was rare. The hyperthyroidism which arises in some
cases of endemic goiter is uncommon at the age period of the men we
were examining, and we were dealing with a condition which was not
at all uncommon. It was noted also that similar signs and symptoms
were found in men who had no thyroid enlargement, and on this
account a statistical study was made of the incidence of tachycardia,
tremor and various other abnormalities in large groups of men with
and without increase in the size of the thyroid gland. ^ This survey
showed that there was no definite relationship between the thyroid
enlargement and the occurrence of tachycardia, tremor and other evi-
dences of vascular and nervous instability. It appeared, then, that the
thyroid enlargement was only a chance concomitant which was fre-
quently present simply because endemic goiter was so extremely
prevalent in many of the districts from which the recruits were drawn,^
It thus seemed still more unlikely that hyperthyroidism could be a
frequent cause of the condition. \\'e felt confident that the great
majority of these men were suftering from the condition described
under the names irritable heart, neurocirculatory' asthenia and effort
syndrome. But in individual cases there was often uncertainty, and
this led to a search for some objective clinical evidence of an increase
in basal metabolism, to take the place of the direct measurements of
oxygen consumption which we could not at that time obtain.

A relationship has been shown to exist between the pulse rate and
the metabolism when they are measured under the same conditions, and
for some time we placed a great deal of weight on the pulse rate
counted in the early morning before the patients had risen from bed.
When the pulse rate was less than 70 in patients who during the day
had tachycardia and tremor, we felt that we could probably exclude
hyperthyroidism. However, the relation between pulse rate and rate
of metabolism is not always close. It seems likely that the reason for
the relation which does exist is to be found in a more general
relation between the metabolic activity of the body as a whole and the
activity of the circulatory system of which the pulse rate is only a
partial expression. The true measure of circulatory activity is the
volume flow of blood jier unit of time. It has been shown by combined
circulatory and metabolic measurements that there is a close corre-
sjiondence between volume flow of blood and rate of metabolism. The
\()lume flow (if blcKid ]».r unit of time has been determined in man by

1. .^cUlis and Kerr: .^rcll. Int. Med. 23:.116 (March) 1919.

2. Kerr: .^rcll. Int. Med. M:M7 (Sept.) 1919.

ADDIS— BLOOD PRESSURE 541

Lindhard ^ and by Means and Newburgh.* The output of the heart
was measured by gas analysis methods over a short space of time during
which the pulse rate was counted. The average output at each beat of
the heart was thus calculated, and the volume flow of blood per minute
obtained from the product of the systolic output per beat and the pulse
rate per minute. In subjects on whom measurements were made before
and after exercise a remarkably close agreement was found between
the increase in this product and the increase in the rate of metabolism
produced by e.xercise. Bainbridge ^ recently pointed out that a com-
parison of such experiments shows that the two factors in the product
which measures the volume flow of blood — the systolic output per beat
and the pulse rate — may each vary, although for given metabolic con-
ditions the product will remain constant. The required volume flow of
blood may at one time be obtained mainly by increase in systolic output
and at another time mainly by increasing the pulse rate. This circum-
stance seems to account for the absence of any very direct relation
between pulse rate and rate of metabolism, and it also indicates that if
we were able to get some clinical measure of the systolic output at each
beat of the heart, even though it were only approximate, we might have
a better index of the rate of metabolism than can be obtained from the
pulse rate alone.

Measurements of systolic output by the nitrous oxid or any other
blood analysis method are usually out of the question in clinical work.
But von Recklinghausen " has shown that the pulse pressure varies
directly with the systolic output per beat, except for such variations as
may arise from differences in the coefficient of elasticity in the arteries.
This has recently been experimentally confirmed by Bazett." It
appeared, therefore, that the product of the pulse pressure and the
pulse rate might have a close relation to the rate of metabolism unless
differences in the elasticity of the arteries in different individuals were
so marked as seriously to distort the relation between systolic output
and pulse pressure. Even if that should be the case the P. P. (pulse
pressure) X P- R- (pulse rate) product might still be a useful clinical
method for the purpose of obtaining an indication of the direction of
metabolic changes in the same individual at different times.

These considerations, but especially the results of exi)eriments
on the effect of exercise on the P. P. X P- R- product which are
given in the next paper, made it seem worth while to collect data on

3. Lindhard: Arch. f. d. ges. Physiol. 161:2.-i3. 1915.

4. Means and Newburgh : J. Pharmacol. & E.xper. Therap. 7:441, 1915.

5. Bainbridge : Physiology of Muscular E.xercise. 1919.

6. Von Recklinghausen: Arch. f. exper. Path. 56:1. 1907.

7. Bazett: Proc. Roy. See. London. Ser. B 90:415. 191/.

542 ARCHIVES OF INTERNAL MEDICINE

the P. P. X P- R- product in normal persons under basal conditions
in order to get a control for similar observations on patients. The
results, of course, do not allow of any conclusion as to whether the
P. P. X P- R- product has a closer or even as close a relation to the
basal metabolic rate as the pulse rate alone. That question can only
by answered by direct comparison with a parallel series of determina-
tions of basal metabolism. But the figures have an interest of their
own apart from any possible significance they may have in connection
with basal metabolism. In addition to these early morning observa-
tions, the pressure and pulse rate was measured on another group of
normal individuals during the day under the conditions we have defined
above for daytime observations.

In both the basal and daytime observations the pressure and pulse
rate were observed simultaneously, the pulse rate being counted by an
assistant while the systolic and diastolic pressures were being read. A
mercury sphygmomanometer with a broad arm band was used. The
diastolic pressure was taken at the end of the third phase or at the
cessation of sound in those subjects in whom no fourth phase could be
distingitished.

In Table 1 the basal and daytime results on normals are compared.
The basal averages were obtained from eighty-nine obser\-ations on
seventy-six persons, and the daytime figures from 300 measurements
on 300 persons. Both groups comprised soldiers on active service
between the ages of 21 and 31.

These results are a contribution toward the accumulation of data
required for a definition of what is meant by normal pressure and
pulse rate. Though a great deal of work has been done on the subject,
the figures have usually been presented in such a way as to preclude
the application of the statistical methods which are essential for an
adequate understanding of their significance. The variability of the
systolic, diastolic and pulse pressure has been admirably dealt with by
Kilgore,* and Alvarez and his associates ' have published a complete
statistical review of a large series of measurements of systolic pressure
in normal individuals. Alvarez arranged his data in groups according
to the ages of the subjects, but no significant change in the systolic
pressure was found between the ages of 21 and 31. We may, there-
fore, compare his average systolic pressure of 126.5 on 2,930 men with
our average' of 127.4 on 300 men, and his coefficient of variation of
12 per cent, with ours of 13 per cent. So far as the diastolic and pulse
pressure are concerned, the only data available for the determination

8. Kilgore: Lancet 2:236. 1918.

9. Alvarez, Wulzen, Taylor and Starkweather: Arch. Int. Mod. 26:381
(Sept.) 1920.

ADDIS— BLOOD PRESSURE 543

of variability seem to be the frequency distributions of Barach and
Marks '" and of Kilgore.* All these results were obtained under con-
ditions similar to those we observed in our daytime observations. Only
a few isolated measurements were found which were carried out under
what we have called basal conditions.

TABLE 1.— XoRM.AL D.-^TA. B.-\s.\l Me.\sureme.n-ts of Sevextv-Six Norm.^i^
Compared with D.wtime Measurements on Three Hundred Normals

Averages

Basal

Day

Systolic

99

12T

Diastolic Pulse Pressure
n 28
78 M

Standard Deviations

Pulse Rate
63
80

P. P. X P. R.
1764
39S0

Basal

Day

Systolic

H-n.O

±17.0

Diastolic Pulse Pressure
-1- 9.7 -+■ 8.5
-11.1 ±13.8

Coefficients of Variation

Pulse Rate
-1- 8.1
±12.4

P. P. X P. E.
-4-635
±1550

Basal

Day

Systolic

11%

13%

Diastolic Pulse Pressure
14% 30%
14% 28%

Frequency Distributions

Pulse Rate

13%
16%

P. P. X P. B.

Pulse Pressure

oac € S. c. u a c. o IS 5

71-80 9% — 41- 60 5% 2% 11-20 27% 1% 37- 44 1% — 0-999

81- 90 16% 2% 51- 60 51% 6% 21-30 45% 9% 45- 52 12% — lOOO- 1999

91-100 34% 6% 61- 70 36% 28% 31-40 25% 23% 53- 60 40% 5% 20CO- 2999

101-110 34% 11% n- 80 39% 34% 41-50 2% 33% 61- 68 30% 13% SCOO- 3999

111-120 5% 22% 81- 90 6% 23% 51-60 1% 20% 69- 76 10% 27% 4(K»- 4999

121-130 2% 28% 91-100 — 5% 61-70 — 7% 77- 84 3% 29% 600O- 6999

131-140 — 16% 101-110 — 3% 71-80 — 4% 85- 92 2% 14% 6000- 6999

141-150 — 10% 81-90 — 3% 93-100 — 11% 7000-7999

151-160 — 4% 101-106 — 3% 8000-8999

161-lTO — 1% 109-116 — 0.6% 9C0O-9999

ra-180 — 1% m-124 — 0.6% 10000-10999

125-132 — 0.3% 110O!M199»

The differences between the basal and day averages show how
markedly the level of pressure and pulse rale is influenced by factors
associated with daytime activities. A measure which would be "nor-
mal" for the day would be unusually high if it were found under basal
conditions in the early morning, so that the interpretation of the sig-
nificance o£ any given pressure or pulse rate depends, in the first place,
on a knowledge of the conditions under which the observation was
made.

During the day the P. P. X P- R- product increases more markedly
than the other measurements because both pulse pressure and pulse
rate rise. The increase in pulse pressure occurs in spite of a slight

10. Barach and Marks: Arch. Int. Med. 13:648 (June) 1914.

544 ARCHIVES OF IXTERXAL MEDICIXE

increase in diastolic pressure and is due entirely to the rise in systolic
pressure. Since the systolic pressure, other things being equal, is
determined by the amount of blood pumped out by the heart, we may
assume that during the day the output of the heart at each systole is
increased. The increase in pulse rate shows that there are a larger
number of systoles per unit of time. The changes induced by daytime
conditions are, therefore, such as we should expect to find if there
were an increase in the volume flow of blood.

There are two factors which are certainly of importance in raising
the daytime levels. One of them is the effect of food, which has been
shown by Weysse and Lutz ^^ to result in an increase in systolic pres-
sure and pulse rate. The other factor is exercise. In Table 2 averages
are given from a group of ten normal persons under basal and day-
time conditions and after a shorter or longer period of exercise.

T.^BLE 2. — Effect of Exercise on Blood Pressure and Pulse Rate Levels
OF Normal Individuals

Conditions Systolic Diastolic Pulse Pressure Pulse Rate P. P. v p. R.

Basal 100 65 35 60 2100

Daytime 1-21 " U 74 3256

.Short exercise 137 59 78 81 6320

Longer exercise ' 168 42 126 114 14360

There is another factor — excitement — which may ha\e been
equally operative under "both basal and daytime conditions. In Table 3
daytime averages from a group of twenty-seven men who were not
excited are compared with averages from a group of twenty-seven men
who admitted that the examination excited them because they were
afraid that something would be found which would prevent them going
over with their regiment.

TABLE 3.-EFFECT

OF Excitement on the Daytime Levels of Pressure and
Pulse Rate in Normal Individuals

ConditioDS

Systolic Diastolic Pulse Pressure Pulse Rate P. P. x P. R.
l*"' 79 43 80 3440

Excitement

154 89 65 92 5980

These three factors, food, exercise and excitement, all have their
most marked effect on the systolic pressure and pulse rate. The
measurement which is least influenced is the diastolic pressure. Food
has little or no effect; exercise lowers it, and excitement increases it,
but it remains the most stable measure, so that in daytime observations
on patients when all three factors are operative, an increase in diastolic

U. Weysse and Lutz: Am. J. Physiol. 37:330. 1915.

ADDIS— BLOOD PRESSURE 545

pressure is likely to be of more significance than the same degree of
increase in systolic pressure.

The significance of any measurement which deviates from the nor-
mal average depends on the variability or degree of dispersion of the
normal measurements. This variability is given in the standard devia-
tion. The relation between the actual deviation of the measurement
and the standard deviation can be expressed in a concrete way as odds
against the possibility that any normal subject would give as high or a
higher level of pressure or pulse rate than the measurement in question.
A discussion of this statistical method will be found in a recent very
complete study of blood pres.sure and pulse rate in children by Faber
and James.'^ In Tables 4 and 5 the odds for a series of measurements
for basal and daytime conditions are given.

TABLE 4.— B.AS.\L Co.ndition's. The Odds That a Normal Individual Under

Basal Conditions Will Show as High or a Higher Level

OF Pressure and Pulse Rate as Those Given Below

Pulse

Pulse

tolic

Odds

tolic

Odds

Pressure

Odds

Bate

118

1 iQ -24

8S

1 in

■'5

43 1

1 in

■'0

77

1 in 24

.2600

1 in 27

33

78

2950

1 in 33

120

1 in ■ 36

90

1 in

4<>

45

1 in

44

121

1 in 44

1 in

46

M

80

122

1 in 55

92

67

47

Wl

81

1 in 76

3100

48

1 in

106

83

1 in 106

3150

124

1 in 86

94

1 in

11'?

49

1 in

147

125

1 in 104

95

1 in

147

50

1 in

aiH

S4

1 m 208

51 :

im

1 in 303

3300

1 in 128

1 in

128

1 in 233

96

1 in

385

53

H2.1

87

1 in 666

.526

54

I in

i»ll»

88

1 in 1000

3450

1 in a5r,

130
131
132
133

1 in 417
1 in 555
1 in 741
1 in 1000

lOO

101

1 in

770

1000

55

3550

1 in 417

3600
3650

1 in 666

3750

1 in 1111

The odds given in these tables refer only to the normal, they can-
not be interpreted as odds in favor of abnormality. To know the latter,
we should have to measure the standard deviation of the group of
abnormals to which the patient happened to belong. However, for
practical purposes of classification of cases it is permissible arbitrarily
to decide that any patient whose pressure exceeds a certain level shall
be regarded as "abnormal." The choice of that level will depend on
the nature of the work. It is sometimes desirable to have a standard
so narrow that all cases with a possible pathologic tendency to hyper-
tension will be detected. Then a level at which the odds are that only
one normal out of, say, twenty-five would give as high a pressure may
be taken. But at other times one may desire to separate a group of

12. Faber and James: Am. J. Dis. Child. 22:7 (July) 1921.

546 ARCHIVES OF IXTERXAL MEDICI XE

cases in which it is practically certain that there are no normals. Then
a level at which the chances are that only one out of 1,000 normals will
be so high may be selected as the dividing point.

These statistical methods give no information as to the nature of
the factor responsible for any unusually high pressure which may be
found in a patient. If it is a daytime observation it may be the result
of an abnormal susceptibility to purely external and evanescent causes
such as food, excitement or exercise, rather than the inner and more
lasting perversion of function we are accustomed to think of in con-
nection with hypertension.

TABLE 5. — Daytime Conditions. The Odds Th.\t \ Norm.\l Individual

Under Daytime Conditions Will Show as High or a Higher

Level of Pressure or Pulse R.\te as Those Gi\-en Below

Sys- Dias- Pulse Pulse

tolic Odds tolic Odds Pressure Odds Rate Odds P. P. x P. R. Odds

in 44 102 1 in 65 T9 1 in 56 107 1 in 69 7200 1 in 53

in 51 103 1 in 82 80 1 in 67 106 1 in 84 7300 1 in 62

in 58 104 1 in 104 81 1 in 80 109 1 in 101 740O 1 in 74

in 69 105 1 in 133 82 1 in 98 110 1 in 128 7500 1 in 86

in 78 106 1 in 170 83 1 in

in 91 107 1 in 222 84 1 in

in 106 lOS 1 in 286 Ml in

. in 125 109 1 in 385 86 1 in

in 147 110 1 in 600 87 1 in

in 175 111 1 in 666 88 1 in

in 207 112 1 in 910 89 1 in

in 250 113 1 in 1250 90 1 in

in 294 91 1 in

. in 345 92 1 in

in 417 93 1 in

in 500

1 in 161

112

1 in 204

IIH

1 in 250

1 in 323

116

1 in 526

117

1 in 714

118

119

1 in 1250

The striking difference between the averages of normal individuals
under basal and daytime conditions are the clearest illustration of the
necessity for uniformity in the conditions under which the observations
are made. It is not possible to use the basal normal for the evaluation
of pressures obtained in patients in the morning if they have been out
of bed even for a moment. The normal values for daytime measure-
ments cannot be taken as a standard for observations made on patients
who are standing or sitting, or on those who have just walked up a
flight of stars. The variability of normal blood pressure under such
conditions is not known.

1. IRKIT.\BLE HKART

The measurements on patients were carried out under the same
basal conditions as were observed with the normal controls. The cases

ADDIS— BLOOD PRESSURE 547

were in each instance tentatively diagnosed as cases of irritable heart
with the exception that a reser\'ation was made in regard to the possi-
bility of true hyperthyroidism. These cases were selected from a
larger group on the following basis. The patients all complained of
one or more of three cardiac symptoms — dy.spnea, palpitation, pre-
cordial pain; one or more of the three symptoms of vascular insta-
bility— dizziness, flushing, fainting, and, in addition, they gave some
evidence of general nervous instability. In the great majority of cases
these symptoms antedated enlistment, and often dated back to child-
hood. Persons presenting these symptoms following some infectious
disease were excluded, and in none could a diagnosis of organic cardiac
disease be made. In almost all cases, tachycardia and tremor were
present at one time or another. The hands were usually cyanosed,
cold and clammy. These patients thus seemed to belong to what has
been called the constitutional type of irritable heart, or neurocircula-
tory asthenia. Basal pressure and pulse rate measurements were made
on 138 of these patients. In Table 6 the averages, standard deviations
and frequency distributions are given.

T.'^BLE 6. — B.\s.\L Me.^surements on Patients Provisionally Diagnosed
AS "Irritable Heart" Cases

Averages

Systolic
105

DiastoUc

Pulse Pressure
33

Standard Deviations

Pulse Bate
65

P. P. X P. B. '
2145

Systolic
+11.8

Diastolic
±10.9

Pulse Pressure
±9.0

Coefficients of Variation

Pulse Rate
±8.9

P. P. < P. R.
±708

Systolic

11%

Diastolic

15%

Pulse Pressure
Frequency Distributions

Pulse Bats

14%

P. P. ;< P. R.

33%

Pulse Pressure

Pulse Bate

CI. Int.
11-20

31-40
41 .W

bi-m

Per
Cent.
16
47
26
9

CI. Int. Cent.
37-44 1
45-52 12
5360 28
61-68 38
69-76 12
77-84 6

ffi-Si I

Cl.Int. Cent.

The average level of all the measurements on the patients is slightly
higher than the basal averages for normals given in Table 1. The
greatest increase is in the P. P. X P- R- product. In this case the
difference is statistically significant, for when the "probable difference
between the averages" is determined the odds are found to be about 64

548 ARCHIVES OF IXTERXAL MEDICINE

to 1 against the possibility that the increase from 1,764 for the normals
to 2,145 for the patients can have arisen simply as the result of chance.
It would appear, then, that some factor not operative in the normal
cases had increased the P. P. X P- R- This might have been the
inclusion of some cases of hyperthyroidism or the same difference
might be due to a greater nervous excitability in the patients. But
whatever the cause may have been, it is evidently of relatively slight
importance, since all but 1 per cent, of the observations on patients fall
within the range of variation of the normal. It seemed to us to be of
particular significance that patients whose condition during daytime
examinations showed so many points of resemblance to hyperthryroid-
ism should in the early morning, under basal conditions, give evidence
of an inactivity and quietude of the circulatory system which seemed
inconsistent with the hypothesis of a state of continuing metabolic
activity, such as exists in hyperthyroidism. It is true that in individual
cases we were sometimes still in doubt, but the important question at
the time was one of group diagnosis ; and the fact that a normal basal
product, as well as a normal basal pulse rate, was found in practically
all these patients was of aid, in conjunction with other evidence, in
leading us to reject a diagnosis of hyperthyroidism for this group of
patients. At a later date Peabody, Wearns and Tompkins '^ showed
the correctness of this conclusion by demonstrating that patients of this
type, many of whom had been diagnosed as having hyperthyroidism,
had an entirely normal rate of basal metabolism.

An increase in the P. P. -X P- R- product may be found in patients
in whom there is no reason to suspect any increase in the rate of
metabolism. This is shown in the daytime measurements given in
Table 7.

TABLE 7.-

-D.WTIME

Measurements Showing Increase

IN

P.P X P.R.

Product

Conditions

Normals

Patients

Systolic
127
190

Diastolic
78

Pulse Pressure
60

77

Pulse Rate P.
80
79

p. X P B.
3960

6083

The patients were nineteen middle aged or old people who had
hypertension. They were unselected cases, except that those with
cardiac decompensation or advanced renal disease were excluded.
There were no indications whatsoever of any increased metabolic
activity, rather the reverse. Yet their P. P. X P. R- product is
markedly increased, but it is an increase due to a rise in only one factor
of the product, the pulse pressure. Furthermore, it has been shown
that any decrease in the elasticity of the large vessels between the heart

13. Peabody. Wcarn and Tompkins: Med. Clin. N. .'America 2:507.

ADDIS— BLOOD PRESSURE 549

and the brachial artery will result in an increase in pulse pressure
because the pressure will rise higher at each systole, if the vessels are
rigid than if they give way to some extent when the blood is forced
into them by the heart.'' Hence, when a product is found to be high
only because of an increase in pulse pressure it would be well to suspect
the presence of an inelastic aorta, rather than an increased output of the
heart.

Daytime measurements were also made on 156 patients who were
believed to have an irritable heart. This group includes the 138 per-
sons whose basal pressures and pulse rates were obtained. The
averages given in Table 8 are derived from 580 observations.

T.'\BLE 8. — Daytime Me-^surements on P.^tients Provision.^lly
Diagnosed as "Irritable Heart" Cases

Averaecs

Systolic
131

Diastolic

75

Pulse Pressure
55

standard Deviatious

Pulse Rate

P. P. X P.
4(i60

B.

Systolic
±18.5

Diastolic
±11.6

Pulse Pressure Pulse Bate
±17.5 ±13.4

Coefficients ol Variation

'■\^^-

R.

Systolic
14%

Diastolic

15%

Pulse Pressure Pulse Rate
32% 16%

Frequency Distributions

P. P. X P.

43%

R.

Systolic

Diastolic

Pulse Pressiue

Pulse Rate

P. P. -< P.

B.

Per
CI. Int. Cent.
71- 80 0.2
81- 90 0.5
91-100 5
101-110 8
111-120 21
121-130 27
131-140 18
141-150 10
151-160 (i
lCl-170 3
171-180 2
181-190 0.4
191-200 0.5

CI. Int.

31- 40
41- 50
51- fiO
61- 70
71- 60
81- 90
91-100
101-110
111-120

Per
Cent

2
10
26

36
21

0.2

CI. Int.
11- 20
21-30
31- 40
41- 60
51- 60

?}:^

81- 90
91-100
101-110
111-120
121-130

Per
Cent.

17
27
20
15
6

i

0.2
0.2

CI. Int.
37- 41
4»-52
53-60
61-68
6»- 76
77- 84
85-92
93-100
101-108
109-116
117-124
125-132

8
22
23
20
14

5

1
1

Cl.lDt.
0- 999
1000- 1999
2000-2999
3000- 3999
4000- 4999
600O- 5999
6000- 6999
700O- 7999
8000- 8999
900O- 9999
10000-10999
11000-11999
12000-12999
13000-13999

Per

Cent.

2

13
23
23
16
10

6

3

0.2

0.2

The averages show a little increase in systolic pressure and pulse
rate as compared with the normals— 131 as compared with 127. There
is also a somewhat greater variability in systolic, diastolic and pulse
pressures and P. P. X P- R- in the patients as compared with the con-
trols. But these differences are slight. This is of interest since it will be
shown in the next paper that a marked variation from the normal can
be demonstrated in this group of patients under conditions which
impose a strain on the cardiovascular system.

2. HYPERTENSION IN YOUNG MEN

In eighteen soldiers between 21 and 31 years of age a systolic pres-
sure of more than 150 was found on repeated daytime examinations.

550 ARCHIVES OF IXTERXAL MEDICIXE

These cases are briefly reviewed here because a comparison of their
basal and day measurements brought up a point which may prove to be
of some importance. They can be separated into four groups in accord-
ance with the conditions associated with the liypertension.

The first group comprises six cases in which no other evidence of
disease than the hypertension was found. The following is a summary
of the record of the only case in this group in which cardiac enlarge-
ment was found.

Case 1. — Sa., aged 26; no complaint.

Mother died of heart disease. Measles and diphtheria in childhood. Pneu-
monia when 8 years old.

Six months ago, while doing heavy work in France, he became breathless
on exertion. He reported for examination, and has been kept on light duty
since then. Neither before nor since that time has he had any complaint.

The heart appeared to he enlarged and this was confirmed by roentgen-ray
examination, which showed an increase in the transverse diameter. There
was a systolic murmur best heard over the aortic area and audible in the neck.
There was no thrill. The Wassermann was negative.

After restriction of fluids the night urine had a specific gravity of 1.025 on
one occasion, and 1.030 on another. No albumin or casts were found. Phenol-
sulphonephthalein excretion was 80 per cent, in two hours and ten minutes
after intramuscular injection.

Ophthalmoscopic examination showed a greater tortuosity than usual in
the retinal vessels, but no thickening of the arteries.

TABLE 9. — B.\s.\L and Daytime PRESStmES and Pulse Rates on

Hypertension Cases Not Associated with

Any Discoverable Disease

Basal

Daytime

Name

Sa

Ft

W

L. ..

170
. 140

; 130

. 120

Dias-
tolic
105
90

13
80
85

Pulse
Pressure
65
50
56
46

m

35

64
96
80
80
60
60

^P^R.^
4160
4800
4480
3600
3000
2100

Sys-
tolic
223
155
164
164
156
168

Dias-
tolic
115
80
78
99
90
95

Pulse

Pressure

108

75

86
65
66
73

Pulse
Bate
96
122
93
91
91
86

P.P. X
P.R.

10380
9150
8000
5910

Sto

Stoc

6006
6278

The second group includes eight patients who were all typical
instances of constitutional neurocirculatory asthemia. The record of
the patient Ta is characteristic of this group.

Case 2. — Ta., aged 28; complains of dyspnea, palpitation, precordial pain,
dizziness, frequent fainting and extreme "nervousness." Duration, twelve years
or more.

His father is very nervous. His mother is subject to fits. His sister has
heart trouble.

The only serious illness he remembers is typhoid fever when he was 16
years old.

He has never been able to do hard work. While he was in Italy he was
three times drafted into the army, liiit each time he was discharged on account
of disability.

ADDIS— BLOOD PRESSURE 551

The heart showed no evidence of enlargement. There was a faint systolic
murmur at the apex. The pulse rate was always regular, but usually rapid.
The hands were often blue and cold and there was a marked tremor. There
was no enlargement of the thyroid or protrusion of the eyes. The urine con-
tained no albumin or casts.

T.ABLE 10.— Basal and Daytime Pressures and Pulse Rates on

Hypertension Cases Associated with Neurocirculatory

.\STHENIA (Irritable Heart)

Basal

Daytime

Sys-

Dias-

Pulse

Pulse

P.p. X

Sys-

Dias-

Pulse

Same

toUc

tolic

Pressure

Bate

p. K.

tolic

tolic

Pressure

Bat«

P. B.

Ta

. 125

70

55

80

4400

160

m.

69

94

&t90

E

135

93

42

72

3024

158

106

52

96

4980

Sch

106

O

. 120

70

50

60

300O

154

72

82

90

7380

Fri

120

90

30

72

2160

157

88

69

5720

33

2380

177

97

80

90

7200

Greenb. ...

110

80

30

72

2160

162

97

65

6305

The third group contains two cases as.^ociated with active pyogenic
infection.

Case 3. — Ri. had experienced some shortness of breath and palpitation on
exertion two years before, but these symptoms had been greatly aggravated
following a mastoid and frontal sinus infection from which he was still suffering.

Case 4. — Ch. also complained of most of the symptoms experienced by the
irritable heart group, but he had a pronounced infection of the urinary tract
associated with evidences of renal decompensation. His urine contained much
pus and was always of low specific gravity even after restriction of fluids. His
phenolsulphonephthalein excretion was 15 per cent., two hours and ten minutes
after intramuscular injection. After intravenous injection. 8 per cent, was
excreted in sixteen minutes from the left kidney and none from the right. Pus
was seen coming from both ureters. No tubercle bacilli were found.

TABLE 11.— Basal and Daytime Pressures and Pulse Rates on Hyper-
TENSio.N Cases .■\ssociated with Pyogenic Infections

P.P. X

P.E.

X

Dias-
tolic

Pulse
Pressure

^^

P.P. X

P.E.

3060
4556

172
160

108

88

84

,t380
7040

In the last group there are two cases with advanced Bright's disease.

Case 5.— R. was 21 years old. He complained of occasional hcadajhes.
He had scarlet fever when a child. Two years ago his ankles were swollen
and painful for some weeks, and a year ago there was a recurrence of this
condition.

There was a diflfuse retinitis in both eyes. The urine contained a moderate
amount of albumin. The specific gravity never rose above 1.016 in spite of fluid
restriction. The sediment showed coarsely granular and highly refractile casts,
many of them three to four times broader than the usual cast. Only a trace
of phenolsulphonephthalein was excreted. The blood urea concentration was
171 mg. per hundred c.c.

S3J

ARCHU'ES OF IXTERXAL MEDICI XE

Case 6. — C. also complained of occasional headaches. His urine contained a
moderate amount of albumin, and the sediment showed a fair number of blood
casts. The phenolsulphonephthalein excretion was 25 per cent, in two hours
and ten minutes.

TABLE 12.— B.^SAL and Daytime Pressures and Pulse Rates on Hyper-
tension Cases Associated with Advanced Bright's Disease

Pulse Pulse P. P. x Sys- Mas- Pulse Pulse P. P. x

Pressure Rate P. E. tolic tolic Pressure Bate P. R.

51 60 3060 173 123 50 69 34SO

2S 59 1652 16- 107 60 66 3960

Only a few of the basal measurements on these patients would have
been regarded as unusual if the low average pressure and narrow range
of variation of normal individuals had not been known. But taking all
these figures together, the average systolic pressure in the early morn-
ing is 32 mm. above the normal basal average as compared with a day-
time systolic pressure 40 mm. in excess of the normal daytime average.
As a whole, then, these hypertension cases, selected because of their high
dajtime systolic pressures, still showed hypertension in the early morn-
ing when the pressor stimuli of the day were no longer in action. The
point, however, which seems to me to be of special clinical significance
is the wide variation in the degree of reduction of pressure during the
night shown by the different individuals of this series. The first case in
Group I and the two nephritic cases are distinguished from the others
by the relatively slight decrease of pressure in the early morning and
especially by the maintenance of high diastolic pressures. In a large
series of cases it might be possible to distinguish two types of hyper-
tension, one in which there is a pronounced fall in jjressure under basal
conditions and another in which the decrease is only slight. It is true
that such a distinction would be one of degree only and wnukl not
necessarily depend on any difference in etiology. But from the point of
view of prognosis it is surely of importance. In the patient Greenw.,
for instance, the average daytime pressure of 177 systolic and 97
diastolic will be more easily borne than the systolic of 167 and the
diastolic of 107 in the patient C. with Bright's disease, because ,in the
.first case the cardiovascular system is rested each night by the fall to
the basal levels of 113 and 80, whereas in the renal case there is no
remission, and the heart has continually to work against a high diastolic
pressure.

SUMMARY

1. The blood pressure and pulse rate of normal, persons were mea-
sured under what are called basal conditions, i. e., in the early morning
before the subjects had risen from bed or taken food. These results

ADDIS—BLOOD PRESSURE 553

are contrasted with similar measurements on normal persons under
what are called daytime conditions, i. e., at any time dui-ing the day
after the subjects had risen from bed and had had food, but in all cases
in the recumbent position and with the exclusion of those who had
recently undergone any muscular exertion, such as stair climbing or
drilling. The averages obtained are shown in Table 13.

TABLE 13.— AvER.vGEs of Pulse Pressure .^nu Pulse R,me

Conditions Systolic Diastolic Pulse Pressure Pulse Bate P. P. -< P. R.

Basal 99 71 28 63 1764

Daytime 127 78 60 80 3980

2. The variability of normal basal and daytime pressures and pulse
rates is defined by statistical methods and probability tables for use in
clinical work are given.

3. The significance of the diflference between basal and daytime
pressures and pulse rates is discussed, and data on the influence of
exercise and of psychic disturbance on pressure and pulse rate levels
are given.

4. Measurements of basal and daytime levels were made on patients,
and the deductions which may be drawn from comparison with the
normal data are discussed.

BOOK REVIEWS

DIAGNOSTISCHE WINKE FUR DIE TAGLICHE PRAXIS. Dr. E.
Graetzer. Verlag von S. Karger, Berlin, 1920.

Books of this tj-pe are intended for the medical student and the general
practitioner. They have a certain value and are a type of reference a busy
and perplexed practitioner will most readily consult. This text will appeal because
it deals largely with the atypical forms and symptoms of various diseases.
These are given in considerable detail. A brief resume of the normal train
of symptoms of a given disease precedes the description of the atypical forms.
This should prove a valuable feature of the book. Differential diagnoses are
generally only mentioned. It is obviously difficult and probably hardly intended
that a text of this kind should cover any subject in detail. Its function is
limited and chiefly lies in the fact that it serves as a quick reference and that
it emphasizes the unusual features of a given disease. The frequent allusions
throughout the book to other sources of reference should prove very helpful.

THE EVOLUTION OF DISEASE. By Prof. J. D.«vsz. Translated by
Francis M. Rackemann, M.D. Philadelphia : Lea & Febiger, 1921.

The subtitle of Professor Danysz' book, which is "A Discussion of the
Immune Reactions Occurring in Infectious and Noninfectious Disease. A
Theory of Immunity, of Anaphylaxis and of Antianaphylaxis," indicates, in a
general way, the scope of the subject matter. In effect, it is an argument for
the selective rather than the specific action in the process of immunity and
anaphylaxis. Throughout most of the book the reviewer follows the argument
with considerable interest. As long as the discussion is largely theoretical,
in spite of the fact that it is an illustration of special pleading, and that it
is somewhat involved with specialized terms, the argument is somewhat con-
vincing. The effect of the illustrative cases reported by Professor Danysz.
many of which are tactfully omitted by the translator, is, however, to shake the
confidence in the earlier theoretic discussion. The results from the administra-
tion of a bacterial vaccine derived from certain bacteria of the intestinal flora
in a wide variety of conditions, including neurasthenia, psoriasis, and asthma,
are too strikingly successful. It should be stated, however, that Danysz par-
ticularly emphasizes the fact that the successful issue in these cases is not
dependent on the theoretical assumption of any specific therapy, but depends
rather on the theoretic assumption of a selective action of these bacterial
antigens. Furthermore, Danysz does not believe that these conditions are due
to any of the components of the bacterial antigens. The book represents an
interesting speculation on the nature of the obscure processes of immunity and
anaphylaxis, rather than the record of scientific achievement, or of sound
application of the theory.

Archives of Internal Medicine

STUDY OF SOME CASES OF DIABETES INSIPIDUS
WITH SPECIAL REFERENCE TO THE DETEC-
TION OF CHANGES IN THE BLOOD WHEN
WATER IS TAKEN OR WITHHELD*

C. D. CHRISTIE, M.D.. and G. X. STEWART, M.D.

Several years ago ^ we published a study of a case of diabetes
insipidus in which, among other points, attention was directed to the
question, whether any well marked changes could be detected in the
blood when the water intake was greatly restricted or water taken at
discretion. The patient had an enormous diuresis and a correspond-
ingly great thirst, so that the conditions seemed unusually favorable
for the inquiry. The conductivity of the serum and the relative volume
of the serum and corpuscles were .selected for study because the con-
ductivity can be measured with great accuracy, and from the con-
ductivities of the serum and blood the percentage volume of serum can
also be obtained to a close approximation. Even when great changes
were taking place in the rate of absorption, elimination and transporta-
tion of water, it was found that the two cjuantities measured altered
only very slightly, although there seemed to be a small increase in the
conductivity and a small decrease in the relative volume of the serum
when water was severely restricted. But the extreme variation for the
conductivity of the serum was only 5 per cent, in observations made
at an interval of five days (from 78.6 to 82. 5 were the extreme
values of K X 10 ' at 5 C). The percentage volume of serum varied
from 83 to 79, again about 5 per cent. The woman had a severe anemia.
In the absence of a greater number of observations than it proved
possible to obtain on this patient, we cannot be quite sure that even the
small differences observed were directly related to the changed intake or
output of water.

We have since observed two additional cases. In neither case, how-
ever, was the diuresis as great as in the first case. Only in one of
the cases (Case 1) were we able to obtain what we considered a fairly
sufficient number of observations. Thev are summarized in Table 1.

* From the Department of Medicine of Lakeside Hospital and the H. K.

Gushing Laboratory of Experimental Medicine. Western Reserve University.

\. Christie. C. D., and Stewart. G. X.: Arch. Int. Med. 20:10 CJuly) 1917.

£

■;

1

E

£

■>!•§•-«

£

" "

5 ^ & °

E-

11

E

g 1 E 1

d

3 9

^ o : 1

E P

« c.

^ til

^ 1 £ !

= S

1 5

g =

S

O o

!5 =
E ^ o S

= 1

-c

a

n

0

%

B

^

a 1 '

JMt

1 1
1 '

ill

1

« a
I 1

1

if

i

1

t^

ill

1 o^ o S S

eall

= 1

5 "

a

i3

'II 311

iiiiiij

1 I

i !

ii

\

U

< fc c

< < fc s

< <

< <

<

S

D

>

s

£

&

__

H

a •"

J

a

s I

!tii

- -f

PS

1

i

i ^ i

EBsl

g S

S

s g s

g -^ " s

II gS

■'OS

i ' ' 5
s|| e"

S 2

<

i

1

S S E

iiis

:

'I

^

£

aaaa

^^1

a

3

a;

S - '

"= ^ 5 S

%%%%

s s

5

Si

? i s

K " s? ;;r

iiis

^. -?

5

g

i

a '^ '"

S '^ E '^

'^ a

a'

Z

u

£

" 3 E

S S S E

E s E a

E I

s

g

£

=^ g S

C o C g

ScSS

s S

s

j.

8 S ^

S 5 i s

issi

g s

3

s

■5=1

,„.

00 rH CO r-rHrH

«^ o

t-: "O"^

=--.»

TT

gg^

5sl S

s !3 B sas

gsisg

g SB

SffiS

^s

<

w a

H

d

■§

oot-jo a

a 00 ^ -,rHU5

fflc1N(N

„Jh«J,

o«5j

n!i

5

5S5 S

s s g ass

BSStS

SS5SS

sas

gjst

S

s

i

-« « m i-t-o

oo-vaic

»> ..&

,„&

2|

M

1

;J5^ I

t t t KSE:

SKiSS

^ SB

JSgJK

a s a

a 5 S E

saaa

a a

as

d d Q

d a cj d

dddd

d d

d d

a

ass

^ '5 S 8

S8SS

B !

!!

» g «

=> S ^ '^

ai:^"-.

la

<o

I-- aJ

Kg-

2 S

gs

?i

1

■S €

^«

« a

^&

i

fe

^

6.P^

? "

S

<

1

C HRISTI EST E\V ART— DIABETES IXSIPWUS 557

The observations were made at different times during eleven weeks,
and the relative constancy of the serum conductivities, in spite of the
changes imposed in the intake of water and the accidental changes
which might be expected to occur o^■er so long a period, is quite
striking. The greatest variation was from 72.8 to 80.7, and the
average of all the observations 76. The values obtained after injec-
tion of Congo red solution for estimation of the blood volume are
left out in calculating the average, although there would be practically
no change if they were included, since the quantity injected causes
no sensible alteration in the conductivity. The variation in the per-
centage of serum was greater than that in the serum conductivity,
from 42.6 to 63 per cent., the average percentage for all the observa-
tions being 55.7. It so happens that all the serum percentages in the
first series of observations (February 6) are lower than in any of the
other series. It is not known that this was due to any experimental
error. The only error which could possibly have caused such a result,
so far as we can see, would be the loss of some of the serum in the
manipulations before the blood was brought to the laboratory. But it
is very difficult to understand how this could have afifected all the
specimens. Also duplicate specimens were taken in two observations,
and the duplicate determinations are identical. It seems, therefore,
more likely that the obserA'ations are correct, and that, for some reason
unknown to us, there was a considerable increase in the relative volume
of serum between February 6 and February 7. The hemoglobin con-
tent was seen to diminish somewhat throughout the series of experi-
ments, accompanied by a slight diminution in the erythrocyte count.
The average quantity of blood taken for each determination was not
less than 20 c.c, so that it is possible that the mere loss of blood
might, in part at least, account for this. The conductivities of the
serum specimens of February 6 are not out of line with those of the
rest of the series. If the serum percentages of February 6 are omitted
the variation for the rest of the observations is from 54.1 to 63 per
cent, and the average 58. There was only slight anemia in this patient,
the erythrocyte count being in the neighborhood of 5,000,000, at the
beginning, declining to 4,500,000 toward the end of the period.

As in the previously reported case, the percentage of serum was
determined by the hematocrit as well as by the electrical method. It
will be seen, as before, that the longer the centrifuge is turned the
closer does the hematocrit reading approximate to the result of the
electrical determination. The number of minutes rotation of the
hematocrit (at about 4.000 turns a minute) is given in parentheses
after the corresponding serum percentages.

The general plan of the observations was as follows : The patient
being on his usual diet and taking water at discretion, a blood sample

558 ARCHIVES OF IXTERXAL MEDIC IX E

(or generally two samples for duplicate determinations) was obtained,
and tile blood defibrinated. The electrical conductivity of the blood
and serum and the percentage volume of serum were determined. The
[latient was then deprived of food and water, in the case of C. S. (Case
1 ) for twenty-four hours or longer, as he bore the deprivation well.
Blood samples were obtained at the end of the period. Then he was
allowed food and as much water as he could drink, and blood drawn
at the end of thirty minutes, and again at the end of five or six hours,
water and food being taken. Several sets of observations of this type
were made on C. S. In none was there any material difference in the
percentage of serum in the samples taken before and at the end of the
period of abstention from water. The conductivity of the serum was
also practically unchanged, except in the observations of March 10
to 11. when there was an apparent increase of about 10 per cent, at
the end of twenty-four hours abstention from water.

The samples taken half an hour after renewed water ingestion
showed practically no change in the serum conductivity, while in several
of the sets of observations it seemed that a slight diminution in the
percentage of serum occurred. More striking was the increase in the
percentage of serum in the samples drawn from five to six hours after
the taking of water had been resumed. This is seen in the observations
of February 6. February 7 to 8, and February 9 to 10. The increase
in the relative volume of the serum was not apparently accompanied
liy any diminution in the concentration of the salts, since the serum
conductivity remained unaltered, or if anything, underwent a slight
increase.

In Case 2 (U..A..). like Case 1. one of medium severity, the range
of variation in the conductivity of the serum in all the observations
was from 77.i to 86.8. .Although the mean of all the observations
(82.2) was somewhat higher than in Case 1. the maximum range
was about the same. However, the number of observations obtained
in Case 2 was smaller than in Case 1. and they were spread over a
much shorter period. In the most complete series in Case 2 (May 5
and 6) (Table 2) there was an increase in the conductivity of the
serum after abstention from water for twelve hours (from 79.2 to
86.8. This was accompanied by a more marked diminution in the
percentage volume of the serum (from 70.9 to 58.3 per cent.). On
taking water, the proportion of serum increased. However, in the
first experiment (April 20 to 22) when abstention was carried to the
pf)ssible limit in this case (twenty-seven hours) a slight diminution
of the conductivity was seen (from 82.0 to 77.3) accompanied by
a marked increase in tlie proportion of the serum, while on April 30
the ingestion of 2 liters of water in twenty minutes after twelve hours
abstention was associated with a slight increase in the conductivity of

itc

1

1

!

1

£
1

i

Hi
llj

IK

M

3<;

\

i

j

1

1

E

E

1^1

iii

o
2

1

iia

1

iiS

1

g

1

$

ARCHll-ES OF IXTERXAL MEDICINE

the serum, and a distinct diminution in the proportion of serum. We
prefer not to attempt a hypothetic explanation of these variable results,
simply pointing out that in Case 2 the relative volume of the serum
was more variable than in Case 1, the extreme range being from 50.6
to 74 per cent, and the average of all the observations 62 per cent.
It should be noted that in Case 2 deprivation of water was not nearly
as well borne as in Case 1, the patient coniplaining much more of
thirst. Also U.A. (Case 2) was allowed food during the period of
deprivation of water, although he ate less than usual.

The observations in these cases illustrate, perhaps, even more clearly
than those previously published, how trifling the changes in the con-
centration of the blood plasma, as regards the electrolytes, may be
when great changes in the ingestion, excretion and transportation of
water are in progress. So far as our experience goes it would seem that
it is only in exceptionally favorable circumstances that even minute
changes in the conductivity of the serum, a quantity capable of being
measured to so considerable a degree of accuracy, can be detected with
certainty. Observations which profess to demonstrate considerable and
constant variations, associated with the taking or withholding of water
should, we think, be received with reserve.

The relative volume of the blood plasma may undergo somewhat
greater variations. This agrees with the conclusion of Farkas,= in an
extensive research on the influence of water and salt given with the
food on the water content of the organs in some of the domestic
animals. In sheep caused to drink large quantities of water with or
without sodium chlorid, the osmotic pressure and the concentration of
the electrolytes remained the same; but the water content of the blood
was increased in the sheep which received salt as well as water.
According to Adolph,^ the drinking of isotonic salt solution by normal
persons is accompanied by a measurable diminution in the hemoglobin
content of the blood, indicating some increase in the water content.
The possibility should not be lost sight of that in cases of diabetes
insipidus the variations may be even less than in normal individuals,
the urine secreting mechanism being, perhaps, even more responsive
to slight changes in the blood, or the tissues less capable of storing
any excess of water.

We do not intend to discuss the mechanism by which the osmotic
pressure and the concentration of electrolytes in the plasma (as
measured by the electrical conductivity) are mamlained relatively con-
stant during the absorption or excretion of large quantities of water.
The exchange between the erythrocytes and the plasma, as well as the
exchange between the tissue liquids and the plasma through the

2. Farkas, K. : Mitth. aus. d. Konigl. Ungar. Tierphysiol. Vcrsuchsstatioii
Budapest. No. X Berlin. 1908; Landwirtscliaftliclie Tahrb.. 1908.
X Adolph, E. F.; J. Physiol. 55:114, 1021.

CHRISTIE-STEirART— DIABETES IXSIPIDUS 561

capillary walls, must play a part. As to the relative volume of plasma
and corpuscles, although the changes were probably greater than those
in the conductivity of the serum, they were too small in amount, and
not constant enough in sign, to permit the assumption that when
water was withheld, any important part of the water which continued
to be excreted could have been credited to a diminished water con-
tent of the blood, or that when water was again taken, any important
part of it went to recoup the blood for its previous loss. The chief
changes must have been in the tissue water. The slight apparent
diminution in the percentage of serum in the first half hour after
resumption of water intake, if it is a genuine diminution, might
possibly have been associated with a preliminary speeding up of the
diuresis. The increase in the proportion of serum a few hours later,
perhaps to something more than the amount present at the beginning
of the period of abstention, is most naturally associated with a rapid
absorption slightly outstripping the diuresis. However, this is not the
only possible explanation, and as has been previously said, speculations
founded on such data as we have been able to obtain would, in the
present state of our knowledge, be of little value. This is well illus-
trated by the marked increase in the proportion of serum, accompanied
by a slight decrease in the conductivity, seen in the observations of
.April 20 to 22 in Case 2, after twenty-seven hours abstention from
water. Whether the fact that some solid food was taken in this case
influenced the result, it is impossible to say.

REPORT OF C.XSES

Case 1.— C. S., male, married, aged 39, was admitted to the Lakeside Hos-
pital medical service, Feb. 3. 1921.

Past //uforv.— Essentially negative, aside from an attack of pneumonia five
years ago. No historj- of any venereal infection.

Present Illness.— The patient dates his present trouble from July. 1920, when
he noticed that he was becoming more irritable and much more easily fatigued.
He had been an active man, but he now became so tired that his one desire was
to lie down and sleep. He arose late each morning and retired early, and usually
spent most of his Sundays in bed. Four months ago he began to have head-
aches, which he described as "heavy aches," coming every two or three days
and lasting from a few minutes to half a day. They seemed to distress him
more at night, and would sometimes awaken him. About one month after the
headaches started, he noticed that, he was gradually drinking more water
each day and was passing an excessive quantity of urine. He said that his
thirst reached such a degree that he would drink from three to four glasses of
water every half hour and pass a corresponding amount of urine.

Physical EAratnination.—Tht patient is a well developed and well nourished
man, and gives the appearance of being in good health. He is extremely
neurotic, apprehensive, prone to complain and very restless. There is consid-
erable drvness of the skin and a slight anemia. The eyes and eye grounds are
normal. There is no evidence of cardiac enlargement. Blood pressure is not
elevated. There is no acceleration of the pulse rate; no abnorrnal physical
signs about the abdomen or the extremities. A complete neurologic examina-
tion revealed no abnormal findings.

562 ARCHirES OF IXTERXAL MEPICIXE

Urine. — The urine was pale and of low specific gravity; no albumin or sugar.
The quantity varied from 6 to 10 liters in the twenty-four hours.

Blood. — The blood showed: hemoglobin from 80 to 90 per cent. (Taliquist);
white blood count, from 5,000 to 7.000: red blood count varied from 4.500,003
to 5,000,000. Blood sugar estimations showed from 0.085 to 0 050 gm. per
hundred c.c. (by Lewis and Benedict method) ; blood urea varied between
0.025 and 0.032 gm. per hundred c.c.

Spinal Fluid. — Lumbar puncture revealed no increase in pressure : fluid nor-
mal in color ; no cells. The Wassermann, globulin and gold chlorid tests were
negative.

Head. — Roentgenograms of the region about the sella turcica showed no
evidence of a pathologic process.

Kidneys. — Renal test meals, given when the patient was getting three daily
intramuscular injections of 1 c.c. of a pituitary e-xtract ( pituitrin) showed an out-
put of from 2 to 3 liters and that the patient's kidneys were quite able to con-
centrate the urine.

Dec. 15, 1921, the patient, who had become insane, shot himself in the left
chest and died December 25. Diagnosis : "encysted hematoma, empyema atid
collapse of the left lung." The necropsy gave no information as to the possible
cause of the polyuria. The pathologic examination of the brain, including the
pituitary, showed nothing abnormal.

C.'\SE 2. — U. A., male, single, aged 16: admitted to Lakeside Hospital medical
service, April 7, 1920.

Past History. — He had had when a child the ordinary diseases.

Present Illness. — He dates his present trouble from the time he was 5 years
old, when he had a cold and was feverish one night, and began drinking water
excessively and urinating frequently. He was apparently worse at that time,
from the statement of his parents, as he was not allowed to enter school until
he was 7 years of age because of the trouble. Since the complaint began,
eleven years ago, there has been no marked abatement of symptoms.

Physical Examination. — The patient is well developed and well nourished,
and does not have the appearance of being ill. There are a few small brown
pigmented areas over the face. The eyes are normal : the pupils react to light
and accommodation; eye grounds are normal; no disturbance in the field of
vision. The heart is not increased in size, and the pulse is regular and of good
volume. The systolic blood pressure varied between 130 and 120 and the
diastolic between 79 and 60. Examination of the abdomen and extremities was
negative. A complete neurologic examination revealed nothing pathologic.

Urine. — The urine varied between 5 and 10 liters per twenty-four hours. It
was pale, and the specific gravity varied between 1.002 and 1.005. except when
the fluid intake was restricted or when the patient was given piti'itary extract
(1 c.c. three times daily, intramuscularly), and then the output was cut down and
the specific gravity elevated. No albumin, casts or sugar.

lilood. — The blood showed: hemoglobin, 90 per cent. (Taliquist) : white cells,
10.000: red cells, 5,200,000. Blood sugar. 0.10 gm. per hundred c.c, and blood
urea, 0.030 gm. per hundred c.c.

Head. — Roentgenograms of the region about the sella turcica revealed noth-
ing abnormal.

Kidneys. — The phenolsulphonephthalein excretion was 78 per cent, in two
hours.

Discussion of the Clinical .Ispccts of Cases 1 and 2. — There was
nothing unusual ahout these cases other than that tlie patients had a
diabetes insipidus of medium severity. Nothing was made out, either
on physical examination or from the laboratory procedures, which gave
lis any clew to a possilile etiology.

CHRISTIE-STEIVART— DIABETES IXSIPIDUS 563

Both Cases 1 and 2 responded to the intramuscular injection of
pituitary extract when given in 1 c.c. doses. The thirst and diuresis
promptly subsided, and if three daily injections of 1 c.c. were given,
the urine output could be cut down from 5 to 12 liters to below 21^
liters in the twenty-four hours. Feeding of the fresh pituitary gland
to the patients had no effect, and the oral administration of commercial
pituitary extract did not modify the diuresis. There was no evidence
obtained from any of the tests for kidney function to bear out the
contention of Erich Meyer ' and others that the diuresis in diabetes
insipidus is in any way associated with, or dependent on, a pathologic
alteration in the kidneys. Both our patients had i>erfectly normal
phenolsulphonephthalein excretions. There was no evidence of nitrogen
retention in the blood : and when the diuresis was lessened by pituitary
extract and the patients were put on a renal test meal their kidneys
showed ample ability to concentrate the urine. These clinical observa-
tions bear out in the main the findings which we ha\e reported in
another ca.se.*

In these two cases the condition was chronic, of Imiij standin<j; and
of unknown etiology. A third case (C.\'.) was studied in wliicli ihe
polyuria, apparently associated with a lesion of the base of tlif lirain.
came on suddenly, and disappeared after lumbar puncture.

C.VSE 3. — C. v.. male, married, aged 28. was admitted tn the Lakeside Hos-
pital medical service. Nov. 27. 1920.

Past History. — This patient had had the ordinary diseases of childhood,
and typhoid fever when about 11 years of age (?). .■\t 25 years he had pneu-
monia, and for some time after was weak. He was thought to have tubercu-
losis of the lungs at that time and was advised to go to California, which he
did. He came back apparently in good health, and then had influenza in
February, 1920.

Present Illness. — .\pril 29, 1920, after moving his household effects, at which
he had worked very hard, riding around most of the day in a truck, he felt
numb over the whole left side of his body. He said that he felt as if the
left side of his body belonged to someone else. He was totally unable to
move his left arm and hand and also his left leg. This paralysis disappeared
in about two hours. It had never appeared prior to that time.

Since the attack of influenza and paralysis the patient enjoyed good health
until last night (Xovember 26) at 4 a. in., when he awakened with a dull
headache, which he had not had for two years at least, and a slight vertigo,
both of which persisted until today (Xovember 27). .As he awoke last night
he says he had a most extreme sense of thirst in the interior of his nose and
not in his mouth. He has since emphasized the fact that the thirst was in
his nose and very intense. He also noticed that he had a markedly overfull
bladder, which caused him pain above the pubic crest. When he voided urine
he said he was sure that he passed more than a quart. His thirst persisted
all day yesterday and is still present today in the same intensity. Yesterday
he voided from 12 to 14 quarts of urine, and he says that he is maintaining
the same average of urine output today. Today he feels weak and tired, in
addition to his other complaints, and has a general soreness over his abdomen,
which he attributes to his frequently overdistended bladder.

4. Meyer, Erich: Deutsch. .Arch. f. klin. Med. 83:1, 1905: Ztschr. f. klin,
Med. 74:352. 1912.

364 ARCHH-ES OF IXTERXAL MEDICIXE

f'hysical Examination. — The patient is an intelligent young man, well pro-
portioned and of healthy appearance. There were no abnormalities about the
head, eyes, ears, nose and throat, except those detailed under the neurologic
examination. There were a few enlarged lymph glands and a moderate uni-
form hypertrophy of the thyroid gland. Aside from a slight impairment in
the movement of the left upper chest, the examination of the lungs was nega-
tive. The heart showed no evidence of enlargement or valvular defect. The
pulse was not accelerated ; systolic blood pressure was 125, diastolic, 80. The
abdomen was normal and an examination of the rectum, genitalia and extremi-
ties revealed nothing abnormal.

Xcurologic Examination. — The mentality of the patient is good and he
cooperates well. Cranial Nerves: There was a loss of the sense of smell in
the left side of the nose. The gross vision not impaired. There was a con-
centric limitation of the left visual field. The left pupil reacts more slowly to
accommodation than the right. The external muscles of the eye were normal.
There was impairment in the sense of touch and pain over the left forehead
and zygomatic area. The lower and lateral part of face were unaffected.
There was slight paresis of the left facial nerve. There was a diminution in
both bone and air conduction in the left ear. There was impairment in the
sense of taste and common sensation on the posterior third of the tongue,
but muscles of palate and pharynx seemed intact. There was some atrophy of
the left side of the tongue.

Sensory Examination. — There was a very gross impairment in all sensa-
tions over much of the left side of the body. This included primarily touch,
but there was a corresponding diminution to pain, heat and cold and to
vibration. The areas which were most markedly involved were the left fore-
head, left zygomatic area, left neck, left finger tips, distal phalanx of all toes
on the left foot and the medial surface of left leg, etc. There was no ataxia;
gait and station normal, and no impairment in complemental opposition; mus-
cles were normal. Both skin and deep reflexes were apparently normal. There
was nothing made out to suggest any involvement of the sympathetic nervous
system.

Our conclusion from the neurologic examination was that the patient had
a basilar lesion which was either a tumor or a serous meningitis.

Urine. — The urine was pale, with a specific gravity of l.OOS. The urine
never contained any pathologic elements.

Blood. — Blood examination showed hemoglobin, 100 per cent. (Tallquist) ;
white blood cells, 7,000; red blood cells, 5,800,000.

Spinal Fluid. — .About 8 c.c. of clear colorless fluid was removed by lumbar
puncture. It contained 2 cells to the field; no increase in pressure or in the
globulin content. The Wassermann was negative. Blood Wasscrmann was
also negative.

Head. — Roentgenograms of the area about the base of and of the whole skull
were negative.

Kidneys. — Phenolsulphonephthalein excretion was 57 per cent, in two hours,
and a renal test meal, given after the diuresis had subsided, showed that the
kidneys had ample ability to concentrate the urine.

Pisdiarge .Xotc. — This patient had polyuria for less than three days. It
. made its appearance early in the morning of November 27, and November 29
a lumbar puncture was done for diagnostic purposes, after which the polyuria
disappeared, even though nothing pathologic was found in the spinal fluid and
there was no increase in pressure. The patient, while the polyuria existed,
excreted about 12 liters of urine per day. He had a severe headache for
several days following the lumbar puncture. He was discharged from the
hospital Dec. 7, 1920. His urine output had not exceeded 1,300 c.c. in any
twenty-four hours after the lumbar puncture. There were no demonstrable
changes in the neurologic signs on discharge.

CHRISTIE-STEWART— DIABETES IXSIPIDUS 565

Case 3 was one of severe polyuria of very acute onset. There
was evidence that the patient had an intracranial lesion which was
probably located at the base of the brain. It had apparently involved
the olfactory nerve, the optic nerve peripheral to the chiasma and the
fifth, seventh, ninth and twelfth cranial nerves, the involvement prob-
ably being nuclear, with, perhaps, also some encroachment on the
sensory areas in the thalamic region. Our impression was that there
was either a tumor or a serous meningitis at the base. A lumbar
puncture was decided on for aid in diagnosis. Although the cerebro-
spinal fluid was not under increased pressure and its examination was
negative, after this procedure the increased thirst and diuresis dis-
appeared. In Case 1 lumbar puncture was also done, but there was
no effect on the thirst or diuresis.

There has been no return of the diuresis since C. V. (Case 3) left
the hospital, and for some months he appeared to be in good health.
However, after from three to four months, he began to show signs
of loss of mental balance, the derangement taking largely a religious
turn, and he had to be discharged from his work. He is now (October,
1921) in a sanitarium, suffering from "nervous breakdown."

Any hypothesis which we could advance as to the cause of the
polyuria in this case would be mere conjecture. The course of the
onset, the recovery following the lumbar puncture and the fact that
there has probably been no increase in the neurologic signs since he
left the hospital, make the diagnosis of serous meningitis seem the
most likely. Herrick ^ reported a case in 1912 in which the polyuria
ceased after lumbar puncture, and in 1918 Cammidge " reported a case.
Cammidge thought his patient had a serous meningitis due to a para-
syphilitic state.'

It had been intended to study the blood in the same way as in the
other cases. But owing to the disappearance of the polyuria after
lumbar puncture only two sufficient samples of blood were obtained,
one before abstention from water was begun and the other one hour
after the taking of water had been resumed. The first specimen was
secured on the day when lumbar puncture was done and the second
on the following day when the diuresis had already subsided. The
conductivity of the serum (Table 3) in the second sample was some-
what greater than in the first and the percentage of serum was
somewhat less. The blood specimen obtained at the end of the twenty-
four hour period of abstention was so small that only a hematocrit
determination could be made. The percentage of serum after thirty-
eight minutes rotation was 45. whereas in the sample taken before

5. Herrick: Arch. Int. Med. 10:1 (July) 1912.

6. Cammidge, P. J.: Practitioner 105:244, 1918.

7. We desire to express our thanks to Dr. R. G. Pearce,
referring Case 3 to us for study.

566 ARCHIVES OF IXTERXAL MEDICIXE

abstention it was 49 after twenty-five minutes rotation. Probably,
therefore, there was some diminution in the serum percentage at the
end of the period of abstention. At any rate there was no increase.
It is impossible from the hematocrit determination alone to say more
than this.

SUMMARY

The regulation of the excretion of water by the kidneys was studied
in two cases of diabetes insipidus presenting the typical features, and
in one case of polyuria of acute onset, apparently associated with a
brain lesion. In the last case the polyuria disappeared permanently
after lumbar puncture, but the patient eventually developed symptoms
of mental derangement.

Blood specimens obtained immediately before and immediately after
a long period of complete deprivation of water (twenty-four hours
or more) showed no definite differences in the electrical conductivity
of the serum, which could be associated with changes in the rate at
which water was being absorbed, transported and excreted, although
the conductivity can be measured with great accuracy. The same was
true of the percentage volume of serum.

Comparison of blood specimens procured within half an hour, and
again after five or six hours, after the resumption of water drinking
with the specimens obtained just before or just at the end of the period
of water deprivation also revealed differences in the conductivity of
the serum so slight and so inconstant that it was impossible to connect
them definitely with changes in the intake of water.

The percentage of serum, in the observations which we were able to
carry out completely, seemed to be somewhat greater in the specimens
taken after five or six hours, than in the specimens taken half an hour
after resumption of water drinking.

The regulation of the concentration of electrolytes in the pla.sma
and of the relative volume of plasma and corpuscles in the blood was,
therefore, at any rate as fine in these cases of diabetes insipidus, in
spite of the great variations induced in the quantity of water trans-
ported, as in normal persons. It is possible, indeed, that in this condi-
tion the renal excretory mechanism is even less tolerant than normal
of any excess of water in the blood, or the tissues less capable than
normal of storing an excess of water.

As in the ca.se previously published,' no evidence was obtained that
the condition was associated with any patliologic change in the kidneys
The various tests of efficiency of renal function gave normal results.
When pituitary extract was administered the kidney showed normal
])Ower of concentrating the urine.

CLINICAL CALORIMETRV. XXX. METABOLISM
IX ERYSIPELAS *

WARREX COLEMAX, M.D.. DAVID P. BARR, M.D

AND EUGENE F. Du BOIS, M.D.

With the Technical .■^ssist.'^nce of G. F. Soderstrom

NEW YORK

In the study of fe\er in the human subject it is difficult to select a
disease which lends itself well to e.xperimental conditions. Patients
with certain fevers, such as pneumonia, are so seriously ill that one
hesitates to make even the simplest observations. Some of the other
fevers are so mild that they do not give one a chance to study high
temperatures. In others, the possibility of contagion must be con-
sidered and it is hardly ju.stifiable, for instance, to study measles or
scarlet fever in a room adjoining a general ward. Typhoid fever, which
is in many respects ideal for experimental work in the respiration
calorimeter, was thoroughly investigated by Shaffer and Coleman ' and
later by the staff of the Russell Sage Institute of Pathology. -

It seemed desirable to determine whether other acute infectious dis-
eases present phenomena similar to those of enteric fever. For this
purpose er3'sipelas was cho.sen. There were several reasons for its
.selection. In the first place, the inflammatory' process can be observed
and a fairly good prognosis can be made from day to day. The
temperature is high and the toxemia often severe and yet the patient is
not exhausted by simple movement such as the necessary transfer
from the bed to the calorimeter. Moreover, temperature fluctuations
are often rapid. It was hoped that further information might be
obtained concerning temperature regulation in the body.

In many respects, erysipelas was a disappointment from an experi-
mental standpoint. Some of the patients were of alcoholic habits'
which dulled their intelligence and accentuated the usual delirium of
the disease. Cooperation in the collection of twenty-four hour speci-
mens was obtained with difficulty. The mental state of the jjaticnt
often rendered calorimetric observations impossible at times when it
would have been otherwise desirable to make them. The appetite was

* From the Russell Sage Institute of Pathology in affiliation with the Second
Medical Division of Bellevue Hospital.

1. Shaffer and Coleman: Protein Metabolism in Typhoid Fever, Arch. Int.
Med. 4:538 (Oct.) 1909.

2. Coleman, W.. and Gcphart. F. C. : Clinical Calorimetry. Paper 6, Notes on
the Absorption of Fat and Protein in Typhoid Fever. Arch. Int Med. 15:882
(•May) 1915; Coleman. W.. and Du Bois. E. F. : Paper 7. Calorimetric Obser-
vations on the Metabolism of Typhoid Patients with and Without Food. .^rch.
Tnt. Med. 15:887 fMay) 1915.

568 ARCHIVES OF IXTERXAL MEDIC IX E

always capricious and it was many times impossible to induce the
patients to take the requisite amount of food. Furthermore, the
expected wide fluctuations of temperature were not observed in the
calorimeter. Special efforts were made to observe these changes.
Experiments were undertaken at night in the hope of obtaining the
usual sharp drop in temperature during the morning hours. For some
reason, however, fluctuations in temperature always seemed to be less
marked when the patients were in the calorimeter than when they
were in the erysipelas ward. The same difiiculty was experienced in
the study of falling temperature in tuberculosis.

Several determinations of the total heat production have been made
in facial erysipelas by Riethus ^ and Grafe.* Riethus found an increase
of 41 per cent, in the metabolism, Grafe, in one case with a temperature
of 39.5 F, found the heat production 40 per cent, above the level which
it assumed after recovery. Loening,^ in a comparative study of the
nitrogen losses in various fevers, published the results in eight cases
of erysipelas. RoUand " in one case with a range of temperature between
37.5 and 39 C. gave 46 calories per kilo in the food with 12.1 gm.
protein daily and found a negative nitrogen balance averaging 0.67
gm. per day. She considered this as evidence against a toxic destruction
of body protein. Kocher " was able to administer to four erysipelas
patients diets containing from 3,200 to 4,300 calories with only 1.8 to
2.2 gm. nitrogen. On such diets, normal men excrete only from 2 to
4 gm. nitrogen even though they perform severe muscular exercise. The
nitrogen excretion of the erysipelas patients was from 9 to 20 gm.
even after several days of this diet. Grafe,* one of the chief opponents
of the theory of toxic destruction, confirmed these results. He gave
an erysipelas patient a diet containing 66 calories per kilo and no pro-
tein. The urinary nitrogen dropped from 25.9 gm. to 7.7 gm. on the
fifth day of the diet but would not fall below this point.

The various urinary constituents have been determined by most of
the investigators who have studied the nitrogen metabolism. Unusually
complete analyses were made by Kocher.' He found during the febrile

3. Riethus, O. : Beobachtungen iiber den Gaswechsel kranker Menschen und
den Einfluss antipvrctischer Medicamente auf denselben. Arch. f. exper. Path. u.
Pharmak. 44:239, 1900.

4. Grafe, E. : Untersuchungen iiber den Stoff- und Kraftwechsel in Fieber
Zur Genese dcs Eiweisszerfalls bei Fieber und bei Arbcitsleistung, Deutsch.
Arch. f. klin. Med. 101:209, 1911.

5. L,oening, K. : Experimentelle und klinische Untersuchungen iiber Eiweiss
und Stoffwcchsel im Fieber, Klin. Jahrb. 18:199. 1908.

6. Rolland, A. : Zur Frage des toxogenen Eiweisszerfalls im Fieber des
Menschen, Deutsch. Arch. f. klin. Med. 107:440, 1912.

7. Kocher, R. : Ueber die Grosse des Eiweisszerfalls bei Fielier und bei
Arbeitsleitung, Deutsch. Arch. f. klin. Med. 115:82. 1914.

8. Grafe, E. : Zur Genese des Eiweisszerfalls ini Fieber. Deutsch Arch. f.
klin. Med. 116:328, 1914.

COLEMAX ET AL.— METABOLISM IX ERYSIPELAS

periods a considerable increase in the excretion of creatinin. At the
height of the disease, this reached from 2.4 to 2.6 gm. per day, the
uric acid from 0.8 to 2.0 gm. and the ammonia from 1.8 to 3.0 gm.

Our own work includes eight observations on the basal metabolism
of five patients during the acute stage of the disease. Two of the
five were studied on the first day of normal temperature. The respira-
tion calorimeter of the Russell Sage Institute of Pathology was
employed. The methods have been described in Paper 4 ^ of this series.
Observations were also made on the nitrogen equilibrium and on the
weight curves during the infection.

The character of the cases studie.d can be judged from the follow-
ing histories.

REPORT OF C.^SES

Case 1. — Erysipelas of neck and back.

History. — Arshel A., a peddler, born in Russia, 29 years of age, was admitted
Oct. 9, 1916, and discharged cured Oct. 24, 1916. He drinks one glass of beer
a day and smokes cigarets lo excess.

September 25 a boil developed on the left side of the back of the neck.
It was incised on the twenty-eighth but the area of incision became red and
swollen. The inflammation spread very rapidly until it covered the neck and
back. He felt feverish but had no chill.

Physical Examination. — The patient is an undersized, fairly well developed
and nourished young Jew, acutely ill but mentally alert and rather appre-
hensive. The tongue is moderately dry with a thick white coat. Lymph nodes
are not enlarged. On the back of the neck is a small incision nearly healed.
The area of inflammation extends from the hair line to two inches below the
inferior angles of the scapulae and from the right to the left deltoid. The area
is brawny and dark red in color and is sharply demarcated from the surrounding
skin, but without a definitely raised edge. There are many broken blebs. The
spleen is not palpable.

OCT. 9

10

M 12

13

1*

15

16

17

18

19

TEMP.r'

A

A ^

K

h

f

/A

^t"

./)

\h

/\

A

V IV IV

V v

\l

v/^

\

V/\

A

-Arshel .\. (Ca

iperature chart.

Laboratory E.raminalion.—flihe urine shows a trace of albumin : no casts.
Blood : October 16 : Leukocytes, 25,000 ; polymorphonuclears, 93 per cent.

October 13 he was in the calorimeter from 10:30 a. m. to 2:30 p. m.
October 14 the inflammation had extended to the elbow on the left side to
within two inches of the elbow on the right. The inflammation of the back was

9. Gephart, F. C, and Du Bois. E. F. : Clinical Calorimetry, Paper 4, The
Determination of the Basal Metabolism of Normal Men and the Effect of
Food, Arch. Int. Med. 15:835 (May) 1915.

.^RCHlfES OF IXTERXAL MEDICIXE

less marked, and his general condition was improved. From 6 p. m. October
15 to 2:45 a. m. October 16 he was again in the calorimeter. By the twenty-
first all active inflammation had disappeared and he was discharged as cured
October 24.

Case 2. — Facial erysipelas.

History. — James \V.. a fireman born in the United States, 51 years of age.
was admitted Oct. 11. 1916, and discharged as cured on Oct. 2\, 1916. He
had gonorrhea in 1896 without complications. He denies syphilis and says
he does not drink.

October 9 he had noticed that the left side of his nose was swollen
and red. He had a chill in the afternoon. During the ne.xt twenty-four hours
the swelling spread rapidly over the left side of the face.

Physical Examination. — Patient is a well developed, poorly nourished, rather
surly American, acutely ill. His tongue is red at the edges, shows a white coat
and is very dry. Over the bridge of the nose and spreading over the entire
left side of the face is a diffuse, red, hot area of inflammation. The edge is
raised, firmly and sharply demarcates the area from the surrounding skin. The
pulse is slow, full and dicrotic. The spleen is not palpable. Lymph nodes
are not enlarged.

OCT. II 12 13 14 15 16 17 18

TEk
104

102

100

P.F.

A

'^

A

N

/

1 -A

\

^^

I

s

Fig.

nines W.

Temperature char

Blood Pressure: Systolic,

Laboratory E.vainination. — Urine is negati'
130 mm. ; diastolic, 70 mm.

October 13 the area of inflammation had spread to include the left car.
The left eyelid was swollen and shut. On the fifteenth the right forehead
and ear were swollen, tender and red. The condition of the left side of the
face was much improved. October 17 he was placed in the calorimeter from
10:15 a. m. to 1:15 p. m. At that time his temperature was normal and most
of the signs of inflammation had disappeared. Both ears were slightly swollen.
The left ear was desquamating. He was discharged as cured October 21.

Case 3. — Facial erysipelas.

History. — Odysseus B., a cigaret maker, born in Greece, 46 years of age,
was admitted Oct. 24, 1916, and discharged cured Xov. 1, 1916. He was
operated on for fistula in apo in March, 1916, at an Italian hospital. He drinks
two glasses of beer a day ; no whisky.

October 20 he had a slight pain in the abdomen. On the afternoon of
October 21 he had a severe chill with high fever. On the morning of October
22 he noted a slight redness on the right side of the nose. This spread gradually
to the left side and ultimately covered his cheek and forehead.

Physical E.vamination October 27. 1916. — Patient is a well developed, fairly
well nourished man, acutely ill. The tongue is moist with a thick, white coat.
Over the left cheek, car, forehead and scalp and over the right forehead, is a
diffuse swelling, hot and tender, dark red in color, show-ing over a portion

COLEMAX ET AL.— METABOLISM IX ERYSIPELAS

571

of its pcriphen,-, particularly in the scalp, a distinct raised edge. The nose
shows no abrasion. However, there is an occasional slight nasal heitiorrhage.
The throat is red but not swollen. The lymph glands of the neck are not
enlarged. Over both lungs are a few scattered rales. Coughing is frequent.
The spleen is not palpable.

OCT. 24

\^

26 1 27

26 29

30

31 I

/

^^

VA^

"\

^

jV ^y K

Fig. 3. — Odysseus B. (Case 3) Temperature chart.

Laboratory Examination. — Urine shows a faint trace of albumin ; no casts.
Blood : Leukocytes, 22,000 : polymorphonuclears, 84 per cent.

October 27 he was in the calorimeter from 10 :30 a. m. to 1 :30 p. m. By the
twenty-eighth the area of inflammation had increased in extent to cover the
scalp to the occipital prominence and the neck for two inches posterior and
inferior to the left ear. The inflammation, however, had decreased in intensity,
desquamation had begun and recovery from that time was rapid. He was
discharged, cured, October 31.

Case 4. — Facial erysipelas.

History. — Robert H., a waiter, born in liermany, ,38 years of age, was
admitted Oct. 31, 1916, discharged as cured Xov. 10, 1916. He has been in this
country since 1902. He drinks moderately of beer; no whisky. He has had
gonorrhea several times; had a chancre in 1915, no secondaries.

October 27 he had some fever and muscular pains but no chill. He remained
in bed until October 30, when he went to a choral club rehearsal. It was
first noted there that his lace was red and swollen. The swelling began on
the nose and spread rapidly to both cheeks.

OCT. 31 NOV I

TEN

P.F.*

A/^ A

(

/

A

V

V

\l

/

V

V

U

nr."

V-

/\

—

Fig. 4. — Robert H. (Case 4) Temperature chart.

Physical £.raMiiH<i(io>i.— Patient is a well developed and well nourished
man, very nervous, acutely ill. The tongue is slightly dry and covered with
a brown coat. Over the nose, the lower half of the forehead and the cheeks
down to the angles of the mouth there i.s an edematous, red, hot, slightly
tender area of inflammation, the edges of which are moderately raised and
indurated and sharply demarcated from the surrounding skin. The eyelids are
swollen and shut. The conjunctivae arc red and edematous. At the angle of

572 ARCHIVES OF IXTERXAL MEDICI XE

the left jaw there is a moderately tender, nonfluctuating lymph node. The
throat is slightly red. The lungs shows no signs although the patient coughs
frequently. The spleen is easily palpable two finger breadths below the
costal margin.

Laboratory Examinatioyi. — Urine shows a trace of albumin; no casts. Blood:
Wassermann is negative.

November 3 the inflammation had spread to involve the whole face. From
11:22 a. m. to 4;22 p. m. he was in the calorimeter. November 4 the ears
were involved but the severity of the inflammation of the face w^s less marked.
Desquamation had begun. During the night, from 11:30 p. m. November 4 to
2 a. m. November S, he was in the calorimeter. On the sixth a large grayish
patch appeared on the uvula. A culture was. negative for diphtheria. By the
eighth the redness and swelling had disappeared from the face. A new area
of inflammation had appeared on the back of the neck similar in character to
the first one. Lymph nodes were no longer palpable. He was in the calorimeter
from 11 :30 a. m. to 2 :30 p. m. By the tenth practically all signs of inflammation
of the skin and the grayish patch on the uvula had disappeared. The spleen
was still palpable below the costal margin. The heart murmur was still present.
He was discharged as cured on November 13.

Case 5. — Facial erysipelas.

History. — Joseph S.. a sailor born in Russia, 25 years of age. was admitted
Feb. 27. 1917, and discharged cured March 23. 1917. He had smallpox in
Russia in 1902. He drinks moderately. He denies venereal infection.

During a boxing match February 25 he received a severe blow over his left
eye and left ear. Two days later he was admitted to Bellevue Hospital with a
hematoma and a marked cellulitis of the left eyelid and ear. By March 3 the
wound of the eyelid had become definitely erysipelatous.

R

A^r

Fig. 5. — Joseph S (Case 5) Temperature chart.

Physical Examination March 6 (after development of erysipelas). — Patient
is a muscular, well nourished, very surly young Russian, prostrated by
disease: very toxic. The conjunctivae of both eyes are swollen, bright red and
are exuding pus. There is a slight cloudiness of the right cornea. In the left
upper eyelid there is a badly infected cut, extending the width of the lid and
exuding pus. The erysipelatous area extends over the nose, both cheeks to
the angle of the mouth on the right, to the neck on the left. The skin is red
and indurated and there are numerous blebs containing purulent serum.
In the upper part of the pinna of the left car is a hematoma the size of a
walnut. There is a wide sinus extending deep into ihe mass and exuding pus.
There is no sign of rupture of the drum membrane, no evidence of infection
of middle ear. The cervical nodes are enlarged and on each side is a diffuse,
tender swelling about the region of the parotids. The tongue is dry and coated
with brownish pus. Fauces are red and slightly swollen. The heart shows
marked overaction. Lungs show a few scattered rales. The soft edge of the
spleen is palpable one finger breath below the costal margin.

Laboratory Examination. — The urine is negative. Blood (March 6, 1917) :
Leukocytes, 7,000; polymorphonuclears, 75 per cent. Roentgen ray showed no
fracture of skull.

COLEMAX ET AL.— METABOLISM l.\ ERYSIPELAS 573

March 6 he was in the calorimeter from 12 noon to 1 p. m. March 12 the
cut on the eyelid was practically healed. The inflammation of the conjunctivae
and the diffuse redness and swelling of the skin had almost disappeared. There
was, however, a localized swelling over the upper right cheek and the bridge of
the nose. When this was opened it discharged considerable amounts of pus.
Following this he recovered rapidly and was dicharged as cured March 23.

The data of the calorimeter experiments are presented in Table 1.
A summary of results will be found in Table 2.

DISCUSSION OF RESULTS

Basal Metabolism- — In the ten experiments the agreement between
the total calories measured by direct and indirect calorimetry is decep-
tive. The calories by the direct method totaled 2,152.9 by the indirect
method 2,118.6, a divergence of only 1.1 per cent. In the individual
experiments, however, the divergence ranged between -f- 9.6 and — 13.5
per cent. Perhaps this discrepancy may be explained, in part, by the
necessarily short periods during which some of the patients were
observed. In typhoid fever,^ it was found that during the first hour,
some heat was probably lost in warming the bed frame and bedding.
This, however, cannot explain the discrepancy on the plus side and
can answer only for a part of the minus divergence. We must look
elsewhere for complete explanation. In the work on malaria,^" it
was demonstrated that the rectal temperature is a rather inaccurate
measure of general body conditions. It may rise more rapidly or less
rapidly than the average body temperature during sudden changes
in heat elimination and production. In the calculation of the heat
production by direct calorimetry the rectal temperature is assumed to
represent accurately the temperature conditions of the whole body.
This assumption is not valid during rapid fluctuations of temperature,
but probably holds where the temperature rises and falls gradually as
in the cases of erysipelas studied. In Figmcs 11 and 12 the curves
of both rectal and average body temperature are charted. It will be
seen that they are practically parallel.

One of the patients observed on the first day of normal temperature
after the subsidence of fever exhibited a basal metabolism 8 per cent,
below the average normal level. The other afebrile patient, also
observed on the day following the crisis, still showed a heat produc-
tion 12 per cent, above the normal. During the course of the fever,
the metabolism was always high, the variations being between 19 and
42 per cent, above the average normal basal.

Relation of Basal Metbolism to Temperature. — In typhoid fever and
tuberculosis '' the increase in heat production was found to be roughly

10. Barr, D. P.. and Du Bois, E. F. : Clinical Calorimetry. Paper 28, The
Metabolism in Malarial Fever. Arch. Int. Med. 21:627 (May) 191^.

11. McCann, W. S.. and Barr. D. P.; Clinical Calorimetrv. Paper 29, The
Metabolism of Tuberculosis, Arch. Int. Med. 26:66,3 (Nov.) 1920.

TABLE 1. — Calorimetric —

Subject,

Date,

Weight.

Surlace

Area, Linear

Formula

Period

End

of

Period,

Time

Carbon
Dioxid,
Gm.

""'sr-

E. Q.

Water,
Gm.

Urine N

H-'o^^r.
Gm.

Indirect
Calo-

rimetry,
Cal.

Heat
Elimi-
nated,
Cal.

.\TshelA

10/13/16
56.4 Kg.
1.58 Sq M.

ArshelA

10/15/16
55.6 Kg.
1.57 Sq. M.

James W

Prelim.

2
Aver
Prelim.
1
2
3
4
5

Aver.

Prelim.

1

Aver.
Prelira.

2

Aver.
Prelim.

1
Prelim.

2

3

4

5

6
Aver.
Prelim.

Prelim,

Aver.
Prelim.

1
Prelim.

2
Aver.

11:35
12:35
1:35

6:47

7:47
8:47
9:17
10:47
12:47
2:47

11:32
12:32
1:32

11:07
12:07
1:08

11:24

12:24

11:22

12:52
1:22
2:.3
3:22
4:22

12:11
1:11
11:42
12:42
1:42

11:52
12:52
11:37
12:37
1:37

23.5
24.7

25.2
24.0
11.5
36.3
45.6
43.5

21.3
20.4

27.0

28.9

28.3

30.3,

16.9 J

13.4

27.6

29.7

31.1

29.1
....
24.0
24.1

32.1

28.4
29.7

24.3

23.5
24.0
10.8
33.1

43.6

17.9
13.3

23.9
27.5

44.1

14.6
26.5
27.8
29.3

28.6
22.0

29.9

27.2
25.7

0.74
0.74

0.78
0.73

0.75
0.75
0.73

0.87
0.81

0.76
0.76

0.75

0.78

0.75
0.76
0.78
0.77

0.74

0.79
0.77

n.78

0.76
0.84

....

34.4
37.1

....

33.8
16.1
50.2
61.7
61.6

32.8
33.0

42.6
39.0

43.5
....

|38.4
1 19.4
13.9
36.0
39.8
41.5

57.2

0.60
0.60

0.49
0.49
0.49
0.49
0.49
0.49

0.43
0.45

0.66
0.66

0.65

0.«,J

0.60 1

0.60

060

0.60

^.60

0.73

77.3

z

74.2

59.7
60.9

85.2
90.3

89.6

145.7

49.0
86.8
91.6
96.5-
93.9

93.9

73.2
75.0

99.3
86.9

76.4
79.6

83.5
81.5
38.6
12i;.0
164.0
163.9

10/17/16
54.8 Kg.
1.67 Sq. M.

Odysseo. B

IO/27/1B

Robert H

11 ! 16
fiO.O Kg.
I;67Sq. M.

Robert H

11 3 16
:.:..:. Kg.
1.67 Sq. M.

Robert H

11/.5/16
60.4 Kg.
1.B7 Sq. M.

C4.4
63.0

83.4
8.3.5

93.5

r80.7

{
147.7

43.3

81.3

90.9

93.1

lOO.O

11/8/16
.=i9.0 Kg.
1.67 Sq. M.

.Josephs

3/6/17
69.0 Kg.
1.81 Sq. M.

Josephs

3/9/17
67.2 Kg.
1.81 Sq. M.

43.1 ' 0.39
37.9 0.39

....

40.9 0.33

43.2 1.05

39.3 1.05

78.2
75.3

98.2

88.8
87.3

Direct
Calo-

R«tal

Aver-

Work
Adder,
Cm.

Xoo-

Per Cent.
Calories from

Calories
per Hour

RcinarliS

Pro-

Fat

Carbo-
hyd.

Per
Kg.

Per
Sq. M.
(Linear)

39.3

••

..

....

Basal

-9.7

39.4

100

....

....

Very quiet

84.3

40.0

96

0.72

21

74

S

1.37

48.9

Very quiet

FaUing temperature

75.4

39.9

90

....

••

Very quiet, voided t

78.0

39.8

Very quiet

37.4

39.8

....

Verj- quiet

lOT.b

39.4

85

....

Very quiet

148.6

39.1

Very quiet

131.2

38.9

80

....

....

Very quiet

..

0.73

17

75

8

1.33

47.3

36.4

62

....

Basal

«8.6

36.5

58

0.88

20

35

45

1.10

Slightly restless

64.6

36.5
38.7

58

1).81

20

51

29

1.11

36.6

Fairly quiet
Uasal

86.5

38.8

67

0.74

70

»

1.57

Fairly quiet

90.3

38.9
39.5

67
106

62(?)

0.75

68

12

1.65

54.6

Restless, coughing
Basal

98.0

39.8
39.7

102
91

0.74

72

9

1.49

53.7

Somewhat restless
Rising temperature

103.1
51.4

40.2
40.3

91
105

:i

.:

l.«3

f Fairly quiet, shivr
i 12:10-1:10 p.m.
ISlightly restless

46.0

40.3

18

0.74

1.64

Slightly restless

81.0

40.5

101

10

0.75

1.45

Quiet, sleeping

91.8

40.6

115

23

0.T7

1.53

Fairly quiet

87.3

40.5

119

35

0.77

66

17

1.62

56.2

Slightly restless
Basal

.,.,1

1'..-.

33

0.72

73

5

1.55

5fi.O

Restless

37.3

....

Basal

75.9

37.2

88

26

0.79

67

19

1.24

Slightly restless

7-...,

37.2

78

''

0.77

68

18

1.27

44.3

Fairly quiet

40.1

....

Basal

100.7

40.3
39.0

91

30

0.78

«8

23

1.44

54.8

Rather restless
Basal

91.1
89.0

39.0
89.1

82

16
11

Fairly quiet dozec
Qui'eT"'

....

0.79

32

48

;«

1.29

48.0

Ife^^lg^ll

CX)

a

=>

oc

^

s

ss

g

?i

S

S3

g

s

a

So ■■^■Bsr

.lillll

S3

2

00

S

ss

a

„

2

s

s

-m^-B

1

+

"*■

+

11^.^1

Ili^«l

5

"

"

S

s

s

S

5

s

'

S|«|s|

M

<s

X

m

tc

-;•

«

M

i I. g Si; a

+

+

-

1

+

1 s-|

i

o

=«

wj5

s

i

s

s

s

s

s

2

§

S

lip

c

=

Id

IN

o

3

i

S

1

s

s

s

S

s

g

h

.

s

3

^

e

f:

s

s

s

o

i

»•

2

"

1^

"

.

o

Ill

n

S8

?

^

n

A

S

f

§

s

I

§

S

g

?s

"

s

s

g

s

.-^

-•2

s

.-S

Tl-S

JJ

.£

.£

■fi

It

?l

11

II

1

il

ll

tl

■s

ss

5^

:¥£

rS2

s

>'

^1

s

e|

Ak

•s"

^1

^1

^

s

g^

S|

h-

iS

s

6

g

s

s

s

5

ffl

a

i<s

'^

-^

s

»

«=

s

2

s

S

^

^

IS

™

s

r?

6

«

^r

00

<o

s

2

2

2

2

-

=

=

s

-r

^

fi

«
1

6

w

W

n

u

.

ai

to

•5
<

1

<

1

1

1

1

1

1

1

COLEMAX ET AL.-METABOLISM I.\ ERYSIPELAS 577

proportional to the degree rise in rectal temperature. The same relation
is found in erysipelas. Figure 6 expresses this relationship graphically
according to the method used by McCann and Barr in tuberculosis. The
abscissas show the level of metabolism in percentage of the average
normal, the ordinates show the rectal temperature in degrees Centi-
grade. The line 90 means 10 per cent, below the average normal;

t 1

X

1 •

' /

/

38
37'

y

/

.

y

C(

YSIP

LAS

■/

130 140 150

Fig. 6.— Relationship of basal metabolism to temperature in erysipelas.
Ordinates represent rectal temperature in degrees Centrigrade; abscissae the
metabolism expressed in percentages of the average normal. Each dot repre-
sents an experimental period in the calorimeter.

150 tneans 50 per cent, above the average. Each dot represents a
calorimetric experiment. The diagonal line represents the average.
Figure 7 expresses the same relation for the more numerous observa-
tions on typhoid fever. The heat production in typhoid increases a little
more rapidly for each degree rise in temperature but on the whole the
curves are strikingly similar.

^

1

! ] J

^

[ '■:y-\.-

y- .™.,.,.w

V

■'\

1 1

90 100^ no 120 130 140 150

Fig. 7.— Relationship of basal metabolism to temperature in typhoid fever.

Nitrogen Equilibrium Kind Weight. — The observations on the nitro-
gen metabolism are disappointing and inconclusive. In spite of the
determined efforts of a well trained staff of nurses, it was impossible
to induce the patients to take nourishment sufficient for their caloric
need during the fever. In figures 8, 9, and 10, the temperature, weight.

578 ARCHirES OF IXTERXAL MEDICIXE

nitrogen intake and output together with the total food intake are
represented graphically. The dashes of the dot dash line at the foot of
the charts represent the heat production as calculated from actual
observations in the calorimeter. The clinical data from which the charts
are drawn will be found in Table 3. It will be seen that during the
febrile course the expenditure of energy was always in excess of the
caloric intake. No conclusions, therefore, may be drawn concerning
toxic destruction of the body protein. Even normal individuals may
show a nitrogen loss during an insufificient caloric intake.

OCT. „^ NOV.

13-14 15-16 17-18 19-20 2122 23 OCT.

|.jg. 8.— .\rshcl .\. (Case 1) Temperature, body weight; food nitrogen
(lotted line; excreta nitrogen, continuous line. .\t the base of chart, columns
representing total calories of food. The dot dash line represents the estimated
heat production for twenty-four hours. The dashes are placed on days of the
observations in the calorimeter.

Fig. 9.— James W. (Case 2) Temperature, body weight, food and excreta
in nitrogen, food calories and dot dash line showing estimated heat production.

Fig. 10.— Robert 11. (Case 4) Temperature, body weight, food and excreta
nitrogen, food calorics and dot dash line showing estimated heat production.

COLEMAX ET AL.~METABOLISM IX ERYSIPELAS 579
TABLE 3. — Clixical Calori metric Data ix Four Cases of Erysipelas

Food

Xarae and

Body

Food

Urine

Exca-eta

Xitrogen

Drine

Date

Weight

Total

Carbo-

S.

S,

Balance,

Volume.

Calo-

hyd..

Fat,

Ale,

Gm.

Gm.

Gin.

Gm.

Co.

ries

Gm.

Gm.

Gm.

James W.

10. 13 ii;

1.598

174

69

9.5

19.4

203

—10.8

1.730

lO/U/lB

ih'.8

2,195

278

84

10.6

12.2

13.3

— 2.7

1..500

1015/lfi

55.5

2,363

273

100

12.0

16.5

17.7

— 3.7

10 Ifi If.

236

71

8.0

13.1

13.9

— 5.9

i;355

10/17/1«

55:i

aieis

301

116

12.7

12.4

13.7

— 1.0

1.800

10 18,16

56.2

2.741

317

114

14.7

13.3

14.8

— 0.1

1.980

10/19/16

56.6

2,573

271

115

15.3

13.5

130

+ 0.3

2.3UO

10/20/16

36.6

2,378

280

92

14.5

13.2

14.6

l.SiU

Arshel A.

10/13 If.

56.4

346

57t

10

0.9

13.9

14 0

—13.1

1.213

10 i4n«

.5.5.9

972

144

31

3.6

12.7

13.1

— 9.3

1,210

10/ir. 1*1

56.0

1,049

100

55

5.0

12.6

12.1

i:070'

lo/iti/io

1.630

218

58

7.2

13.4

14.1

— 6!9

2,360*

lO/lT/lfi

jsie

1.958

228

87

8.5

12.9

13.8

— 5.3

2,480

10/18,16

55.5

2.576

291

lis

12.4

11.4

12.6

— 0.2

1,220

10/19/16

56.0

2,303 ?

239?

106 ?

13.4 ?

11.5

12.8

-1- 1.6

2:200

1020/16

56.0

2,333

252

104

13.0

12.4

13.7

2,370

10/21/16

2,656

236

144

13.7

12.4

13.7

±o;o

2,800

10/22/16

5.5.5

2.863

272

47

14.7

11.6

13.1

-r 1.6

2,200

10/23/16

56.4

2,835

282

44

13.5

11.2

12.6

+ 09

2.230

Odysseo B.

10/26/16

54.5

1.371

134

68

15.7

16 4

— 90

nso

10/27/16

53.9

933

108

38

54

3.2

14.3

14.3

—11.1

076

10/28/16

54.1

2,475

187

127

162

14.4

14.9

16.3

8;o

10/29/16

54.0

2,493

239

117

54

Mi

12.4

13 6

510

10/30/16

54.4

2:493

270

106

14.2

15.6

....

1,060

-10/31/16

55.4

2,116

192

105

13.7

13.B

13.0 -

—'1.3

1.180

Robert H.

11/ 1/16

60.0

1,429

137

69

8.4

18.0

19.8

—11.4

2.125

11/ 2/16

60.9

1,896

176

96

10.6

19.5

— 8.9

1.930

11/ 3/16

59.8

1,109

105

54

6.5

17.1

18.8

—12.3

1.555

11/ 4/16

60.1

1,537

174

65

8.3

16.0

17.6

— 83

1.390

11/ 6/16

60.3

2.em

•282

119

15.1

18.1

19.9

-4.8

2:980

11/ 6/16

59.9

2,315

257

115

15.1

15.6

17.2

— 2.1

1.120

11/ 7/16

60.3

2,339

241

106 1 ...

14.0

15.2

16.7

— 2.7

2.570

11/ 8/16

59.1

1,861

216

80 ...

8.8

10.7

11.8

— 3.0

1.477

11/ 9/16

59.8

2,n5

292

121

17.0

14.2

15.6

+ 1.4

1,230

11 10 '16

59.5

2,496

258

112

...

15.2

13.6

15.0

-1-0.2

1,U0

5f lood nitrogen.

The weight curves are of some interest. In studying pneumonia.
Sandelowsky '^ found that, during the acute febrile course of the
disease, many patients maintained their weight 01 even gained a consid-
erable amount. At the crisis and during the early convalescence, on
the other hand, there was a rapid loss in weight. This, he attributed
to a storage of water in the body during fever and a rapid loss of water
following it. He supported this hypothesis with refractometric studies
of the serum proteins by which he demonstrated to his own satisfaction
that there was a dilution of blood during the fever and a concentration
or return to normal following it. In the light of Sandelowsky's
contention, the weight curves in erysipelas are significant. During the
acute course of the fever, it is a little surprising that the weight remains
practically constant in spite of the insufficient food. At the crisis,

jrizfiitratiiui liei Pneumonic, Deutscli. Arch. f.

12. Sandelowsky. J.
klin. Med. 96:445, 1909.

580

ARCHIVES OF LXTERXAL MEDICJXE

and in convalescence, however, the fall of weight noted by Sandel-
owsky in pneumonia is not observed.

Regulation of Body Temperature. — In the chills of malarial fever,®
it was found that the sharply rising temperature was accomplished by
a great increase in the production of heat while the elimination of heat
remained at the level which had existed before the chill. Following
the paroxysm there was a period of high continuous temperature.
In this stage, the heat production fell while the heat elimination
increased until the two were equal to each other. Both, however,

gms.perIhour.
30

Fig. 11.— Metabolism chart. Robert H (Cai
of heat production to heat elimination.

4) curves showing relationship

were maintained at a level high above the normal basal metabolism.
The conditions under which the erysipelas patients were studied may
be considered analogous to the high constant temperature following
the malarial paroxysm. Both heat production and elimination are at
a high level, which, as we have seen, varies with the degree of fever.
In erysipelas, however, the temperature is seldom truly continuous.
There are usually frequent remissions, sometimes intermissions. It
would be interesting to know how these fluctuations are brought about
and whether the mechanism of the rise and fall of temperature during

COLEMAX ET AL.— METABOLISM IX ERYSIPELAS 581

a high fever is similar to that of malaria in which the temperature
is originally normal. It was with these questions in mind that the
study of erysipelas was undertaken. Unfortunately our data are not
sufficient to answer them. Only two long observations were made
during significant changes in temperature. In both, the changes were
so gradual that it is impossible to draw conclusions from them. Figure
11 represents graphically the results of studies made on Robert H.
during a rising temperature while Figure 12 shows in a similar manner
the results of studies made on Arshel A. during a falling temperature.
It is interesting to note that the rather sharp rise during the first two

40

RECTAl

MP

..^^^

39°
38°
100

AVSRR&E

BOD

^^^

^^^

HEAT CLIM.

PIR^T CAL,.,..

50

---■^

INDIRECT CAL.

VAPORIZATION

-Al..

■0

Fig. 12.— Metabolism chart. Arshel A. (Case 1) curves showing relation-
ship of heat prouction to heat elimination.

hours of the observation on Robert H. was accompanied by shivering
which amounted to a moderate chill. During these periods, the heat
production was high. After the shivering ceased, the production of
heat diminished while the heat elimination increased in a manner quite
comparable with that observed in malaria. During the falling temper-
ature shown in Figure 12, the heat elimination remained at a practically
constant level throughout the seven and a half hours of the observa-
tion. The heat production gradually fell.

He<it Loss by Vaporisation of Water. — The question of water
utilization in fever is of great importance. Because of its ability to

582 ARCHIVES OF IXTERXAL MEDICIXE

absorb heat and its fluid character water otters the chief means of
carrying heat from the cells where it is produced to the surface of the
body where it may be eliminated. Heat loss both through radiation
and conduction and through vaporization is therefore dependent on the
proper mobilization of water within the body. Because of the dry
skin, concentrated urine and other phenomena of fever, it has been
argued that the water supply of the body is depleted or cannot be
mobilized. This has been considered a possible cause of fever by
Balcar, Sansum and \\'oodyatt.'' Facts concerning the heat lost in
vaporization of water from skin and lungs are relevant to this question.
Under the constant temperature conditions of the calorimeter,
normal subjects accomplish about 24 per cent, of their total heat
elimination by vaporization of water. The average percentage was
slightly lower, about 22 per cent, in typhoid fever. In erysipelas,
on the other hand, the heat lost in the vaporization of water is high
in proportion to the total elimination. During the rising temperature
observed in Robert H. (Fig. 11), it constituted 26 per cent. The limits
of variation for all the observations were from 23.6 to 33.4 per cent.

SUMM.ARV .-\XD COXCLU.SION.S

1. Ten calorinietric experiments have been made in five cases of
erysipelas. Two of these were taken on the day following the crisis.
The others were made during the febrile course of the disease.

2. During the fever, the metabolism is increased from 19 to 42
per cent, above the average normal basal. The increase in metabolism
is roughly proportional to the degree of fever. A temperature of 40
C. involves a heat production of about 40 per cent, above the average
normal.

3. The change in rectal temperature is not always an accurate index
of the change in average body temperature in erysipelas

4. The regulation of body temperature is similar to that observed
in malaria during the stage of high continuous temperature. Both
heat jiroduction and heat elimination are maintained at a high level.

5. The heat lost in the vaporization of water constitutes from
23.6 to 33.4 per cent, of the total heat elimination. During rising,
con.stant and falling teinperatures the j)ercentge of heat eliminated in
the vaporization of water was greater than in normal individuals.

6. No specific differences were found between the metabolism in
erysipelas and in typhoid fever. Both fevers show approximately the
same increases in the level of heat production for the .same increase
in body temperature. The ])rotein metabolism is greatly increased in
both diseases.

\X Balcar. J. B. : Sansum, W. D.. and \V.»,<lvatt. R. F. : Fever and Water
Reserve of tlie Body. .Arch. Int. .Med. 24: lU. (.Inly) 191').

CLINICAL CALORIMETRY. XXXI. OBSERVATIONS
OX THE METABOLISiM OF ARTHRITIS*

RUSSELL L. CECIL, M.D. ; DANID P. BARR. M.D.,

AND

EUGEXE F. DC BOLS. M.D.
With the Techmcal Assistance of G. F. Sodf.rstro.m axd Estelle Magill

.NEW YORK

In recent years, bacteriologists have made important contributions
to tlie etiology of arthritis, especially in regard to the part played by
focal infections. The large mass of evidence which has gradually
accumulated indicates that most cases of arthritis are infectious in
origin. In spite of this, however, some practitioners have held to the
belief that arthritis, particularly in its chronic form, is an expression
of a disturbance in metabolism. The relationship between gouty
arthritis and an abnormal purin metabolism gave some ground for
this theory, and for many years the various forms of acute and chronic
arthritis were supposed to be rlependent on the retention in the body
of uric acid or some kindred substance. For this reason rheumatic
patients were often put on a reduced nitrogenous diet, but the results
obtained by this mode of treatment were not encouraging. Pemberton '
has advocated a lowered carbohydrate intake foi arthritis patients.

Comparatively little experimental work was done on the metabolism
of arthritis before the extensive studies of Pemberton and his asso-
ciates - published in 1920. Indeed, it is only within the last few years
that the development of newer and more accurate methods in biochem-
istry have made reliable investigations along this line possible. Tileston
and Comfort.' who studied the non])rotein nitregen and urea in the

* From the Russell Sage Institute of Pathology in affiliation with tlie Second
Medical Division of Bellevue Hospital.

1. Pemberton, R, : The Metabolism and Treatment c,f Rheumatoid .Arthritis,
Fourth Paper, Am, J. M, Sc, 153:678, 1917,

2. Pemberton, R., and Robertson, J. W. : Studies on .Arthritis in the .Army,
Based on Four Hundred Cases. I. Preamble and Statistical Analysis, Arch,
Int. Med. 25:231 (March) 1920. Pemberton, R„ and Tompkins, E. H, : Ibid. IL
Observations on the Basal Metabolism, .Arch, hit, Med. 25:241 (March) 1920.
Pemberton, R„ and Foster, G. L, : Ibid, III. Studies on the Nitrogen, Urea,
Carbon Dioxid Combining Power, Calcium. Total Fat and Cholesterol of the
Fasting Blood Renal Function, Blood Sugar and Sugar Tolerance, Arch. Int.
Med. 25:243 (March) 1920, Pemberton. R., and Buckman, T, E, : Ibid. IV,
Studies in the Relation of Creatin Metabolism to .Arthritis, Arch, Int, Med,
25:335 (April) 1920, Pemberton, R. : Ibid. V. Roentgen-Ray Evidences, Clin-
ical Considerations, Treatment, Summary, Conclusions and Clinical Abstracts
of Cases Studied, Arch, Int. Med, 25:351 (April) 1920.

3. Tileston. W„ and Comfort, C, W., Jr.: The Total Nonprotein Nitrogen
and the Urea of the Blood in Health and Disease, as Estimated by Folin's
Methods. Arch. Int. Med. 14:620 (Nov.) 1914.

584 ARCHIVES OF IXTERXAL MEDICI XE

blood in various diseases, obtained low figure^ in rheumatic fever.
Pemberton and Foster ^ estimated the blood urea and nonprotein nitro-
gen in seventeen cases of chronic arthritis and found that the figures
in every case were well within the normal limits. Pemberton and
Buckman^ made determinations of the nonprotein nitrogen in the
blood in forty cases of arthritis and found an abnormal elevation in
only two cases. Folin and Denis * found an increase of uric acid in
the blood in nongouty arthritis. Similar results were obtained by
McClure and Pratt.' Pemberton and Buckman ^ carried out observa-
tions on the creatin and creatinin of the blood and urine in forty cases
of arthritis. About one-half of the cases showed an abnormally high
\alue for blood creatinin. Pemberton and Foster ^ also estimated the
carbon dioxid combining power of the blood, the calcium of the circu-
lating blood and the total fat and cholesterol of the fasting blood. In
all cases the figures were well within normal limits. These authors
did find evidence of an abnormal rise in the blood sugar following the
ingestion of 100 gm. glucose.

Pemberton and Tompkins ^ studied the basal metabolism in a series
of twenty-nine cases of arthritis, the observations being made by
indirect calorimetry, using the Tissot apparatus. Of the cases studied,
80 per cent, showed a metabolism within normal limits, in 20 per cent,
the rate was slightly below normal limits. From the respiratory
quotients no abnormality could be detected in the percentage of calories
obtained from the three classes of foodstuffs.

The present investigation, carried out in 1916-1917, deals with the
metabolism in two cases of acute arthritis, two cases of subacute
arthritis, one case of gouty arthritis and four cases of chronic arthritis.
Circumstances prevented the study of a largei series but it seemed
desirable to publish the data at hand in order to complement the report
of Pemberton and Tompkins and to render the records accessible to
future workers.

METHODS

The experiments reported in this study were carried out on
patients in Bellevue Hospital. The calorimeter of the Russell Sage
Institute of Pathology was employed. The heat production was
measured by both the direct and indirect methods. The actual details
of technic have been described fully in previous articles in this series.*

The basal metabolism of four cases of acute and subacute arthritis
was determined. Edward C. (Case 1) liad a typical case of acute

4. Folin, O., and Denis, W. : The Diagnostic Value of Uric Acid Deter-
minations in Blood, Arch. Int. Med. 16:33 (July) 1915.

5. McChire, C. W., and Pratt, J. H.: A Study of Uric .Acid in Gout, Arch.
Int. Med. 20:481 (May) 1917.

6. Clinical Calorimetry, Arch. Int. Med. 15:793-945 (Mav) 1915; 17:855-1059
(June) 1916; 19:823-957 (Mav) 1917.

CECIL ET AL.— METABOLISM OF ARTHRITIS 585

rheumatic fever. John Bl. (Case 2) was considered to have subacute
rheumatic fever. John Br. (Case 3) had acute arthritis associated with
gonorrheal urethritis and a conjunctivitis of gonococcus origin. Joseph
]VIcC. (Case 4) had a subacute arthritis and an intermittent urethral
discharge of several years duration. Observations concerning the
nitrogen equilibrium were made on one patient, John Br. (Case 3).

REPORT OF C.\SES

Case 1. — Acute Rheumatic Fever.

Historx. — Edward C, a waiter, born in Ireland, 33 years of age, was
admitted, Jan. 13, 1917, and discharged unimproved, Jan. 29, 1917. Since child-
hood he has had frequent attacks of severe sore throat, and he has had quinsy
four times. In 1906, he had symptoms of scarlet fever and was treated at a
contagious hospital. During the course of the disease, he developed a very
severe joint attack which lasted two weeks. In July, 1912, he was operated
on for appendicitis, and in November, 1912, he w^as operated on for ulcer of the
stomach. He was kept on a diet for one year after the operation. He was a
convivial drinker, but denied venereal disease.

November 24 an operation was performed on his nose. Following this he
had a severe hemorrhage from the nose lasting four hours. November 29
he was seized with a severe quinsy w^hich was lanced December 1. He rapidly
recovered from the throat trouble, but December 20 he was seized with pain in
the muscles and joints which remained severe for about a week. Vague pains
were present until January 13 when they again became very severe.

JAN.I3 14,15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

f^^

VH^rMY

Fig.

-Edward C. (Case 1). Temperature Chart

Physical Examination. — The patient was a rather pallid, poorly nourished
young Irishman of cheerful disposition. The tonsils were much enlarged, the
right being larger than the left. The crypts are deep but there are no follicles
present. The teeth are in fair condition.

Heart: The left border of dulness in the fifth space is 11.5 cm. from the
median line: the right border is beneath the sternum. There is a definite,
short, soft systolic murmur, heard best at the apex, transmitted to the mid-axilla
and the sternum.

.Abdomen : Shows scars of previous operations. Rectal examination disclosed
a normal prostate and seminal vesicles. There was some pain, on motion, in
both shoulders, but no swelling or redness.

Laboratory Examination. — Urine: Moderate trace of albumin; no casts.
Blood pressure : Systolic, 140 mm. ; diastolic, 85 mm. Blood cultures : Negative.
Tonsil cultures: Streptococcus viridans (blood agar plates); from right tonsil.
.9. viridans and a moderate number of Staphylococcus aureus colonies and a
diphtheroid bacillus from the left tonsil. Wassermann test: Negative. Gono-
coccus fixation test: Negative.

January 22 he was in the calorimeter from 10:50 a. m. to 1 :50 p. m. He had
moderate pain in both shoulders, irf the right wrist and in both knees. None of
the joints showed swelling, tenderness or redness. This condition was practically
i.nclianged when he was discharged January 29.

586

ARCHIJ'ES OF IXTERXAL MEDICI XE

Case 2. — Subacute Rheumatic Fever.

History. — John Bl., a bookkeeper, born in Scotland, 58 years of age, was
admitted Jan. 22, 1917, and discharged cured Feb. 7, 1917. In 1914 he had an
attack of joint pain similar to the present one. He had pneumonia in 1912.
Thirty years ago he had gonorrhea without complications. He drinks one
glass of whiskey and three glasses of beer daily but is rarely intoxicated. His
father died of "rheumatism and heart trouble."

January 8 he began to have vague pains in the muscles. January 18, follow-
ing exposure to the weather, both knees became very painful and swollen. He
felt chilly during the day and was nauseated. Two days later his ankles began
to swell, and finally his arms and hands became stiff and sw-ollen.

Physical Examination. — The patient was a large, robust Scotchman, not very
ill. The tongue showed a moderate white coat. The tonsils were atrophic but
red with chronic congestion. Most of the teeth were missing, although there
was no definite pyorrhea. Heart : Not affected. Lungs : Show signs of a
moderate emphysema. At the right base are many coarse, leathery friction
rubs and diminished breathing. There was swelling and exquisite tenderness
of the right and left knee joints; wrists and. to a lesser extent, of the ankles.
On the right shin is the scar of an old varicose ulcer.

JAN 22

23

24

25

26

27

28

29

TEM
102

100°
9R

RF.'

A

'"

\

V

\^

^ /

u

^

^

Fig. 2. — John Bl. (Case 2). Temperature Chart.

Laboratory Exaniiiiation. — Urine: Negative. Blood pressure: Systolic. !30
mm. ; diastolic, 90 mm. Blood culture : Negative. Blood : Leukocytes, 12,000 ;
polymorphonuclears, 80 per cent. ; hemoglobin, 80 per cent. Wasserniann test :
Negative. Gonococcus fixation test : Negative. Culture . of each tonsil shows
abundant and almost pure growth of Streptococcus viridans and a few colonies
of Staphylococcus aureus.

January 26, from 10:45 a. m. to 1:45 p. m., he was in the calorimeter. At
that time the hands, wrists and knees were moderately hot and swollen. The
elbows and shoulders were tender and painful on motion. The joints improved
gradually and he was discharged as cured February 7.

Cask 3. — Acute Gonococcus Arthritis.

History. — John Br., a clerk, born in the United States, 19 years of age, was
admitted Jan. 25, 1917. and discharged improved March 23. Both his mother
and his father had rheumatism. The mother died in 1916 of chronic alcoholism.
He had chorea in early childhood. He says that between the ages of 5 and
13 he averaged one attack of "rheumatism" a year, but has had none since.
The attacks were usually mild and did not force him to go to bed. He never
had tonsillitis.

January 7, seven days after intercourse, he began to have a urethral dis-
charge. January 13 he had severe pains in the right knee. Later, the right
foot, left foot and right thumb were affected. Pains became very severe and
motion was limited.

CECIL ET AL.— METABOLISM OF ARTHRITIS

587

Physical Examination. — The patient was a well developed, somewhat emaci-
ated boy, acutely ill, rather toxic. The tongtie was slightly dry with a brown
coat. The eyes showed a moderately severe conjunctivitis. The tonsils were
small with deep crypts and no signs of inflammation. The heart was not
involved. The left knee joint was distended with fluid, hot, slightly tender
and held in semiflexion. Considerable motion was possible without great pain.
The right knee showed signs of inflammation and contained some fluid. Both
ankles were swollen, the left more than the right. There was marked tenderness

JAN. FEB.

31 I 2 3 4 5 6-7 8-9 10-11 12-13 1415 16-17 18-19 20-21 22-23 2426 2fr? 7 28

mfi"^^^

l±

^\7'

A77

w(;t k

62^

vllTIOGEfl^
"r'cOD NIT Rod EN.

Fig. 3. — John Hr. ( Ca.sc 3). .\cutc arthritis, k'^i'iococcus origin. Tempera-
ture, body weight ; excreta nitrogen, continuous line ; food nitrogen, dotted
line. At base of figure, columns representing total calories of food. The dot
dash line represents the estimated heat production in calories for twenty-four
hours. The dashes are placed on days of the observations in the calorimeter.
Note that the calories of the food exceed the estimated heat production except
for the first two days of observation.

over the tarsometatarsal joints of both feet; tenderness and pain on motion in
the metatarsophalangeal joint of the great toe and of the metacarpophalangeal
joint of the thumb. There was a moderate, white, purulent urethral discharge
containing gonococci in large numbers. The prostate was not enlarged. The
urine from the posterior urethra was clear and contained only an occasional
pus cell. The urine from the anterior urethra contained many pus cells but no
albumin or casts.

588 ARCHirES OF IXTERXAL MEDICI XE

Eaboratory Examination. — Blood; February 2 (before vaccine), leukocytes
11,000; polymorphonuclears, 75 per cent. Blood culture: January 30 and
February 3, sterile. Wassermann negative. Gonococcus fixation test :
January 26, doubtful; January 31, negative. Tonsil culture. February 3: Both
tonsils showed Streptococcus viridans predominant but with scattering colonies
of Staphylococcus aureus and Micrococcus paratetragenus.

January 29 he was in the calorimeter from 11 a. m. to 3 p. m. ; January 31,
from 10:30 a. m. to 2:10 p. m. February 2 a faint presystolic murmur was heard
for the first time, just medial to the apex of the heart. On the same day, at
10 :50 a. m. he was given intravenously a 25 million dose of the New York
Board of Health typhoid vaccine and was observed in the calorimeter from
11 a. m. to 3:15 p. m. February 4 he received a second dose of typhoid vaccine.
February 5, 10 c.c. of cloudy fluid w-as removed from the left knee joint. The
smear showed large numbers of polymorphonuclear cells and a few endothelial
cells. Occasional gram-negative intracellular bodies, resembling gonococci, were
seen. The cultures were negative on ascitic glucose agar plate, deep ascitic
glucose agar tube, blood broth, ascitic and plain broth. February 9. from 11 a. m.
to 2:15 p. m., he was in the calorimeter. February 14 he received a third injec-
tion of typhoid vaccine. He remained in the metabolism ward until March 1
when he was discharged to the general service.

January 29 he weighed 137 pounds ; February 9, 126 pounds; March 1, 131
pounds. During the time of observation none of the joints originally affected
were cured. There was frequently marked improvement with equally frequent
relapse, sometimes following the administration of typhoid vaccine by twenty-
four hours, sometimes by thirty-six hours, more frequently occurring spon-
taneously without reference to treatment. During the last three weeks, however,
there was a very slow, general tendency to improvement. The heart murmur,
which was first heard February 2, became much louder by February 9. March 1
it was again very feeble and March 5 it could be heard only after exercise.
During the month the right side of the heart increased moderately in size.
The electrocardiogram taken March 2 showed well marked right sided enlarge-
ment. The urethral discharge lessened with local treatment but never dis-
appeared entirely. It remained localized to the anterior urethra.

In the general ward, from March 3 to 13, he received gonococcus vaccine
intravenously in doses of from 20 to 30 million. His improvement was remark-
able. The temperature, which had been continually elevated since entrance to the
hospital, became normal and remained so except during the reactions following
vaccinations. The pain disappeared. \ moderate amount of fluid in the knee
joints, however, remained unabsorbcd. He received bakes and massage until
March 23, when he was discharged. Both knee joints contained small amounts
of fluid but there were no other symptoms. He was seen again in .\pril. His
right knee still contained a small amount of fluid Init he was otherwise i:i
excellent health.

Case 4. — Subacute Gonorrheal .Arthritis.

History. — Joseph McC, an elevator operator, born in the United States. 27
years of age, was admitted Jan. 14, 1917, and discharged unimproved March 1.
1917. In 1910 he had an attack of gonorrheal urethritis followed by epididymitis.
Since then he has had a urethral discharge several times, the last in November.
1916. In 1912 he had an acute arthritis involving all the joints, and he was ill
for nine weeks. A second attack of the same character, in 1914, lasted two
months. He has never had sore throat, chorea or other manifestations of acute
rheumatic fever.

Jan. 3, 1917, he began to have a dull pain in the lower end of the spine
and in the lumbosacral muscles, which radiated down the right thigh. His
right heel became so painful that he could not walk, .\bout tlie same time the
urethral discharge recommenced.

CECIL ET AL.— METABOLISM OF ARTHRITIS 589

Pliysical Examination. — This shows an undeveloped, poorly nourished man.
He held himself very stiff because of pain in the back. There was pain on
pressure over the sacrococcygeal joint and marked tenderness over the spine
of the ninth dorsal vertebra. The under surface of the right heel was exquisitely
tender. There was a considerable watery, purulent urethral discharge, a smear
of which showed gonococci. The prostate was moderately enlarged. The right
seminal vesicle was palpable.

Laboratory Examination. — Urine from both anterior and posterior urethra
contained pus but no albumin or casts. Gonococcus fixation test was negative.
Wassermann reaction was negative. The roentgenogram showed periosteal
exostosis of the os calcis of both heels.

His symptoms remained unchanged during the first seven days in the
hospital. February 1 he was given intravenously 40 million of the New York
City Board of Health typhoid vaccine. February 3, 5 and 7 similar doses were
given. The character of the reactions following the injections appears on the
temperature chart. The chills began from thirty to seventy-five minutes after
injection. February 13 a dose of 50 million was given.

February 7 and 13 the patient was observed in the calorimeter. No noticeable
improvement resulted from the vaccine therapy. On the first day following
each injection he was more stiff and uncomfortable. On the second day the
condition returned to that which obtained before the vaccine was given. Except
at the times when he was given vaccine this patient had a normal temperature.

Results of the study of basal metabolism in the cases of acute and
subacute arthritis will be found in the summary, Table 4, the detailed
calorimeter data in Table 1.

It will be noted that the patients studied showed but slight eleva-
tion in the basal metabolism as measured by the surface area standards
of Aub and Du Bois. All but one determination came within 12 per
cent, of the average normal figure. This single exception was found
in the case of John Br. who showed a basal metabolism 26 per cent,
above the average normal when he was slightly restless with a con-
stant temperature of 38.4 C. This case deserves mention from another
standpoint.

In Figure 3 and Table 2, it will be seen that during the period of
high temperature, 100-104 F., there was a persistent negative nitrogen
balance similar to that observed by Coleman and Du Bois ' in typhoid
fever. This occurred during a time when he was receiving a well
balanced ration greatly in excess of his heat production as measured by
several calorimeter observations. This phenomenon was so distinctly
abnormal that it is necessary to assume a toxic destruction of body
I)rotein caused by the gonorrheal arthritis. Since this negative nitro-
gen balance continued during a period when the temperature was
comparatively low, it is natural to infer that the destruction of protein
was due to toxins rather than the hyperpyrexia.

7. Coleman and Du Bois : Clinical Calorimetry. Paper 7.— Calorimetric
Observations on the Metabolism of Typhoid Patients With and Without Food.
Arch. Int. Med. 15:882 (May) 1915.

Subject.

Sunace

Area, Linear

Formula

Period

End

of

Period.

Time

Carbon

DIoxid,

(im.

Oxygen,
Um.

R.Q.

Water,
Gm.

Urine N

H-'o'^.^r,
Gm.

Indirect

Calo-
rim^e^try.

Heat
Elimi-
nated,
Cal.

William B

12/6/16

Prelim.

12:14

1

1:14
2:18

26.2
28.9

24.2
25.3

0.79
0.83

31.1
35.1

0.39
0.49

80.2
84.7

78.6
82.6

Aver.

....

80.0*

William B

:2 11/16

63.6 Kg.

1.80 Sq. M.

Prelim.

1

2

Aver.

11:50
12:50
1:50

23.9
24.2

21.3
22.2

0.82
0.79

28.3
29.8

0.48
0.48

72.0

77.0

76.4

WiUiamB

12/8/16

63.6 Kg.

1.80 Sq. M.

Prelim.

11:57

....

....

2

12:57
1:57

27.6
28.8

21.5
21.8

0.93

31.7
31.7

0.58
0.58

73.7
74.9

75.0

75.6

3

2:59

33.0

24.5

0.98

37.0

0.58

82.3

S4.6

William B

12/13/16

lAM Kf.
1.80 Sq. M.

Prelim.

11:16
12:17
l:16

30.7
31.8

24.1
28.4

0.93
O.Sl

35.7
46.8

1.00
l.OO

79.7
9-2.6

S4.2
91.9

3

2:16

33.9

29.6

0.84

55.2

l.OO

96.9

99.8

4

3:16

32.9

29.5

0.81

58.2

2.24

96.0

102.0

Edward R

1,'3/17
37.8 Kg.
1.42 Sq. M.

Prelim.

1

2

Aver.

12:06
1:06
2:06

15.7
15.8

14.8
14.5

o.rr

0.79

23.5
21.7

0.12
0.12

48.8

48.4
49.1

Edward B

Prelim.

12:19

1/9/ir
38.8 Kg.
1.42 Sq. M.

1

1:19
1:59

23.7

16.8

1.03

34.7

0.09

59.1

57.9

3

2:59

25.2

19.0

0.97

40.9

0.94

66.4

66.6

Edward McK

1/5/17
38.0 Kg.
1.37 Sq.M.

Prelim.

1

12:03
1:03

19.3

17.0

0.82

23.8

0.10

....
56.7

2

2:03

20.0

17.4

0.84

24.2

0.10

58.2

Aver.

58.1

....

Edward C

Prelim.

11:52

1/22/17
51.7 Kg.
1.56 Sq. M.

Aver.

1:52

21.1
20.1

20.1
19.4

0.77
0.75

33.1
30.2

0.43
0.43

64.4

70.7
68.7

Prelim.

1/26/17
80'4 Kg.
2.01 Sq. M.

1

2

Aver.

12:46
1:46

26.8
26.9
....

25.7
26.0

0.76
0.75

33.4
34.7

0.63
0.63

84.7

76.9
832

1/29/17
61.2 Kg.
1.80 Sq. M.

1
2

12:58
1:58

30.6
30.2

29.3
28.3

0.76
0.78

42.3
40.6

0.93
0.93

95.6
92.8

94.0
93.8

3

2:68

30.6

29.2

0.76

41.9

093

95.3

90.0

Aver.

....

Average calories per hour.

-Calorimeter Data

Direct
Calo-
rimetry
(Rectal

Rectal

Aver- '
Piflse

Work
Adder.

Cm.

Kon-
proteiD
R.Q.

Per Cent.
Calories from

Calories
per Hour

Remarks

Pro- j
tein

Fat

Carbo-

hyd.

11^ i sTu.

(Linear)

....
81.7
82.4

37.1
37.2
37.2

37.0

60
64

16

28

O.Sl

16

55

29

1.26
....

44.6
44.1

IQ chair

[Apparently quiet, pa-
tient says he exercised

[ fingers constantly

Second period prolong-
ed because of tailing
barometer

Basal, in chair

V6.6

37.0

60

11

Quiet

S2.4

37.1

€4

20

Quiet, voided

O.Sl

18

53

29

1.13

40.0

69.9

37.2
37.1

65

17

0.9«i

IG

12

72

1.16

40.9

Dextrose 212 gm., 10:56-

11:07 a. m.
Quiet, asleep 15 min.

76.3

37.1

64

21

0.99

16

2

82

1.18

41.6

Quiet

8.5.1

37.2

74

21

1.01

16

0

84

1.25

44.3

Quiet

76.8

37.6

37.4

68

^

1.03

;;

1.25

44.3

Chopped beel, 662 gm.

(24.3 gm. N)
Moderately quiet

90.9

37.4

68

26

0.82

28

44

28

1.48

53.0

Moderately quiet

98.7

37.4

24

0.S5

27

38

35

1.53

54.6

Quiet

101.4

37.5
36.9

69

21

0.82

62

23

1.48

52.9

Quiet
Basal

47.9

36.9

90

6

Very quiet

45.6

36.9

12

Very quiet

....

0.78

7

70

23

1.29

34.4

57.3
63.0

37.2
37.2

37.1
37.0
36.9

145
57

6

....

1.04

0.97

4
3

0
10

96
S7

1.52
1.71

41.8
46.7

Dextrose, 212 gm., 10:19

Vei-y 'quiet, pain; un-
comfortable at start
ol second period

Removed from calorim-
eter, pillows shifted

Very quiet, pain

Basal

52.4

36.7

12

Fairly quiet

57.6

36.7

82

2

very quiet

0.83

4

55

41

1.53

42.4

37.4

Basal

60.5

37.2

76

22

....

Somewliat restless

71.0

37.3

13

Quiet

0.75

19

68

13

1.25

41.3

37.7

Basal

73.9

37.6

80

6

Very quiet

81.5

37.6

4

Very quiet

0.74

20

70

10

1.05

42.2

38.4

92

Basal

89,9

38.4

87

21

0.75

26

64

10

1.57

Restless, voiding

95.9

38.4

84

14

\ 0.76

27

59

14

1.52

Fairly quiet

ViO

38.4

1 ••••

90

18

0.75

26

03

11

::" 1 ..

Fairly quiet

ARCHIVES OF IXTERXAL MEDICIXE

TABLE 1. — Calorimeter-

Subject.

Date,

Weight,

Surface

Area. Linear

Formula

Period

End

ol

Period.

Time

Carbon

Dioxid,

Gm.

Oxygen.
Gm.

R. Q.

Water.
Gm.

Urine X

Indirect
Calo-

Cal.

Heat
Elimi-
nated,
Cal.

John Br

Prelim.

11:58
12:58
1:58

26.4
26.8

25.1
24.0

0.76
0.81

47.0
35.2

0.72
0.72

1/31'17
l.Tsli^^Wt.)

87.8
81.9

Aver.

80.8

John Br

Prelim.

1

11:51
12:47
1:51

25.2
28.0

25.6

0.61
0.81

....
28.2
32.2

0.57
0.57

z

S/9/i7

:.r/;^at.

73.9

88.5

Aver.

80.2

Timothy S. ^

53.4 Kg.
1.64 Sq.M.

Prelim.

1

aI,

12:12
1:12
2:12

26.1
26.1

24.0

0.79
0.79

31.2
30.2

0.57
0.57

79.5

78.7
78.1

Tiinothv S

4/27/17
53.4 Kg.
1.64 Sq. M.

Prelim.

2
Aver.

11:33
12:33
1:33

23.6
24.0

21.1
21.5

,0.82
0.81

28.6

27.7

0.62
0.62

70.6

69.2
70.1

Timothy S

l.M Sq. M.

Prelim,

2
Aver.

11:06
1:06

21.9
23.2

IS.l
20.3

0.88
0.83

25.7
25.8

....

0.53
0.53

64.3

61.0
65.0

Joseph McC

2/7/17, 51.4 Kg.
1.46 Sq.M.

Prelim.

1

11:09
12:09

20.9

18.6

0.82

27.8

^::

62.4

61.3

Average calories per hour,

METABOLISM IN GOUT

The basal metabolism in one case of gout was determined.

Case 5.—

History.— Timothy S., a laborer, born in Ireland, 42 years of age, was
admitted .-Xpril 21. 1917. and discharged improved May 12, 1917. He has never
had any serious illness and says that he is not alcoholic. Attacks of gout began
in 1S05 when he was 30 years of age. The first attack involved both great
toes. Since that time the attacks have occurred at intervals of from 'wo months
to one year. The toes, knees and fingers have been involved. The present
attack involved both knees and the small joint of the right hand.

Physical Examination. — The patient was a moderately emaciated, fairly well
developed man. The pinnae of both ears showed large and small tophi. There
• were small tophaceous spots in the ear drum. The throat was moderately con-
gested. The heart was not enlarged. Blood pressure: Systolic. 155 mm.:
diastolic, 105 mm. The arteries were palpable but not sclerosed. The meta-
carpophalangeal joints of the thumb and the first three fingers of the right hand
were tender, swollen and painful on motion. The right wrist and elbow were
slightly involved. The left hand showed marks of deformity Init no active

CECIL ET AL.-
-(Continued)

■METABOLISM OF ARTHRITLS

Direct
Calo-
rimetry
(Rectal

R«tal

Temp.,

Cal.

Aver-
/^fe

Work

NOD;

Per Cent.
Calories from

Calories
per Hour

Remarks

'r q'.° "p

?n

Fat

Carbo-
hyd.

It

Per

Sq. M.
(Linear)

38.1

78

Basal

-4:5

.37.9

76

21

....

Fairly quiet

77.4

37.8
37.8

74

12

0.78

52

1.35

45.4

Quiet
Basal

BC7

37.7

»1

27

Restless

90.1

37.7

80

27

Restless

....

0.81

52

29

1.40

46.1

37.7

Basal

74.8
80.9

67.1

37.6
37.7

37.4
37.3

80

75

13
13

0.78

9

59

1.49

48.5

Very restless
fDnsatistactory because
1 of pain and restless-
(ness

Basal

Very quiet, turned twice

71.1

37.4

8

Very quiet

081

3

49

1.32

43.1

37.6

''

Basal

62.4

37.6

76

2

Very quiet

67.8

37.7
37.5

..

10

0.87 :

2

.

1.23

39.2

Very quiet until 20 min.
before end

Basal

60.3

37..i

G4

23

0.82

1.21

40 0

Very quiet

inflammation. The left knee joint was slightly swollen, e.xquisitelv tender and
very painful on motion. The left great toe showed marked deformity and
moderate tenderness.

Laboratory Examination. — Urine showed a trace of albimiin with many
granular casts. Phenolsulphonephthalein test: 17 per cent. Blood: leukocytes,
11,000; polymorphonuclears. 66 per cent. Blood uric acid: 7.9 mg. Roentgen-
ray examination showed hypertrophic osteoarthritis of all the joints affected.

April 26 he was observed in the calorimeter. His metabolism was found
to be 23 per cent, above the average normal. He was restless and suffered
much pain. On the following day he had less pain and was quiet. He was
again observed and the metabolism was found to be 10 per cent, above the
average normal basal. By May 12 he had entirely recovered from the symptoms
of his attack. He was again observed in the calorimeter. This time his
metabolism "was found to be normal. He was discharged May 12, 1917.

The results of the study of basal metaboHsm will be found in
Tables 1 and 4. The first observation on .April 26 was unsatisfactory
because of the pain and restlessness of the patient. In the two sub-

594

ARCHU'ES OF IXTERXAL MEDICIXE

sequent observations, the metabolism was 12 per cent, and 22 per cent.,
respectively, above the average normal level, practically within the
limits of normal.

TABLE 2. — Clixic.al Data on Tohn Browi

Body
Weight

Esti-
mated

Produc-
tion per
24 Hours

Food

Food
N,
Gra.

Urine
N,
Gn,.

Ex-
Gm.

Nitrogen

Balance

Gm.

!4-Hour
Urine
,-olume.

Date

Total
Calories

Car-
bohy-
drate,

Gm.

I'J:

1/28 '17
1/29/17
1/30/17
1'31'17

2/ 3/17
2/ 4/17
2/ 5/17
2/ 6/17
2/ 7/17
2/ 8/17
2/ 9/17
2/10/17
2/11/17
2/12/17
2/13/17
2/14/17
2/15/17
2/16/17
2/17/17
2/18/17
2/19/17
2/20 '17
2/21/17
2/22/17
2/23/17
2/24/17
2/25/17
2/26/17
2/27/17
2/28/17

62.4

61.3

60.5 av.

59.7

60.0

58.5

59.1

59.7

59.5 av.
59.4
69.0
68.5
57.4

57.6 av.
57.8
57.7

58.0 av.
58.3

58.3 av.
58.2

58:3 '"'■

58.6
59.0

59.1 av.

59.3 av.

69.4 av.
59.6
69.6
59.8
59.8 av.
69.9

i'eoo
■i.m

2,666

2,126

1,270
1,320
1,910
2.344
2,700
2,590
3.060
3,130
3,130
3,090
3,100
2,930
2.880
3.810
3,500
3,760
3,510
3,730
3,500
3,610
3,500
3,660
3.480
3,470
3,490
3,520
3,520
3.500

3,220
3,460

138
156
156
261
243

1?
322
340

i

292
381
•378
428
380
416
378
402
377
389
381
S75
380
368
353
382
388
398
323
368

105

109
141

141

144
146
144
147
139
177
168
173
169
176
169

Ji
169
165
167
165
174
179
167
167
166
162
168

8.4
7.3
11.3
10.2
15.4
12.2
15.1
15.2
15.7
16.2
15.1
15.0
15.2
15.6
15.2
15.3
15.0
15.2
15.1
15.3
15.1
15.5
15.2
15.2
15.4
15.3
15.6
14.9
15.1
15.2
15.0
15.2

23.5
24.4
24.3
21.3
21.1
20.2
21.3
20.3
20.6
19.8
20.3
18.5
18.7
18.5
17.7
17.4
15.9
14.6
13.7
14.2
15.0
14.0
14.7

11:?
11.7
11.5
11.6
11.5
11.9
11.3
11.5

24.3
25.1
25.4
22.3
22.6
21.4
22.8
21.8
22.2
21.4
21.8
20.0
20.2

18.9
17.4
16.1
15.2
15.7
16.5
15.6
16.2

Iti
13.2
13.1
13.0
13.0
13.4
12.8
13.0

—15.9
—17.8
—14.1
—12.1

— 7.2

— 9.2

— 7.7

— 6.6

— 6.5

— 5.2

— sio

— 5.0

— 4.5

— 4.0

— 3.6
—■2.4

— o!i

— 0.4

— 1.4

— 0.1

— 1.0

-1- 2li
-1- 2.5
-1- 1.9
+ 2.1
+ 1.8
-f- 2.2
-1- 1.5

1,670
l,490t
1,460
1:220
1,510
1,150
2,420
1,560
2,520
1,680
2,080
1,680
1,250
1,560
1:370
1,790
1,580
1,300
1,960
1,450
1,470
1,100
1,800
1,300
1,470
1,290
1,360
1,390
l,4iO
1,330
1,300
1,240

METABOLISM IN" CHRONIC DEFORMING ARTHRITIS

The basal metabolism of three cases of arthritis deformans was
studied. Four cases were examined for evidence of the toxic destruc-
tion of protein. The eflfect of the ingestion of large amounts of protein
and carbohydrate was determined in two of the patients. The histories
of the four patients follow :

Case 6. — Arthritis Deformans.

History. — Wm. M., a traveling salesman, born in the United States, SO
yefirs of age, was admitted Oct. 16, 1916. and discharged improved Jan. 17,
1917. In 1899 he had gonorrhea followed by epididymitis. The discharge
lasted for two weeks. In 1909 he was told that he had syphilis and was treated
for a week. ./\s far as he knows, he never had a chancre nor has he shown
any secondary symptoms. His wife had three miscarriages at about three
months. He has had occasional slight sore throat Init no previous attacks of
joint pain.

July 21, 1916, he was awakened in the middle of tlie night by severe pain in
the left arm, left wrist and right ankle. He thinks he had some fever. Two
(lays later he went to St. Vincent's Hospital where he was treated until

TABLE 3.— Clinical Data ox Arthritis Deform.

Body Heat Car-

Wt. Produc- Total: bohy-' Tat,
tionpen Calo- drate,] Gm.
24 Hrs. I ries Gm. |

Food Urine Ex-

Nitro-| j

gen I Urine
Bal- Vol- Length of

II/-26 Ifi

11 2- 16
11/28/16
11/29/16
11/30/16
12/ 1/16
12/ 2/16
12/ 3 '16
12/ 4/16
12/ 5/16
12/ 6/16
12/ 7/16
12/ 8/16
12/ 9/16
12/10/16
12/11/16
12/12/16

12 13 16
12/14/16
12/15/16
12/16/16
12/17/16

12/24/16
12/25/16
12/26/16
12/27/16
12/28/16
12/29/16
12/30/16
• 12/31/16
1/ 1/17
1/ 2/17
1/ 3/17
1/ 4/17
1/ 5/17
1/ 6/17
1/ 7/17
1/ 8/17
1/ 9/17
1/10/17
1/11/17
1/12/17
1/13/17
1/14/17
1/15/17

Edward McK.
12/21/16
12/22/16
12/23/16
12/24/16
12/25/16
12/26/16
12/27/16
12/28/16
12/29/16
12/30/16
12/31/16-1/1/17
1/ 1/17
1/ 2/17
1/ 3/17
1/ 4/17
1/ 5/17
1/ 6/17
1/ 7/17
1/ 8/17
1/ 9/17
1/10/17
1/11/17
1/12/17

2,190
2,600

207

271

2.000

2,690

293

2.510

2,5.50

257

2.510

260

2„100

2,500

296

30O

2,520

297

2,930

2,970

355

3.020

366

2,970

355

2,860

351

3.060

.369

2,980

356

2.660

296

2,530

266

3,a30

2,800

269

3.050

3.060

327

3.020

3,010

3.010

317

317

318

3.010

319

3,060

2,9(iO

318

2,940

315

1,780

138

3,020

316

2.990

3,050

319

3,010

323

1.9T0

207

1.910

181

1,350

1.220

123

1,2.50

133

1,210

148

l,'5fl6

1,490

174

1,490

174

1.490

2.05O

244

2,180

244

1.940

241

1,990

235

1,880

232

1,530

163

1,530

161

1,.590

1,520

1,550

161

1,560

162

3.4

3.3

2.8

3.2

2.9

! 3.2

3.1

i 3.4

3.0

15.2

15.0

7.1

15.7

7.1

1.5.2

1.5.4

12,6

14.8

11.7

15.2

15.0

10.2

i 14 9

10 7

1.5.1

14.7

11.4

14.9

14.6

9.9

15.4

10.1

15.3

11.6

10.2

9.1

7.4

6.8

6.5

7.3

7.6

7.4

7.5

7.1

2.6

4.o:

.5.1

4.5

2.6

3.6

2.6

3.3

2.6

2.3

3.4

2.4

3.2

2,2

2.7

2.7

2.6

7.4

3.3

.5.9

7.6

7.6

5.1

7.6

5.5

7.7

5.9

5.3

+2.5

2,.5fl0

24 hrs

+ 1.3

l.»iO

24 hrs

+0.5

2.020

24 hrs

+1.9

24 hrs

+1.2

l.,«.S5

24 hrs

+1.0

2,170

24 hrs

2.000

24 hrs

+0.3

2,070

24 hrs

+0.4

24 hrs.

+1.3

2,5(;0

24 hrs.

—7.9

1,410

24 hrs.

1,20-)

-3.3

1,670

24 hrs.

—1.4

1,470

24 hrs.

1.220

24 hrs.

—0.3

2,100

24 hrs.

-0.3

—0.7

1,700

24 hrs.

—0.3

1,510

24 hrs.

1,8.30

24 hrs.

+0.1

1,700

24 hrs.

+8.6

2,010

24 hrs.

+6.4

+7.0

1.320

24 hrs.

+2.5

2,450

24 hrs.

+2^

1,770

24 hrs.

2,280

24 hrs.
24 hrs.

+1.8

2 000

+5 8

1,270

+3.2

1,720

+0.6

1,976

+4.0

+2.8

2.030

+0.7

2,310

+1.3

815

—01

1,1 fO

-0 6

850

-0.4

780

1,020

1.550

-1.2

1,230

—1.0

1,080

—0.8

1,370

—1.1

890

—1.3

1,260

-1.0

—0.7

1,140

—0.5

1,080

—0.2

900

+0.8

1,320

+1.9

1,540

+1.7

+1.3

890

+0.4

1.220

24 hrs.

24 hrs!
24 hrs.
24 hrs.
24 hrs.
24 hrs.
24 hrs.

24 hrs. 5 min.

24 hrs.

17 hrs.

■25 hrs. 40 min.

24 hrs. 10 min.

23 hrs. 45 min

24 hrs. 15 min.
23 hrs. 50 min.

23 hrs. 50 min.

24 hrs.

23 hrs. 50 min

24 hrs.
24 hrs.
24 hrs.

23 hrs. 50 m-n.
23 hrs 50 min.

Urine nitrogen plus 10 i
Approximate.
Incomplete specimens.

of food nitrogen.

ARCHIVES OF IXTERXAL MEDICI XE

TABLE 3.— Clinical Data on Arthritis De

-(Continued)

Esti-

Food

Nitro-

mated

Food

Urine

Ex-

gen

Urine

Name and

Bodv

Car-

N,

N,

Bal-

Vol-

Length ol

Date

Wt.

Produc-

Total

bohy-

Fat,

Gm.

Gm.

",*"

ance,

ume,

Period

tion per

Calo-

drate.

Gm.

Gm.

Gm.

C.c.

24 Hrs.

ries

Gm.

Edward R.

r2/2I/lfi

39.5

1,990

196

10.6

7.3

8.4

4-1.1

450

24 hrs. 15 min

12/22/16

1,340

159

64

7.5

6.6

7.4

4-0.1

600

23 hrs. 55 min

12'23'ie

1,410

162

58

5.9

-H.2

80O

24 hrs. 10 rain

12 '24; 16

1,220

123

57

7:6

6.6

7:4

-1-0.2

1,400

23 hrs. 50 min

12/25/16

12 26 It;

38.S

1.35<

150

56

7.f

5.5

£0 hrs. 10 min.

12/27 16

1,204

147

67

2.6

3.9

940:

21 hrs. 55 min

12 2S/I6

1,28(

168

57

2.1

3.8^

1,870-

22 hrs. 25 min

12/29/10

1,490

174

77

2.6

'600!

14 hrs. 45 min

12/30/16

1,500

174

2.6

4:2;

1,450!

11 hrs. 20 min

12/31/16

39.2

1,490

174

77

2.6

3.1

'3:4

.^'.8

1,£0!>

24 hrs. 35 min

1/ 1/17

1.490

77

2.6

3.1

3.4

—0.8

1,610

23 hrs. 50 min

1/ 2/17

:::::

1,490

174

2.6

3.4

-1.1

1,760

23 hrs. 40 min

1/ 3/17

1,268

2,060

242

108

2.4

2.3

2:6

-0.2

695

22 hrs. 40 min

1/ 4/17

2,180

244

121

2.4

2.7

2.9

-0.0

1,420

1/ 5/17

1,970

231

104

•2.3

26

2.8

—0.5

1,340

26 hfs: 15 min

1/ 6/17

1,990

235

104

2.3

2.3

2.5

-0.2

1,270

21 hrs. 55 min

1/ 7/17

37!6

1,77C

169

81

12.5

5.3

6 6

-1-5.9

1,330

23 hrs. 50 min

1/ 8/17

i;45C

155

67

7.2

5.4

5.9

—1.3

1.800

56 hrs. 25 min

1/ 9/17

i,764

3.4)

1.050'

17 hrs. 3' min

1/10'17

1,550

iei

'74

'7:8

5.6!

1,810!

IS hrs.

1/11/17

1,660

175

73

8.1

6.0

'6:8

1,660

24 hrs. SO min

1/12/1-

■1,530

163

72

7.5

4.3

5.1

-1-2:4

1,700

21 hrs. 25 min

1/13/17
1/14/17
1/15/17

1,540

161

7.6

5.2

6.0

-1-1.6

1,250

24 hrs. 10 min

....

1,602

170

77

7.4

5.5

6.2

4-1.2

1,440

24 hrs.

Frank H.

11/24/16

4n.a

2,520

256

115

14.8

89

10.4

4-4.4

1,150

24 hrs.

11/25/16

41.1

2,590

277

114

15.3

9.s:

11.3

4-4.0

785!

24 hrs.

11/26/16

41.1

2,520

260

115

15.1

11.2;

12.7

4-2.4

l,050i

24 hrs.

11/27/16

41.6

2'470

249

114

15.0

12.6

14.1

4-0.9

1,200

24 hrs.

l]/2!r/16

J.....

2,470

249

112

15.7

12.6

4-1.5

980

24 hrs.

11 29 16

4i!5

2,520

117

14.6

12.8

14^3

4-0.3

960

24 hrs. •

11/30/16

41.3

2,1€0

80

13.6

12.8

14.2

-0.6

860

24 hrs.

12/ 1/16

41.5

2,500

255

115

15.0

12.3

13.8

4-1.2

900

24 hrs.

12/ 2/16

2.4.50

112

15.0

12.1

13.6

4-1.5

1,000

24 hrs.

12/ 3/W

42!2

2,490

114

14.8

13.4

149

1,030

24 hrs.

12/ 4/16

42.8

2:530

304

131

2.4

14.5

14.7

—12:3

1,200

24 hrs.

12/ 5/16

41.8

2,580

308

135

2.4

4.7

—2.5

790 24 hrs.

12/ 6/16

42.3

2,620

299

144

52;

5:4

—3.3

980! 24 hrs.

12/ 7/16

41.4

2:530

301

132

2.6

4.0

4.3

850

24 hrs.

12/ S/16

42.0

•268

132

2.4

4.0

—1:8

780

24 hrs.

12/ 9/16

41.7

2:520

132

2.7

4.6

49

—2.2

960

24 hrs.

12/10 16

41.6

2„530

301

132

2.6

3.6

3.9

— l.S

1,050

24 hrs.

12 11/16

2,540

304

132

2.6

3.5

38

—1.2

1,060

24 hrs.

12/12/16

ii'.i

2,590

317

132

2.7

3.6

3.9

-1.2

930

24 hrs.

12/13/16

2,540

304

132

3.5

3.8

—1.2

1,300

24 hrs.

12/H 16

4i!4

2,510

2»7

132

"•-.

3.6

-0.9

675

24 hrs.

12/15/16

41.3

2,480

286

J33

2:7:

4:2

4.5

—l.S

600

24 hrs.

12 16, Hi

41.1

2,520

298

133

2.6

37

—1.1

660

24 hrs.

12'17/1<;

42.0

2:290

229

106

14.4

e.i

7.9

4-6.5

1,050

24 hrs.

12/18 11;

40.7

2,590

268

117

15.7

7.8

9.4

4-6.3

1,000

24 hrs.

12/19/16

42.0

2,650

296

113

15.2

8.6

10.1

4-5.1

1,100 24 hrs.

October 16. During this time, his left ankle and right hand were involved.
The right ankle improved slightly but the elbows caused great pain. The right
elbow gradually stiffened until only slight motion was possible.

Physical Examination.— lie was poorly nourished, well developed, rather
apathetic man. He had no temperature and was not toxic. The teeth showed
moderate pyorrhea. The tonsils were small with rather deep crypts. The right
elbow was the site of very painful inflammation. The joint was swollen; the
forearm was held at a right angle to the arm. Movement was impossible :
the effort was very painful. All of the muscles of the arm and forearm were
flabby and wastc<l. Tliere was but little redness about the swollen parts. The

CECIL ET AL.—METABOLISM OF ARTHRITIS

597

fingers of the right hand could not be flexed. There was no sign of inflamma-
tion, the stiffness probably being due to disuse. There was some limitation of
motion, but no pain, in the right shoulder. The left elbow showed thickening
of the soft parts but no pain or active inflammation. The right ankle was
swollen and slightly painful but not red or hot. The actual involvement of
the joint was hidden by a diffuse superficial edema of the foot and ankle.
There was slight limitation of dorsal and plantar fle.xion and some limitation of
eversion and inversion. The left ankle was normal. Massage of the prostate
caused no discharge. The seminal vesicles were not palpable.

TABLE

Summary of B.as.«il Met.^bolism Studies in- .Arthritis

Acute rlieuniotic
fever (mild)

Gonococcus
arthritis

John Br

Gonococcus
arthritis

John Br

Gonococcus
arthritis

JohnBl

Rheumatic fever

Joseph Mc

Gonococcus
arthritis

Timothys

Gout

Timothy S

Gout

Williams

Severe arthritis
deformans

Edward R

Severe arthritis
deformans

Edward McK

Severe arthritis
deformans

tion

1/22/16

1/29/17

1/31/17

2/ 9/17

l/2(i/17
2/ 7/17

4/27/17
5/12/17
12/11/16

1/ 3/17

Sur-

.iver-

«^effal
.■emp..

H

.\ver-

Ties

In-
direct

Per
Cent.
Devia-
tion
from
Nor-
mal

N-

M.ni-

Remarks

1.56

37.3

0.76

64.4

-r 5

Slig!,t!y
re»tle.-!S

ISO

38 4

0.77

94.6

-1-29

Slightly
restless

1.78

37.9

0.79

S0.8

+11

Quiet

1.74

37.7

O.Sl

80.2

-t-12

Restless

2.01

37.6

0.76

84.7

+ 12

Very
quiet

1.46

37.6

0.82

+ 8

Very
quiet

l.(i4

37.4

0.82

70 6

-1-12

Very
qu:et

1.64

37.6

0.85

64.3

+ 2

Quiet

ISO

37.0

0.80

72.0

+ 7

2.9-3.3

Quiet

1.42

36.9

0.78

'

—13

26-3.1

very
quiet

1.37

36.7

0.63

53.1

+ 7

2.6-2.7

Quiot

• Urinary nitrogen for 24 hours on low nitrogen intake.

Laboratory Examination. — Urine: Negative. Blood: Leukocytes, 6,000;
polymorphonuclears, 80 per cent.; erythrocytes, 5,000,000: hemoglobin (Sahli)
90 per cent. Nonprotein nitrogen : 57 mg. per 100 c.c. Uric acid : 4 nig. per
100 c.c. Wassermann: negative. Gonococcus fi.xation test: once positive; twice
doubtful.

Roentgen-Ray Reports The teeth showed a considerable amount of alveolar
recession, with accentuation of the pericemental membrane, suggestirg pyorrhea.
There was also an area of rarefaction around the root of the second right
upper bicuspid, with imperfect root canal filling. The ankles showed a moderate
amount of periarticular atrophy, with a narrowing and clouding of joint spaces,

ARCHIVES OF JXTERXAL MEDIC IS E

most marked in the right ankle joint and tarsometatarsal joint, the appearance
suggesting an atrophic adhesive osteoarthritis. A similar condition was notite-
able in the right elbow.

October 30 the second right bicuspid was extracted. The tooth was dead
and somewhat carious in the root canal. On extraction there was a marked
putrid pus odor. A culture from the cavity of the gum was negative.

November 27 he had an acute attack of follicular tonsillitis. The temperature
rose to 103 F. The right elbow was more swollen and extremely painful.
There was pain on any motion of the right shoulder. A tonsil culture showed
a pure growth of a nonhemolytic streptococcus. A blood culture was sterile.

December 8. after receiving 200 gm. glucose, he was observed in the calori-
meter. On the following day, the joint condition was unchanged. Basal obser-
vations were made December 6 and 11. December 13 he was given meat in
large quantities and was afterward observed in the calorimeter. From
December 19 to January IS he received frequent doses of Jobling's proteose,
the observations On which are mentioned in another article. The joint condi-
tion improved very gradually. With constant baking and massage he regained
motion in the fingers of the right hand. ^Motion in the elbow was not increased.

Fig. 4.— Willia

Case 6) . Roentgenogram of right elbow.

The ankles caused little pain bitt the edema of the right foot and ankle con-
tinued to the time of his discharge on January 17. He was seen again in
February. The disease had not advanced. He w^as working every day but had
regained no motion in the elbow.

Case 7. — Arthritis Deformans (many ankyloses').
History. — Edward R., a machinist, born in Austria, 32 years of age, was
admitted Dec. 19. 1916, and discharged unimproved Jan. 16, 1917. A sister has
deforming arthritis. The patient is a skilled mechanic who had excellent work-
ing and home conditions. His habits have been good. He had diphtheria in
early childhood. For about one year preceding the onset of the arthritis he
expectorated large amounts of foul smelling material which came from the
nasal passages and which ceased about the time the joint trouble began. He
persistently denies any venereal infection.

In the winter of 1911 he began to have pain in his right great toe. Two
weeks later pain appeared in the other toes of the right foot. Soon afterward
pain developed in the other foot and before a year had passed both knees were
affected. During this time he was able to continue with his occupation but

CECIL ET AL.— METABOLISM OF ARTHRITIS 599

required help in going to and from work. During the next eight months he
was treated in a hospital where traction was applied to his legs. Roentgeno-
grams were taken of his teeth as a result of which one tooth was extracted.
While in the hospital the shoulders, elbows, hands, jaws, hips and the sacroiliac
synchondroses were involved. In each new joint affected there was some swell-
ing and pain, very severe in the wrists, ankles and toes, moderate in the other
joints. The swelling and pain gradually subsided and stiffening occurred.
Subsidence was a matter of months or years. So far as he knows, he has
never had fever. During the first two years of the illness his weight dropped
from 148 to 96 pounds.

Fig. 5. — Edward R. (Case 7). Roentgenogram of left hand

Physical Examiiialion. — His face was that of a well nourished man. Com-
plexion florid. There was, however, tremendous wasting of the muscles which
gave his body the appearance of extreme emaciation. He laid in dorsal
decubitus, the thighs in line with the body, the knees extended, the elbows
flexed. His skin was moist, shiny and of very fine texture. He was of a
cheerful, philosophic disposition. Ears were normal. The teeth and gums
appeared healthy. There was no pyorrhea. The tongue and throat could not

600 ARCHIVES OF IXTERXAL MEDICIXE

be seen because of ankylosis of the jaw. The joints showed many deformities.
The jaws could be opened only a quarter of an inch. Xo lateral motion was
possible. There was no tenderness but distinct crepitus over the temporo-
mandibular joints. The right shoulder and wrist, both elbows, hips,
knees and ankles were ankylosed but still tender. The sacroiliac joints
were tender; very limited motion was possible in the right shoulder. The
hands showed extreme deformity which w'as best demonstrated by the roentgen
ray. In both hands were numerous subluxations. There was complete ankylosis
of all joints except the distal interphalangeal joints of the right thumb, right
fifth finger and left ring finger, which were not at all affected, and the
metocarpophalangeal joint of the thumb, which showed free motion but was
swollen, very painful and tender. The toes of the left foot were ankylosed.
Those of the right foot showed very little motion.

Laboratory Examination. — Urine : Faint trace of albumin ; no casts. Blood :
Leukocytes, 7,000; polymophonuclears, 60 per cent.; hemoglobin, 85 per cent.;
erythrocytes, 5,000,000. Wassermann : strongly positive. Gonococcus fixation
test: negative. Nonprotein nitrogen: 45 mg. per 100 c.c. Uric acid: 3 mg.
per 100 c.c.

Roentgen-ray report: In the hand there is obliteration of intercarpal, carpo-
radial, metacarpophalangeal and interphalangeal joint spaces with extreme bone
atrophy. Examination of hip shows extensive bone atrophy without excrescence
formation, also a diminution of the joint space. Knees, shoulders and elbows
show a similar condition. There appears to have been in the joints an absorp-
tion of the cartilage and a fibrous ankylosis without bone destruction.

While in the hospital he had pain in many of the joints. Pain was most
severe and constant in the metacarpophalangeal joints of the thumb which, he
says, was the last joint to be affected. About a week after admission he
first noted pain and tenderness over the two upper cervical vertebrae whicli
had not been previously involved. At the time of his discharge, January 16,
the pain was more severe but .no stiffness had developed. His condition was
otherwise unchanged.

Case 8. — Arthritis Deformans (many ankyloses").

History. — Edward McK., a machinist, born in the United States, 28 years of
age, was admitted Dec. 19, 1916, and discharged unimproved Jan. 16. 1917. In
the spring of 1908, while skating, he fell on his right knee. The fall caused
him only temporary discomfort. One month later, however, the knee became
swollen, stiff and slightly painful. In spite of vigorous treatment during the
next two years, the joint did not improve. It was thought at first that it might
be tuberculous. In October, 1908. however, a specimen of the joint fluid was
injected into a guinea-pig at Roosevelt Hospital with negative results. In 1910,
he contracted gonorrhea which lasted three months but did not affect the knee.
In 1911, the left knee was involved in a similar manner but with more pain.
Four months later the left elbow was affected. During the last year his neck
has been stiff and painful and both ankles have been involved. He has received
competent local treatments, besides numerous vaccines. All the joints have
been affected insidiously. There has never been marked inflammation. So far as
the patient knows, he has never had fever.

Physical E.xamination.—Wh face was that of a fairly well nourished man.
The body and extremities showed marked muscular wasting and moderate
emaciation. He was of an unstable, emotional disposition and was overcome
by his misfortunes. Two of the molar teeth were carious. There was slight
pyorrhea about the incisors. The tonsils were small and red with very deep
crypts. The left ear drum showed two large white patches, one posterior,
the other anterior to the malleus. There were no signs of active inflammation.
The right ear drum was normal. The lungs showed very poor expansion.
Massage of the prostate caused no discharge. The shoulders were not affected.
The left elbow allowed less than five degrees of motion; it was practically

CECIL ET AL.— METABOLISM OF ARTHRITIS

601

ankylosed at 100 degrees. The right elbow was ankylosed at 180 degrees.
Neither showed signs of active inflammation. The wrists were swollen and
tender. The right showed ten degrees of motion ; the left even less. The
metacarpophalangeal joints of both thumbs and of the right index finger were
swollen, tender and painful on motion. The cervical vertebral articulations were
completely ankylosed. There was marked tenderness along the cords to the
right of the upper three spinous processes. There was slight motion in the
vertebral articulations of the dorsal spine and still more in the lumbar region.
The hips were completely ankylosed in flexion of 110 degrees to the line of
the trunk. The knees were ankylosed, causing the legs to form an angle of
90 degrees with the thighs. The tibia and fibula were sublu.xated backward and
outward on the femur. There was swelling, some tenderness and great pain on

Fig. 6.— Edy

McK. (Case 8). Roentgenogram of right knee.

attempted motion in the left ankle which was practically ankylosed. Motion in
the right ankle was limited to 30 degrees.

Laboratory Examination. — Urine: Negative. Blood: Leukocytes, 12,500;
polymorphonuclears, 75 per cent. ; erythrocytes, 4.700,000 ; hemoglobin, 90 per
cent. Gonococcus fixation test : doubtful. Cultures from both tonsils showed
an almost pure growth of Staphylococcus aureus with a few colon;es of Strepto-
coccus viridans. Nonprotein nitrogen : 24 mg. per 10<) c.c. Uric acid : too low
to estimate.

Roentgen-ray Report: Examination showed osteoarthritis deformans with
excrescence formation. There was marked subluxation at the knee joint. The
elbows and tarsal joints showed evidence of atrophy, diminution in the size of
the joint space, but no bone destruction.

602 ARCHIJ-ES OF IXTERXAL MEDICIXE

January 11 he developed herpes zoster along the course of the fourth and
fifth intercostal ner\-es. Vesicles were numerous but pain was very slight.
January 16 he was transferred to the Metropolitan Hospital unimproved.

Case 9. — Arthritis Deformans; scabies.

History. — Frank H., a proof-reader, born in the United States, 61 years of
age, was admitted Nov. 4, 1916, and discharged unimproved Dec, 20, 1916. His
wife died of "rheumatic gout" at the age of 41. He has led a sedentary life,
is very moderate user of alcohol and tobacco. He has had gonorrhea five or
six times, the last attack ten years ago. History of "hard chancre" twenty-five
years ago ; two soft chancres since then. No history of secondary syphilis.

Fig. 7.— Frank H. (Case 9). Roentgenogram .of right knee.

or of any specific treatment. He has never had tonsillitis or alveolar abscesses.
He has not suffered from respiratory infections. No history of rheumatic
fever or heart disease.

About seven years ago he developed a stiffness in the right knee which
gradually became worse. Five years ago, the left knee began to get stiff. For
the past two years he has noticed a gradually increasing stiffness and outward
deflection of the fingers of both hands. The riglit knee has been painful at
times, but there has never been much pain in the fingers, even on flexion.
Recently there has been stiffness in both shoulders and in the spine. Patient
thinks he had some urethral discharge at the beginning of his illness. He
has lost some weight, but does not know how much.

Physical Examination.— Patient was a middle-aged man, poorly nourished
and somewhat under-developed. Teeth: .Ml molars missing, except two on

CECIL ET AL.— METABOLISM OF ARTHRITIS 603

lower left side which were markedly carious; considerable pyorrhea present.
particularly of lower incisors which were loose in sockets; no tenderness of
gums. Throat : Tonsils red but not enlarged. Several posterior lymph nodes
enlarged on both sides. Heart: Measured 8 cm. to left in fifth space; apex
impulse not felt ; right border of heart at right sternal margin ; sounds faintly
heard, faint systoHc murmur at apex heard also over sternum ; action slow and
regular. Over the entire body, more marked on abdomen and arms, there
were numerous small petechiae. apparently the result of scratching; also many
small ecchymoses and scabs. Tendon reflexes all exaggerated. Joints ; Spine
straight. Considerable limitation of movement, especially in dorsal region
when patient bends forward. Flexion was from the hips, the spine remaining
rigid. Lateral motion was from the lumbar region. Shoulder joints: Arms
could be elevated to an angle of only 45 degrees. Motion in shoulder joints
considerably limited, especially in abduction and rotation. Flexion and extension
good. Wrist joints showed some limitation of flexion and extension. There
was moderate atrophy of infraspinatus and supraspinatus muscles. Metacarpals,
phalanges and knuckles very prominent on both hands due, in part, to atrophy of
the interossei muscles. There was considerable limitation of extension of all
fingers on both hands (45 degrees). Flexion of fingers not limited. Moderate
contraction of flexor tendons, more marked on left side. Hips : Limitation of
abduction (30 degrees). Patient stood with legs bowed, knees prominent on
account of a marked atrophy of the muscles, more marked on the left side.
Right knee measured 33 cm., left the same. Flexion of knee joints, 90
degrees. No fluid in knee joints, ankles normal. Toe joints normal.

Laboratory Examination. — Urine : very faint trace of albumin ; no casts.
Phenolsulphonephthalein test : first hour, 42 per cent., second hour, 23 per cent.
Sputum : Many streptococci : no tubercle bacilli.

Roentgen ray Report : In the chest the aortic shadow was somewhat
broader than normal ; otherwise, the findings were negative. The hands
showed periarticular atrophy of the bones with narrowing of the joint spaces,
and rarefaction of the articular borders. Roentgenogram of the right femur
showed cortical thickening of the femur with bowing and rarefaction. The
skull showed a similar rarefaction. The findings were those of an atrophic
osteo-arthritis.

During the patient's stay in the hospital he sufifered from a subacute
bronchitis which at one time showed an acute exacerbation. The condition in
his joints remained practically unchanged.

The four cases of arthritis deformans were all of severe type, of
long duration, showing great deformities ; in fact, two of the patients
(Cases 6 and 7) were so crippled that it was difficult to collect twei^ty-
four hour specimens of urine and almost impossible to make them end
the period exactly on the minute according to the custom of the
metabolism ward. Since each voiding was collected in a separate
bottle, and the time of voiding recorded at once, it did not destroy
the value of the observation if the urine secreted during one or two
hours was lost. The last columns in Table 3 (Clinical data) show
the exact time of all complete and incomplete specimens. The results
of observations on the basal metabolism of these patients will be found
in Tables 1 and 4. The efTects of large carbohydrate and protein
meals are represented in Figures 8 and 9.

604 ARCHIVES OF INTERNAL MEDICINE

DISCUSSION OF CHRONIC DEFORMING ARTHRITIS

By a majority of writers, chronic deforming arthritis has been
considered a manifestation of infection from some septic focus within
the body. By others, it is thought to be a disease of metabolism. In
considering the results obtained from the study of these few cases,
the two prevailing conceptions of etiology should be kept in mind.

The wasting and emaciation which accompany long standing infec-
tion have led often to the assumption that in such diseases there is an
increase in basal metabolism and a toxic destruction of body protein.
This assumption has never been properly conftrnied. In the chronic
infection of tuberculosis, several cases of which have been studied
by McCann and Barr,^ only slight toxic destruction of protein was
found. The level of metabolism was in certain cases increased but
the increase was usually accompanied by a considerable elevation of
body temperature. Our knowledge of these factors in other chronic
infections is fragmentary and incomplete. If one thinks of deforming
arthritis as a chronic infection, facts concerning the level of basal
metabolism and possible destruction of body protein are of con-
siderable importance.

Basal Metabolism. — By referring to Table G, it will be seen that
two of the patients gave basal figures 7 per cent, above the average,
in other words, within the limits of normal. One gave results 13 per
cent, below the average but he was emaciated to nothing but skin,
bones and ankylosed joints. One usually finds a low metabolism in
such profound undernutrition. There is a most satisfactory agree-
ment between the methods of direct and indirect calorimetry. T!-.e
total in all experiments by the direct method was 1381.5 calories, by
the indirect method 1376.5 calories, a difiference of 0.3 per cent. The
respiratory quotients were at all times normal and indicated no dis-
turbance in the proportion of calories obtained from the diflferent food
constituents. In the study of these three cases, no disturbance of
the normal basal metabolism has been detected.

Nitrogen Balance. — By referring to Table 3, it will be observed
that nitrogenous equilibrium was maintained in all four cases, except
when the nitrogen intake was cut to a very low figure. Even then
the nitrogen excreted never exceeded the food nitrogen by more than
a few grams. Furthermore, by reducing the nitrogen intake to a
minimum it was possible to reduce the nitrogen excretion of these
patients to as low a level as is possible with normal, healthy indi-
viduals. For instance, William B. for five consecutive days excreted
less than 3.6 gm. nitrogen per day; Edward McK., less than 2.7 gm.
for three days; Frank H., less than 3.6 gm. for five days. Edw.Trd R.

8. McCann. W. S., and Barr. D. P.: Clinical Calorimetry. Paper 29. The
Metabolism in Tulierculosis. •.•\rch. Int. Med. 26:663 (\ov.) 1920.

CECIL ET AL.— METABOLISM OF ARTHRITLS

60S

averaged about 3 gm. for a considerable period, but his figures are
somewhat uncertain since he lost a few specimens, as is indicated in
the last column of this table. All of these results are within the normal
limits found by Landergren. Kocher and others, and they indicate
that there is no toxic destruction of protein in chronic arthritis.

If chronic deforming arthritis is a disease of metabolism, one may
expect to find some evidence of diminished ability to metabolize one
or more of the food stufts. Pemberton's extensrve studies have demon-
strated practically nothing abnormal except a high glucose content of
the blood following glucose ingestion. It must be said, that high
sugar curves are also found in diabetes, hyperthyroidism, nephritis,
severe infections and various other conditions. Even in supposedly
normal individuals the curve of sugar in the blood after ingestion of
glucose varies so much that conclusions drawn from apparently
abnormal figures may be very misleading.

NORMAL CONTROL

E.r.QB. 74.7 KG.

lo lo lo lo lo lO

9 o Q ?? £? P

ARTHRITIS DEFORMANS
WM.B. 63.6 KG.

R.Q.CALiJ
1.00 1 00)2-

R.Q.

BASALJij

jar^:

Fig. 8 — Comparison of normal control and arthritis patient in their response
to large carbohydrate meal. Solid lines — level of metabolism; dotted line,
respiratory quotient.

Perhaps, a more direct way of securing evidence of the body's
ability to handle food stuffs is by a study of the respiratory quotients
and the specific dynamic action of the different classes of food follow-
ing their ingestion. Experiments of this kind were done in two of
the cases of chronic arthritis. December 8, William Bl. received 212
gm. dextrose one hour before going into the calorimeter. He remained
in the calorimeter three hours. December 13, the same patient received
662 gm. chopped meat (24.3 gm. nitrogen) before going into the
calorimeter. January 9, Edward R. was given 212 gm. glucose and
was studied in the calorimeter. In Figures 8 and 9 the curves for
heat production as determined by the indirect method and the respira-
tory quotients are compared with normal controls studied in Paper IV

606 ARCHIFES OF IXTERXAL MEDIC IX E

of this series. The results are in no way conclusive, but there are
no significant differences which would indicate an inability to oxidize
either protein or carbohydrate.

The observations give no indication that arthritis deformans is a
disease of metabolism. The patients studied were of the type usually
considered to be the result of a chronic infective process. If this be

NORMAL CONTROL
LOUIS M. 53.5 KG.

ARTHRITIS
WM. B. 64.3 KG.

CO N- CO ID 10

Fig. 9. — Comparison of norir
a large protein meal. Solid
•y quotient.

il control and arthritis patient in their response
ines — level of metabolism, dotted line, respira-

the correct conception, one may say that the infection wa.s unac-
companied by changes in the level of basal metabolism or by toxic
destruction of protein.

SUMM.VRY AND CONCLUSIONS

1. Three cases of acute and subacute arthritis .showed no varia-
tion from the normal basal metabolism. One case of acute arthritis
observed during a continuous temperature of 38.4 C. showed a basal
metabolism 26 per cent. abo\e the average normal level. Other observa-
tions on the same patient during afebrile periods exhibited a metabolism
practically within normal limits. In this case there was a marked loss
of body nitrogen during a period when the energy requirement was
more than covered by a liberal diet. This indicates a toxic destruction
of body protein.

2. One case of gout showed little change in the level of basal
metabolism.

3. Four cases of se\ere arthritis deformans on the Landergren
diet, very low in protein but high in calories, excreted from 2/) to 3.6
gm. nitrogen per day. figures which are well within the normal limits.

CECIL ET AL.— METABOLISM OF ARTHRITLS 607

Three of these patients when tested in the calorimeter had a metabolism
rate close to the average normal level. The respiratory quotients
were normal, and there was no evidence of abnormal respiratory
metabolism following the ingestion of large test meals of glucose and
protein.

4. The observations on arthritis deformans do not indicate that
it is a disease of metabolism. If infectious in origin, it may be said
that the infection is not accompanied by increase in basal metabolism
or by toxic destruction of body protein.

CLINICAL CALORIAIETRY XXXII

TEMPERATURE REGULATION AFTER THE INTRAVENOUS INJECTION
OF PROTEOSE AND TYPHOID VACCINE *

DAVID P. BARR, M.D., RUSSELL L. CECIL. M.D.

AND EUGENE F. Du BOIS, M.D.

With the Technical Assistance of G. F. Soderstrom

NEW YORK

During the last seven or eight years, many clinicians have been
treating arthritic patients by means of intravenous injections of foreign
protein. These produce chills which resemble malarial paroxysms and
aiiford an ideal opportunity for studying in man the phenomena of
temperature regulation. In 1917 the chills and fever in several cases
of malaria were studied in the Sage calorimeter.^ The present inves-
tigation was undertaken as a supplement to the work in malaria in
order to study in more detail the mechanism of the rise and fall of
body temperature.

The gaseous exchanges of patients have been studied by Kraus
and Chvostek ' after giving tuberculin, and in animals by Freund and
Grafe,^ \'ersar, who gave infusions of sodium chlorid,'' and by
Berrar,'^ who used aloin. In general, these animal experiments showed
a rise in total oxidative processes accompanying the rise in temperature.
Sometimes this increase amounted to 130 per cent.

Tie other phenomena which follow the intravenous injection of
foreign protein have been studied in many clinics since the therapeutic
application of this procedure was developed by Ichikawa. Kraus,
Jobling and Petersen and others. Miller and Lusk " reported favorable
results in arthritis and many others have tried their methods and have

* From the Russell Sage Institute of Pathology in affiliation with the Second
Medical Division of Bellevue Hospital.

1. Barr, D. P., and Du Bois, E. F. : Clinical Calorimetry, Paper 2S. The
Metabolism in Malarial Fever, Arch. Int. Med. 21:627 (May) 1918.

2. Kraus, Fr., and Chvostek, F. : Ueber den respiratorischen Gaswechsel im
Fieberanfall nach Injection der Koch'schen Fliissigkeit, Wien. klin. Wchnschr.
4:104. 1891.

3. Freund, H., & Schlagentvveit, E. : Ueber die Waerme Regulation Kurari-
sierter Tiere, Arch. f. Exper. Path. u. Pharmakol. 69:12. 1912.

4. Verzar. Fritz. : Die Wirkung intravcnoser Kochsalziiifusionen auf den res-
piratorischen Gaswechsel, Biochem. Ztschr. 34:41, 1911.

5. Bcrrar. M. : Die Wirkung des Aloins auf den Stoffwechsel, Biochem.
Ztschr. 49:426, 1913.

6. Miller. T. L., and Liisk, F. B. : The Treatment of Arthritis by Intravenous
Injection of Foreign Protein, J. A. M. A. 66:17.V) (June 3) 1916: The Use of
Foreign Protein in the Treatment of Arthritis, 1. A. M. A. 67:2010 (Dec. 30)
1916.

BAKR ET AL.— TEMPERATURE REGULATIOX 609

investigated the physiologic changes which accompany the paroxysm.
Cecil ' has made a clinical study of forty cases treated in this hospital,
including in his series the cases published in this article. He obtained
fairly satisfactory results but did not consider that treatment with
foreign proteins was indicated until salicylates had been given a
thorough trial. Snyder* used the same New V'ork City Board of
Health typhoid vaccine, starting with small doses.

Scully ^ made a careful study of the blood cuunt, temperature and
blood pressure after intravenous injections and published composite
curves which are most instructive. He used typhoid vaccine in doses
of from 37 to 75 million bacilli with patients suffering from acute
articular rheumatism. The composite temperature curve reached a
maximum of 103.6 F. four hours after the injection and fell gradually
to normal in sixteen hours. The highest individual temperature was
106.6 F. ; the lowest 102 F. The leukocytes showed first a fall and then
a sharp rise. The average count w^s 14,000 at the time of injection.
During the chill it dropped to 5,000 but rose lo 40,000 about eight
hours after the vaccine was given. The highest individual count was
77,200 at six hours; the lowest 13,600. The blood pressure probably
rose during the chill but accurate measurements were impossible.
After the chill, the composite curve showed a fall, reaching 92 mm.
systolic and 60 mm. diastolic pressure six hours after the injection.
Following this there was a gradual rise. The lowest individual reading
was 60 mm. systolic and 40 mm. diastolic, the patient evidently suffer-
ing from marked shock. In addition to these phenomena, Scully
examined the urine but found no marked changes.

Cowie and Calhoun " have emphasized the analogy between the
nonspecific chill and the malarial paroxysm. They made numerous
blood counts and found nucleated red cells and myelocj-tes and many
atypical cell forms, particularly in the lymphocyte group. Jobling,
Petersen and their co-workers ^^ have studied in detail the ferments
after the injection of foreign protein. They have found an instan-
taneous mobilization of a large amount of nonspecific proteose;
decrease in antiferment; increase in noncoagulable nitrogen of the
serum; increase in amino-acids; and a primary decrease in scrum

7. Cecil, R. L. : A Report on Forty Cases of .^cute Arthritis Treated by
Intravenous Injections of Foreign Protein, .•\rch. Int. Med. 20:951 (Dec.) 1917.

8. Snyder, R. G. : A Clinical Report of Nonspecific Protein Therapy in the
Treatment of Arthritis. Arch. Int. Med. 22:224 (Aug.) 1918.

9. Scully, F. J.: The Reaction after Intravenous Injections of Foreign
Proteins, J. A. M. A. 69:20 (July 7) 1917.

10. Cowie, D. M., and Calhoun, H. : Nonspecific Therapy in Arthritis and
Infections, Arch. Int. Med. 29:69 (Jan.) 1919.

• 11. Jobling, J. W.: Petersen, W.. and Eggstein, A. A.: Studies on Ferment
Action, J. F.xper. M. 22:401, 568. etc., 1915; Petersen, Wm. H.: Serum Changes
Following Protein "Shock" Therapy. Arcli. Int. Med. 20:716 (Nov.) 1917.

610 ARCHIVES OF IXTERXAL MEDICIXE

proteoses. There is also an increased flow of lymph from the thoracic
duct. Later there is a progressive increase in the noncoagulable
nitrogen, in proteoses and serum lipase.

We must remember that while all these changes are taking place
in the blood and cardiovascular system, the organism is being subjected
to great variations in the degree of muscular activity, marked fluctua-
tions in the respiratory activity, sudden demands for the mobilization
of foodstuffs with increased products of katabolism, and also rapid
changes in the temperature of the body cells. These metabolic and
physical phenomena form the subject of the present investigation.

The apparatus used was the respiration calorimeter described in
the previous papers of this series and the patients were kept under
close observation in the metabolism ward. On account of the length
of the observations, it was necessary to allow them some food shortly
before the start of the experiment. All, with the exception of Albert G.,
were given a small "standard breakfast" four or five hours before
the start of the observation. It has been shown in Paper 26 ^^ of this
series that this breakfast has no effect on the metabolism, except for
two or three hours after it has been taken.

The subjects were five patients with various rheumatic affections,
one comparatively well man with lumbar and sciatic pains and one
normal control. Three of the patients were studied from the standpoint
of possible changes in metabolism which might occur in chronic
arthritis. The results of observations on their basal metabolism may
be found in the accompanying article on arthritis. At the time of
the observations there were in the general wards of the hospital a
considerable number of patients being treated with intravenous injec-
tions of protein made according to the method of Jobling or with
typhoid vaccine as prepared by the New York City Board of Health.
The other two rheumatic subjects were intelligent men selected from
among these patients. Both had previously given definite response to
injections of foreign protein. In these, it seemed possible to calculate
fairly closely the time interval between a given dose and the onset
of a chill. This was a matter of importance since the technic of
managing a calorimeter in short periods during a chill is extremely
difficult. With one subject (Genaro A.), a rise in temperature followed
the injection but the chill did not occur. Albert G., the normal control,
and R. L. C, had never been given foreign protein before; yet, neither
had the expected chill. All of the others reacted very much in the
manner predicted.

12. Soderstrom. G. F. ; Barr. D. P.. and Du Rois. E. F. : Clinical Calorimctry.
Paper 26, The Effect of a Small Breakfast on Heat Production, .\rch. Int.
Med. 21:613 (May) 1918.

Subject,

Date,

Weight,

Surlace

Area, Linear

Formula

Period

End
ol

Carbon

Dioxid,

Gm.

Oxygen,

R.Q.

Water,
Gm.

Urine X
Gm.

Indirect
Calo-

rimetry,
Cal.

Heat
Elimi-
nated,
Cal.

E.L.C.*

1/19 17. 62.3 Kg.
(Ht.-Wt.)
1.77 Sq. M.

William B

1/15/17

l^S^-^M.

Prelim.
Prelim.

11:48
1:48
11:40
12:20
12:50

46.4

24.1
39.0

20.2
32.6

0.78

0.87
0.87

22.3

o.'is

0.43

142.3
....
68.6

111.2

50.;

4U4

1:50

24.3

28.4

0.62

40.0

0.43

92.8:

889

2:50

28.6

26.8

0.78

39.5

0.43

88.8

92.5

... p

Prelim.

1:07
2:07
3:07

22.5

19.0
21.2

0.86
0.92

28.0

29.7

048
0.42

64.0
72.5

"S

■2.07 Sq. M.
Meeh Formula

69.0
74.1

4:07

25.2

23.8

0.77

30.2

042

78.6

T8.8

5:07

25.0

22.7

0.60

31.2

0.42

75,4

73.2

John Br

2/2/17
68.5 Kg.
1.75 Sq. M.
(Ht.-Wt.)

Prdim.

11:29
12:19

54.8

....
44.7

0.69

29.0

....

....
152.8

70.6

12:49

16.8

14.3

0.86

16.9

0.68

48.5

43.2

1:19

15.6

13.8

0.82

15.6

0,68

46.4

44.1

2:19

31.7

310

0.74

32.1

0,66

102.1

93.5

3:19

31.4

28.8

0.79

46.9

068

96.1

112.2

Joseph McC

.I'M.
1.46 Sq. M.

Prelim.

11:09
12:46

20.9

16.6

0.82

27.8

0^27
0.27

62:;

....

61.5

1:15
1:11

8.8
10.4

10.1 1

7.4 J

0.79

fl3.4
(l3.8

0.27 J
0.27 J

.,.

J 27.8
128.6

2:41

2S.6

27.0

0.60

32.6

0.27

90.3

• 66.1

-•ivor.

Prelim.

3:41

22.7

21.9

0.76

30.6

0.27

72.1

66.4

JOSiphMcC

2/12-13/17

.^1.6 Kg.
1.46 Sq. M.

9:28
9:58
10:20

10.7
7.8

8,9
6.2

0.87
0.91

9.8
7.7

0.16
0.16

30.3
21.1

26.3
20.6

11:00

28.1

22.4

0.92

17.6

0.16

77.1

44.0

12:00

23.3

20.7

0.82

24.5

0.16

69.7

628

l:00

27.1

?34

0.84

27.2

0.16

79.3

68.7

1:40

....

2:40

22.4

19.7

0.83

26.3

0.16

66.5

69.3

3:40

213

190

0.82

29.8

0.16

63.7

75.3

4:40

20.7

18.2

0.83

26.7

0.16

61 .3

66.8

5:30

18.8

17.0

0.81

23.9

O.IB

56.7

63.6

6:00

10.7

9.1

0.85

12.9

0.16

S0.8

34.0

• Direct lost because of leak in pipes of ice tank,
t Spirometer string broke In first and third periods,
t Calculated from R. Q. of 0 70.

—Calorimeter Data

Direct
Calo-

Rectal

Temp.,

Cal.

Aver-
/#4

Work
Adder.
Cm.

Non-

Per Cent.
Calories from

Calories
per Hour

Remarks

Pro-
tein

Fat

Carbo-
hyd.

z

(Linear)

36.9

32

0.78

()2

■■

1.13

40.2

Proteose 4% (Joblingi
0.5 c.c. at 10:0.-) a.m.
Quiet, no chill

6-4

37.3
37.6

60

34

0 88

?7

52

1.55

57.2

Proteose 4% (.loblingl

0.8 c.c. at 9:18 a.m.
Shivering last 5 min.

IM.O

3?.6

21

0.87

26

54

3.36

123.5

Chill 12:20-12:40

115.4

39.1

79

14

0.60

1.40

Fairly quiet

62.6
64.2

38.6
3«.8

76

14
13

0.77
0.90

48
27

39
53

1.34
0.96

49.3

Meeh )
30.9

Quiet

(Typhoid vaccine 50(i

1 million- subcutaneous-

I ly 10:07 a. m.

Quiet

90.5

37.0

. 60

28

0.94

17

68

1.09

35.0

Quiet

77.3

37.0

60

24

0.76

70

16

I.IS

37 9

Rather restless

74.4
139.6

37.0
37.8
39.2

28
25

0.60

0.90

58
30

27
60

1.13
3.14

I 36.4 J
103.0

Bahter restless
[Typhoid vaccine 25 mil-
i lion intravenously at
I 10:50 a. m.

Chill, 30 min.

76.1

39.9

10

0.87

35

46

1.64

54.8

Quiet, voided

38.2

39.8

7

0.82

18

32

1.5S

52.6

Quiet

87.6

39.7

20

0.73

76

0

1.73

57.7

Quiet

86.0

39.2
37.5

98

33

0.79

bO

21

1.63

54.5

Somewhat restless
Basal

60.3

20.4
£9.4

37.5
37.5
37.3
37.3

64

58
56

23

12 1
8 J

J.32
0.79

1.21
1.24

42.7
40.7

Very quiet

[Removed from calorim-
! eter, t5T)hoid vaccine
1 40 million intravenous-
1 ly at 12:31 p.m.
Quiet

1120

38.4

12

0.79

1.76

57.9

Chill. 1:4.5 2:15 p.m.

79.0

38.8

0.75

1.40

46.2

Very quiet

ISO

37.2
37.1

8

65

37

25

1.18

41.5

(Typhoid v.-iccine 50 mil-
! lion intravenously at
1 9:05 p.m.
Quiet

20.5

37.1

2

0.90

30

62

1.11

39.5

Very quiet

92.8

38.2

89

7

0.92

26

70

2.24

79.3

Chill. 10:21-10:53 p. m.

70.2

38.4

120

0.82

58

36

1.35

47.7

Quiet

7S.3
57.4

38.0
38.3
38.2

79

20
9

0.84
0.83

52
56

43
38

1.29

54.3
45.5

Fairly quiet
Removed from calorim
; eter, given 240 c.<
' water at 1:05 (ST")
Very quiet

62.2

37.9

7

0.82

57

36

1.23

43.6

Asleep

62.0

37.8

88

4

0.83

bi

40

1.19

42.0

Very quiet

59.9

37.7

13

0.81

(■2

32

1.32

46.4

Very quiet

f.(>.»

.■57.6

„

n»(;

M

42

1.20

42 2

Restless last 5 min.

ARCHIJ-ES OF IXTERXAL MEDICI XE

TABLE 1.— Calorimeter-

Subject,

Date.

Weight,

Surface

Area, Linear

Formula

Period

Dnd

of

Period,

Time

Carbon

Dioxid,

Gm.

Oxjgen.

R. Q.

Water,
Gm.

Urine N

Indirect
Calo-

rimetry,
Cal.

Heat
Elimi-

Prank G.-t

1.66 Sq.M.

Prelim.

11:15

2

11:45
12:45

11.0
40.9

9.9t
35.6

O.Sl
0.84

37.7

0.29
0.29

33.2
120.1

32.8
68.6

3

1:45

26.0

23.lt

0.82

3S.5

0.29

77.4

74.0

4

2:45

25.2

24.4

0.75

51.0

0.29

80.3

90.6

5

3:53

24.1

27.5

0.64

73.8

0.29

89.7

110.1

3/2/17
52.9 Kg.
1.56 Sq.M.

1

11:45

12.1

9.9

0.89

12.9

0.68

33.3

32.1

2

12:15

13.2

11.9

0.80

15.6

0.68

39.4

37.7

3

l;15

25.9

23.5

0.80

32.3

0.68

77.7

68.7

4

2:i5

26.2

25.0

0.76

33.8

0.68

82.0

71.1

5

3:l5

2r.2

26.2

0.76

36.5

.,

85.8

78.7

• Direct lost because of leak m pipes of ice tank,
■t Spirometer string broke in first and third periods.
J Calculated from R. Q. of 0.70.

CASE HISTORIES

Case 1.— R. L. C, a physician, born in the United States. 35 years of age,
has never been seriously ill, has had no attacks of tonsillitis or of articular
rheumatism. During the winter of the past three or four years he has had
lumbar and sciatic pains, which, at times, have entirely incapacitated him. At
the time of the observation he was in the midst of a particularly uncomfortable
attack.

Physical Examination.— "Sioihm^ abnormal. Tonsils are small and healthy in
appearance. The teeth are in good condition. No areas of tenderness are found
in the back or in the region of the sciatic nerve.

Jan. 19, 1917, he was given intravenously 0.5 c.c. of a 4 per cent, solution
of Jobling's proteose. No rise in temperature or disagreeable symptoms
followed the injection. No improvement of pain or stiffness resulted.

The calorimeter observation was unsatisfactory because there was no chill.
A leak in the pipe made it impossible to use the method of direct calorimetry
in this experiment.

Case 2.— William B., arthritis deformans (gonorrheal?), a traveling sales-
man, born in the United States, SO years of age, was admitted Oct. 16, 1916, and
discharged improved Jan. 17. 1917. In 1899 he had gonorrhea followed by
epididymitis. The discharge lasted for two weeks. He had had occasional sore
throat but no previous attacks of joint pain.

July 21, 1916, he was seized with pain in the left arm and wrist and in the
right ankle. Two days later he went to St. Vincent's Hospital where he was
treated until Oct. 16.

Physical Examination.— Y{e is a poorly nourished, well developed rather
apathetic man. He has no temperature and is not toxic. The right elbow
is ankylosed and is very painful. .Ml of the muscles of the right arm and
forearm are flabby and wasted. The fingers of the right hand are stiff from
disuse. The right shoulder, left elbow and right ankle are very tnoderatcly
involved. There is superficial edema of the feet and ankles.

BARR ET AL— TEMPERATURE REGULATIOX 615

-Data— (Continued)

Direct
Calo-
rimetry
(Rectal

Rectal
Temp.,

Aver-
I^?e

Work
Adder,
Cm.

Non-
protein
R.Q.

Per Cent.
Calories from

Calories
per Hour

Remarks

Pro-
tein

Fat

S,Vd.°-

B'

Per
Sq. M.
(Linear)

29.8
150.7
76.5

37.5
37.4
39.1
39.2

84
96
96

7
19

0.81
084
0.S2

10
10

58

51
54

32
43

1.18
2.12
1.37

42.6
77.0
49.6

ITyphoid vaccine 35 mil-
•j lion intravenously at
1 10:51 a. m.

(ChiU,'ll:.58 a.m. -12:25
i p.m.; drank 240 c.c.
1 water (37 C.1 12:37 p.m.
Restless for 50 min.

75.2

389

6

0.75

10

77

13

1.42

51.5

Restless last 10 min.

93.0
32.0

38.B
37.6
37.6

66

18

1

0.62
0.92

27

20

1.58
1.26

57.5
42.7

fTyphoid vaccine 20 mil-
] lion intravenously at
1 10:50 a.m.
Very quiet

40.4

37.7

5

0.80

23

44

33

1.50

50.5

Almost motionless

85.0

38.1

■■

3

0.80

23

52

25

1.47

49.8

Almost motionless

S0.3

38.3

74

3

0.75

22

68

10

1.55

52.6

Almost motionless

7G.5

.38.3

2

0.74

21

70

9

1.62

■WO

.Almost motionless

November 27 he had an acute attack of follicular tonsillitis from which he
recovered rapidly. From December 19 to January IS he received at rather
irregular intervals eight doses of Jobling's proteose intravenously. From the
first dose he had no reaction. Both the first and second injections v\rere of a 1
per cent, solution : the others were of a 4 per cent, solution, varying in dose from
0.3 c.c. to 1.0 c.c.

January 10 and January IS he was observed in the calorimeter following
proteose injections. Owing to technical errors, the observation of January
10 was lost. That taken on the fifteenth is here presented. No marked change
in symptoms was noted after any of the proteose injections. The joint condition
improved very gradually. With constant baking and massage he regained
motion in the fingers of the right hand. The elbow was still ankylosed at
the time of discharge, January 17.

Case 3. — Albert G.. a normal control to whom typhoid vaccine was given
subcutaneously. a laborer born in Italy. 24 years of age. was admitted Dec. 14.
1914. and discharged Jan. 14. 1915. His health was excellent. He was out of
work and was admitted to the hospital to act as a normal control for
other observations which were being carried on at the time. He was short,
with large muscles and very little subcutaneous fat. He was neurasthenic,
continually fearing that he would become ill.

January 13, at 10:07 a. m., he was given 500 million dead typhoid bacilli
(New York City Board of Health vaccine) subcutaneously into the arm. The
usual reaction occurred with moderate swelling and tenderness of the arm."

Case 4.— John Br., acute arthritis, gonorrheal (rheumatic?), a clerk born in
the United States, 19 years of age, was admitted Jan. 25, 1917, and discharged

1.3. For details of previous observations on this man, consult Soderstrom,
G. F.; Meyer, A. L., and Du Bois, E. F.: Qinical Calorimetry, Paper 11, A
Comparison of the Metabolism of Men Flat in Bed and Sitting in a Steamer
Chair, Arch. Int. Med. 17:872 (Tuly) 1916; Gephart. F. C, and Du Bois. E. F. :
Clinical Calorimetry, Paper 13, The Basal Metabolism of Normal Adults with
Special Reference to Surface Area. Arch. Int. Med. 17:902 (June) 1916.

616 ARCHirES OF IXTERXAL MEDICIXE

improved March 23, 1917. January 7 he began to have a urethral discharge.
Six days later he was seized with pain in the right knee, right foot, left
foot and right thumb.

Physical Examination. — Patient was a well developed, somewhat emaciated
boy, acutely ill, rather toxic. The tongue was slightly dry, with a brown coat.
The left knee joint is distended with fluid, hot. slightly tender and held in
semiflexion. The right knee, both ankles and some of the small joints of the
hands and feet show swelling and moderate tenderness. There is a urethral
discharge containing gonococci in large numbers.

January 29 and January 31 he was observed in the calorimeter, February
2 he was given intravenously a 25 million dose of New York City Board of
Health typhoid vaccine and was observed in the calorimeter. On the fourth
and again on the fourteenth he received typhoid vaccine. On March 1 he was
discharged to the general ward, very slightly improved by the vaccine therapy.
He was later given gonococcus vaccine intravenously. Following this his
improvement was rapid. He was discharged from the hospital March 23 with
slight swelling in his right knee joint but with no other symptoms.

Case 5. — Joseph McC, subacute gonorrheal arthritis, an elevator operator,
born in the United States. 27 years of age, was admitted Jan. 24, 1917, and
discharged unimproved. In 1910 he had an attack of gonorrheal urethritis
followed by epididymitis. Since then he has had a urethral discharge several
times, the last time being in November, 1916. In 1912 he had an acute
arthritis involving all the joints and he was ill for nine weeks. A second
attack of the same character, in 1914, lasted two months. He has never had
sore throat, chorea or other manifestations of acute rheumatic fever.

Jan. 3, 1917, he began to have a dull pain in the lower end of the spine and
in the lumbosacral muscles which radiated down the right thigh. His right
heel became so painful that he could not walk. About the same time the
urethral discharge recommenced.

Physical Examination. — This shows an under developed, poorly nourished
man. He holds himself very stiffly because of pain in the back. There is pain
on pressure over the sacrococcygeal joint and marked tenderness over the spine
of the ninth dorsal vertebra. The under surface of the right heel is exquisitely
tender. There is considerable watery purulent urethral discharge, a smear
of which shows gonococci. The prostate is moderately enlarged. The right
seminal vesicle is palpable.

Urine from both anterior and posterior urethra contains pus but no
albumin or casts. Gonococcus fixation test is negative. Wassermann reaction
is negative. Roentgen ray shows periosteal exostosis of the os calcis of both heels.

His symptoms remained unchanged during the first seven days ' in the
hospital. February 1 he was given intravenously 40 million of the New York
City Board of Health typhoid vaccine. On the third, the fifth and the seventh
similar doses were given. The character of the reactions following the
injections appears on the temperature chart. The chills be.gan from thirty
to seventy-five minutes after the injection. On the thirteenth a dose of 50
million was given. On the seventh and thirteenth the patient was observed
in the calorimeter. No noticeable improvement resulted from the vaccine
therapy. On the first day following each injection he was more stiff and
uncomfortable. On the second day the condition returned to that which
obtained before the vaccine was given. He was discharged improved.

C.ASF. 6.— Frank G., chronic gonococcus arthritis, a barber, born in the United
States. 42 years of age, a widower, was admitted Feb. 13, 1917. to the service of
Dr. C. E. Nammack, transferred to the Metabolism Ward and discharged
unimproved May 8, 1917. He says he has had gonorrhea seven times, the first
attack being at the age of 15. He has had three distinct attacks of arthritis,
all of which had occurred during or immediately following an acute artl.ritis.

BARR ET AL.— TEMPERATURE RECULATIOX 617

During the past year the joint pains have been almost constant. Since January.
1917, he has had a urethral discharge and more severe joint involvement.
He has always used alcohol to excess. (During the past month he has
consumed as much as a quart of whisky a day). February 16, in another
ward, he received intravenously a 20 million dose of New York City Board
of Health typhoid vaccine. February 19 he received 30 million. The joint
condition was not improved but the urethral discharge, which had been profuse,
was checked. He is a poorly nourished, fairly well developed, very dissipated
looking man. He has no temperature elevation and does not appear ;o be
toxic. His throat is congested. The tonsils are normal in appearance and
his teeth are in fair condition. His spleen is felt two finger breadths below
the costal margin. He has moderate pain, tenderness and swelling in the right
wrist, right hand and right knee. His prostate is enlarged. Massage of this
organ causes the discharge of a drop or two of thin, purulent material which
contains large numbers of gonococci.

His gonococcus iixation test is strongly positive. W'assermann is negative.

February 27 he was transferred to the general service. .March 10 the urethral
discharge again became profuse. March 27 he developed a severe gonorrheal
conjunctivitis. He had entirely recovered from this at the time of his discbarge
May 8. The joint condition, however, was unimproved.

Case 7. — Genaro, A., acute rheumatic fever, a munition factory worker,
born in Cuba, 22 years of age, was admitted to the hospital Feb. 25, 1917, and
discharged improved March 16, 1917. He says that he has never been ill before.
He denies gonorrhea and syphilis.

Since February he has had pain in both knees, wrists, shoulders, elliows
and ankles. With it. he has had a slight sore throat, some headache, and at
the onset of the illness several nose bleeds. March 26 he received intravenously
in another ward a 60 million dose of Xew York City Board of Health typhoid
vaccine; March 27, SO million; and March 28, 50 million. He had severe
reactions in each case, .^fter the first injection there was considerable clinical
improvement. The other doses had little or no effect.

Physical Exaiiiination. — He is a well nourished and developed Cuban boy
of remarkabh- sanguine disposition in spite of considerable pain. He is not
toxic. The tongue is moist. His tonsils are small and not inflamed. The
teeth show many fillings and gold crowns. .'Xt the time of admission his heart
was normal but later there developed a soft, blowing, systolic murmur, maximum
at the ape.x and transmitted outward into the a.xilla. The spleen is felt one
fingerbreadth below the costal margin. The left wrist, elbow and shoulder are
painful on motion.

March 2, after receiving a 20 million dose of typhoid vaccine, he was observed
in the calorimeter. He had no chill. On the third, fourth and sixth he received
a dose of 40 millions. Each time he had a chill with severe reaction. Th?
joint pains gradually improved. Improvement, however, seemed to bear no
definite relation to treatment. .After the last injection his temperature reached
normal and remained so until his discharge IMarch 16. .\t that time all joint
pain and swelling had disappeared but the heart murnnir persisted.

In a'l. eight observations were made, lasting from two to eisjlit and
a half hours. When changes in the metaljolism were expected tlie
periods were made as short as possible. One period of only twenty-two
minutes was obtained. It would have been interesting to subdivide
the period of chill but no experimental period can be ended unless the
subject has been quiet for six or seven minutes. Short periods are
not as accurate as long ones since a small error in determining the

ARCHIVES OF IXTERXAL MEDICIXE

"lO °^ °ro °oJ "— °o "CD ''oo
0000000505

BAKR ET AL.— TEMPERATURE REGULATIOX 619

residual carbon dioxid or oxygen will cause a larger percentage change
in the shorter period. This error will, of course, be compensated for
in the next period.

The time of the chill never corresponds exactly to the experimental
period in which it is observed. It is possible, however, to calculate
with fair accuracy the heat production occurring during the chill itself.
We may consider that the metabolism during the few minutes before the
chill is at the same level as that of the preceding period. Similarly
the metabolism in the short interval after the paroxysm approximates
the level of the following period. The heat thus calculated for the
interval before and after the chill is subtracted from the total heat
produced during the entire experimental period. In this manner,
the heat production during a twenty minute chill can be estimated even
though the experimental period be forty minutes long.

The data of the calorimetric experiments are given in Table 1.
Figure 3 shows graphically the results of the calorimeter observation
on John B. Figures 4 and 5 show the results on Joseph McC.
(Case 5). Figure 6 on Frank G. (Case 6) ; Figure 7 on William B.
(Case 2) ; Figure 8 on Genaro A. (Case 7).

DISCUSSICX OF RESULTS

The phenomena observed after the intravenous injection of foreign
protein are almost identical with those observed in malaria.^ For
convenience of discussion the malarial paroxysm was divided into
six periods: (a) a basal period before the chill; (b) a prodromal
phase immediately before the chill; (c) the chill itself; (d) a period
of rising temperature after the cessation of shivering; (e) a period
of high continuous temperature corresponding to the clinical stage of
heat, and, finally, (f) a period of falling temperature. In considering
the reactions to foreign proteins, this same division is useful.

In the chills of malaria, the respiratory quotients were found to be
high. The same thing is observed during the chill following the
injection of vaccines, which indicates a rapid combustion of the glycogen
stores of the body. In most of the experiments, the quotient falls
steadily after the chill.

Four main questions of the mechanism of the rise anrl fall of
body temperature will be considered.

1. The relation of heat production to heat elimination as factors
in the rise and fall of temperature.

2. The divergence of the rectal temperature from the average
body temperature.

3. The relation of heat lost in tlie vaporization of water to the
total heat elimination and to the heat production.

4. The influence of body temperature on heat production.

The data on which the discussion is based are presented in Table 2.

0.2 g? »

ss?i?; 8 SS8S ssss s ss ssj;

HI ii'-

itll ?i ?++ 1ii1 t\ +1

nit

11

mm

2

iSi

s

"!'-:

gs

iii

a

ills

mu

§iS

SEii

s

p

isi

spi

oodo

d

ill

SI5S

i

ii

Sis

5SSS

iSIs S las S|„" 2 =S

2SS &Szi

1 M? I

5 •£

,.a l^a lj.5 ^^a a,,a ;,

BARR ET AL.— TEMPERATURE REGULATION 621

Relation of Heat Production to Heat Elimination.— In malaria, it
was found that during a chill the heat production was enormously
increased while the heat elimination remained practically at its basal
level. After injection of foreign protein, the same mechanism is
observed. It would seem as if the temperature regulation were set
at a higher level and that the body responded by producing heat
sufficient to warm the tissues to the new temperature level, .\ftcr

0)

0^

(T>

f\J

^

f\J

CVJ

0>

40

38

37

150
CALSI

100

50

CHILL

INDIRECT CAL.
PIRECJ.QAL,

PER HOUR
HEAT ELIM

L.

.80

70

AV.BODY TEMP.

VAPORIIAi ION CAL.

1 — r

"BASTfU^

Fig. 3.— Calorimeter observation on John Br., February 2. At 10:o3 a. m.,
25 million typhoid bacilli were given intravenously. The chill lasted from 11 ;40
a. m. to 12 : 10 p. m. The upper curve shows the rectal temperature measured
every four minutes ; the dash line below shows the change in the average body
temperature, as calculated from the difference between heat elimination and heat
production as determined by the method of indirect calorimetry. This line
is started arbitrarily 0.5 C. lower than the rectal temperature. While it is
possible to determine the rise and fall of the average body temperature, it is
absolutely impossible to fix the exact level at which these fluctuations take
place, .^fter the start at 11:29 a. m., the first satisfactory fixation of this Ime
was at 12:49 p. m. Between these points the exact shape of the curve is not
certain. The respiratorv quotient ( R. Q.) is indicated by short lines in each
period. The line at 80.8 calories represents the basal heat production per
hour as determined on January 31.

622 ARCHU'ES OF IXTERXAL MEDICINE

vaccine, the heat production is increased from 75 to 210 per cent,
during the chill, while the amount of heat eliminated is scarcely
increased above its former basal level. This, of course, results in the
storage of large amounts of heat within the body. After the chill
is over, the heat production drops sharply but still remains somewhat
above the normal, as is usually the case during increased body temper-
ature. The rectal temperature still continues to rise after the shivering
ceases. The heat elimination is practically unchanged. During the

CAL.PER HOUR.
100

Fig. 4. — Calorimeter observations on Joseph McC, February 7. The first
period from WW a. n\.. to 12;09 p. m., was a basal determination. .'\t 12:31
p. m. he was given an intravenous injection of 40 million typhoid bacilli. Two
short periods were obtained before the chill, which lasted from 1:45 to 2:15
p. ni. Note that the average body temperature rises more slowly than the
rectal temperature.

stage of high continuous temperature the heat elimination increases
until it is equal to the heat production ; both, however, being at a
level from 20 to 40 per cent, above the normal. During the fall in
body temperature, the heat production drops gradually to the basal
level and the heat elimination increases steadily.

Relation of Rectal to Average Body Temperature. — During the
j)ast ten years, more than 300 observations on patients with normal

Fig. S. — Calorimeter obser^
from 10:21 to 10; 53 p.m. A(
1 : 40 a. m. Note that the cui

ation on Toseoh McC February 12 and 13. At 9: OS p. ra. he was given 50 million typhoid bacilli intravenously. Ihc cln 1 astc.
1 00 am the calorimeter was'opened to give the patient a drink of water. It was closed immediately and the ne.xt per.od sta.tcd at
ve for the average body temperature lags behind that oi the rectal temperature.

622 ARCHIVES OF IXTERXAL MEDICINE

vaccine, the heat production is increased from 75 to 210 per cent.

past ten years, more than 300 observations on patients with normal

BARR ET AL.— TEMPERATURE REGULATIOX 623

temi)erature have been made with the Sage calorimeter. They have
shown a remarkably close agreement between the direct and indirect
methods of measuring heat production. In patients with fever,
however, and particularly in malaria, there has been a wide divergence
between the two methods. This may be explained on the following
basis. The measurement of heat production by direct calorimetry
depends for one of its factors on the rectal temperature which in the

CAL. PER HOUR. DIRECT CAL.
ISO

100

50

i' HEAT ELImJU^.^.^-^.^

VAPORIZATION CAL.

INDIRECT CAL1

Fig 6.— Calorimeter observation on Frank G.. February 23. At 1U:51 a. in ^J
million tvphoid bacilli were given intravenously. The chill lasted frorn 11:3b
a. m. to 'l2 :25 p. m. The average body temperature rose less sharply than the
rectal temperature.

calculation is assumed to represent accurately the average temperature
of the body. During rapid production of heat, such as is seen during
a shivering chill, the distribution of heat will not be immediately
uniform, with the result that the temperature in the rectum may
change more or less than that of the rest of the body.

624 ARCHIVES OF JXTERXAL MEDICIXE

Exact measurement of average body temperature appears impossible
since we cannot have thermometers in all parts of the body. In the
paper on malaria, however, an indirect method was devised which
allows ii> to calculate this value. For a given interval the difference

37

36

CAL.
?00

150
100
50

CHILL

AV. BODY TEMP.

PE

INDIRECT CAL.
R HOUR.

DIRECT CAL.

HEAT ELINU

.VAPORIZATION CAL.

Fig 7.— Calorimeter olservation on William ?.., January 15. .At 9:40 a. ni.
he was given an intravenous injection of 0.8 c.c. of a 4 per cent, proteose
solution. He shivered during the last five minutes of the first period and had
a violent chill from 12:20 to 12:40 p. ra. In this case the average body
temperature seems to liave risen more sliarply than the rectal temperature.

is taken between the number of calories produced, as estimated by the
chemical methods of indirect calorimetry and the calories eliminated
as measured by the direct physical methods. This represents the heat
lost or gained by the whole body. AMien this difference is divided

■AKK ET AL.— TEMPERATURE REGLLATIOX

625

by the hvdrothermal equivalent of the organism, the gain or loss in
average body temperature is obtained. For example, heat production
100 calories, minus heat elimination 60 calories=heat stored 40
calories. In a man weighing 70 kg., the hydrothermal equivalent (heat
necessary to raise the temperature of the whole body 1 C.)^.t-^.1
calories." The gain in average body temperature is, therefore, ■^^ =
0.66 C. This may be compared with the change in rectal temperature
which has been observed during the correspondmg time interval.

In Figures 3 to 8, the average body temperature, as calculated
by this method, is plotted with the rectal temperature. In about half
of the cases, the two methods agree during the chill. In the others.

8. — Calorimeter observation on Genaro .A.. March 2. This patient
received a small dose of 20 million typhoid bacilli intravenously at 10:50 ... m.
He did not have a chill. There was only a slight rise in rectal temperature
and even less of a rise in the average body temperature.

the finding is similar to that observed in malaria. The rectal
temperature rises more rapidly than the average body temperature.
During the period of falling temperature, the heat measurements in
the rectum represent with fair accuracy the conditions of the entire
body. In the previous discussion of malaria, u was considered that
the more rapid rise of rectal temperature indicated a .storing of heat
in the deeper portions of the body during the chill. In the light of
these further observations, however, this cannot be regarded as a con-

14. This value is obtained by multiplying the weight in kg. by the facte
0.83 which is assumed to represent the specitic heat of the liody.

626 ARCHIVES OF IXTERXAL MEDICIXE

slant finding. We can only say that, during rapid varations in the
production and elimination of heat, the distribution is not always
uniform and that the rectal temperature is often an inaccurate index
of the average temperature of the body.

TIic Relation of Heat Lost in the Vaporization of IVntcr to tlie
Total Heat Elimination and to the Heat Production. — Under the
constant temperature conditions of the calorimeter, normal individuals
accomplish about 25 per cent, of their total heat elimination through
the vaporization of water. The same relationship was found in a
large number of obsen-ations made in afebrile pathologic conditions.
\\'ith continuous normal temperature, heat elimination and heat pro-
duction are equal so the calories lost in the vaporization of water also
constitutes 25 per cent, of the heat production. During the sharp
rise in temperature accompanying chill, we have seen that the heat
production is greatly increased while the elimination remains at its
normal level. It is conceivable that the heat lost in vaporization might
follow either one. In malarial fever, it was found to bear a fairly
constant relation to heat elimination. It never followed the cur\-e of
rapid fluctuation in heat production. In the present experiments, the
same relation is observed (Table 2). During the chill, the calories
lost in vaporization constitute a very low per cent, of the heat produc-
tion. During the fall in temperature, the percentage may be high-
The relation to heat elimination, on the other hand, never deviates far
from the limits of normal. This is well illustrated in the experiment
on John Br. where the heat lost in the vaporization of water varies
between 11 and 28.8 per cent, of the heat production while the relation
to heat elimination varies only between 20.4 and 24.5 per cent.

The ability of the body to utilize water in heat elimination is of
the greatest importance in the regulation of temperature. Because
of its high specific, heat and fluid character, it is capable of absorbing
large amounts of heat from the cells and of distributing it to the surface
of the body where it may be eliminated. The dry skin and concentrated
urine of acute fevers has often given rise to the belief that the
mechanism of heat loss is disturbed.

During exercise, in exophthalmic goiter and in other conditions
in which the heat production is increased, the heat elimination increases
correspondingly. Wolpert,^^ Benedict and Carpenter ^^ and others
have shown that during exercise, the heat lost in vaporization is
increased not only absolutelv but relatively to the total heat elimination.

15. Wolpert : Ueber den Einfluss dcr Lufttempcratur auf die ini Zustaiid
anstrengender korperlicher .'\rbcit ausReschiedeiien Mengen Kolilensaure und
Wasserdampf beim Menschen. Arch. f. Hyg. 26:.12, 1916.

16. Benedict, F. G., and Carpenter, T. M. : The Metabolism and Energ\-
Transformations of Healthy Man Dnring Rest. Carnegie Institute of Washing-
ton. Pub. 126. 1910.

BARK ET AL.— TEMPERATURE REGULATIOX 627

Thus, in severe work involving a heat elimination four times the normal
amount Benedict and Carpenter found that the calories lost in vaporiza-
tion constituted 47.6 per cent, of the total. Such findings have been
compared with those occurring during fever in which a relative increase
in the vaporization calories does not always occur. This is one of the
factors which has led Balcar, Sansum and Woodyatt *" to consider
that fever may be due to an inability of the body to mobilize its water
reserve for the elimination of heat. It has also been considered an
argument for the supposed concentration of blood during fever.

It must be remembered that the rise in heat production involved
in even light muscular work greatly exceeds the increase of heat
production in fever. Except under conditions of chill, it is unusual
to find the metabolism more than 50 or 60 per cent, above the normal

TABLE 3. — Comparison Between Exophth.^lmic Goiter .\xd Typhoid Fevi

IX THE PeRCEXT.\GE OF ToT.\L HeAT ELIMINATION LoST IN THE

Vaporization of Water

Heat
Production,
per Cent.
Rise Above
Average
Normal
Basal

Per Cent.
Total Heat
Elimination
Lost in
Vaporiza-
tion ol
Water

Relative
Humidity

Subject

Diagnosis

Date

Begin-
ning

End-
ing

10/29/13
3/10/15

11/15/13
4/-24/U
3/23/U

10/23/14

+41
+39

+37
+53
+39

+2!

24

25

23
20

23

49
39

37
46

41
49

Charles F

Pnitpr

-,r

Edward B...
Edwin T

Typhoid

4S

level during the course of a fe\er. It may be more in keeping, there-
fore, to compare water elimination in fever with some other conditions
involving about the same increase in heat production. In Table 3 will
be found a comparison of conditions in cases of exophthalmic goiter
studied in Paper 14 ^' with those in cases of typhoid fever from Paper
7 ^' of this series. The temperature of the calorimeter was between
23 and 24 C. in all observations. Because the conditions or relative
humidity were not always uniform the table is arranged in three
groups of two each, showing respectively constant, rising and falling
humidity.

17. Balcar, J. O.: Sansum, W. D.. and Woodyatt. R. T. : Fever and Water
Reserve of the Body. Arch. Int. Med. 24:116 CJu'.y) 1919.

18. Du Bois, E. F. : Clinical Calorimetry, Paper 14, Metabolism in Exoph-
thalmic Goiter, Arch. Int. Med. 17:915 (June) 1916.

19. Coleman, W., and Du Bois, E. F. : Clinical Calorimetry, Paper 7. Calori-
metric Observations on the Metabolism of Tvphoid Patients with and Without
Food, Arch. Int. Med. 15:887 (June) 1915.

628 ARCHIVES OF l.XTERXAL MEDICIXE

It is to be noted that all percentages in both conditions are close
to the average normal value. Max W. with a metabolism higher than
any of the fevered typhoid patients lost only 20 per cent, of his
calories in the heat of vaporization. The comparison is more striking
when one remembers the notoriously moist skin of goiter patients.

In the study of typhoid fever, it was found -" that during a rising
temperature, the percentage of heat lost in vaporization of water was
slightly lower than normal. The average was 22 per cent, which must
be compared with the average normal of 24 per cent. The variations
in the typhoid group, however, were great and the number of cases
small. In spite of this, the finding has been considered by others as
evidence of the body's inability to mobilize water and hence as a cause
of fever. That this is not a constant finding during the rising temper-
ature of fever can be shown in the case of George S., a tnalaria patient.
wh(i?e record is shown in Table 4.

TABLE 4.— RELATION of He.\t V,-\poriz.\tion to Tot.al Hr:.\T Elimin.mion in
George S., Malari.^l Fever

Condition

Per Cent. Rise

in Heat Production

.\bove Average

Normal Basal

Percent. Total
Heat Elimination
Lost in Vaporiza-
tion of Water

Rising temp<r.it ,i

;;;;^,;,:,';;,',;',';,

+.24

28

Rising temp. r;. I
Highcontinu.H,.
Falling temr.iiit.

r chill

ting

'.'.'.'.'.'.'. + iO

:'9
36+

It will be seen that the percentage of heat lost in vaporization is
well above the average ncirmal imt only during the rising temperature
but also during the chill and following periods. The same thing is
shown even more strikingly in the case of Frank G in Table 2.
Similar results were obtained in all observations in erysipelas.

Our data present no evidence that the rise of temperature is
dependent on an inability of the body to mobilize its water reserve.
It is true that the heat loss in vaporization of water is low when
compared with the enormously increased heat production of a chill,
but it never falls far below its average normal relation to the total
elimination and as in the two cases just cited may rise considerably
above it.

The Influence of Body Teniperatiire on Heat Production.— Ju
discussing the conditions in the ]3eriod following a chill, it was empha-
sized that the heat production remained from 20 to 40 per cent, above
the normal level after shivering had ceased. It was also remarked

20. Soder.'^troni, G. F.. nnd Du Rois. E. F. : Clinical Calorimetry, Paper 25.
The Water Elimination Through Skin and Respiratory Passages in Health
ar<l Disease. .Arch. Int. :\Ie(l. 19:Q31 (Jnne) 1917.

TEMR C
40°

39

38

37

36

• y^

^

• ^

<^.

'

;>

^•

INTRAVENOUS
FOREIGN PROTEIN

1 1

90

100% 110

120

130 140 150

l.-ig_ 9.— Relationship of basal metabolism to temperature in the lever follow-
ing the injection of foreign protein. Ordinates represent rectal temperature in
degrees centigrade, abscissae the metabolism expressed in percentages of the
average normal. Each dot represents an experimental period in the calorimeter.

41

40

39

38

37

36

c

//

' •

•

•^

^

/

MAL/

>RIAL

EVER

y

^ '

90

100^ 110

■1. lu— Relation

120 130

140 150

to temperature

160

170 180

630 ARCHIVES OF IXTERXAL MEDICIXE

that this was the usual finding in conditions of increased body temper-
ature. Work on typhoid fever, tuberculosis and malaria has
demonstrated that rise in body temperature is accompanied by increased
heat production and that the increase corresponds to the degree of
fever. Figure 9 shows the relation of heat production to the degree
of body temperature after the injection of foreign protein expressed
bv the method utilized by McCann and Barr "^ in tuberculosis. The
ordinates show the rectal temperature in degrees Centigrade. The
abscissae show the level of the metabolism in percentage of the average
normal. The line 90 means 10 per cent, below the average; 150 means
50 per cent, above the average. Each dot represents a calorimetric
observation. Of course, it is necessary to leave out the shivering
periods. It is possible that some of the results in the very high tem-
perature following the chill are slightly afi'ected by previous severe
muscular exercise. Figure 10 demonstrates the relationships during
the fever of malaria. In Figure 11 are grouped in one chart all of
the fevers studied. The continued diagonal line is drawn from
statistical calculations and the dotted lines are placed to represent
divergencies of 10 per cent, in this average. Out of the total of 137
experiments in various fevers, 82 per cent, come within 10 per cent,
of the average. In other words, the percentage variations in the
metabolism for a given temperature are slightly greater than a similar
group of normal individuals.

Most of the patients whose metabolism is very high for the degree
of temperature were typhoid or malaria patients with a high level
of protein metabolism. Most of those with low basal metabolism
were cases of tuberculosis with a normal protein metabolism. We
know that protein increases the metabolism through its specific dynamic
action and this may explain the difference between the groups of
patients. The ingestion of a large protein meal does not increase the
heat production in typhoid where the protein metabolism is already
high, but it does cause a striking increase in tuberculosis where the
protein metabolism is at a much lower level. While the increased
protein metabolism seems to be a factor in explaining differences
between the various fevers we believe that it is outweighed by another
and simpler factor.

The surprising uniformity of results expressed in Figure 11,
suggests that we are dealing with a law of the velocity of chemical
reactions enunciated by van't Hoff." For ordinary temperatures
the van't Hoff law can be expressed as follows : "With a rise in

21. McCann. W. S., and Barr, D. P.: Clinical Calorimetry, Paper 29. The
:Metabolism in Tuberculosis, .-Xrch. Int. Med. 26:663 (Nov.) 1920,

22. van't Hoff. J. H. : Studies in Clicmical Dynamics. Revised ]iy E. Cohen,
Translated by T. Ewaif. Easton. Pa. Chemical Publishing Co.. 1896.

B-ARR ET AL.— TEMPERATURE REGULATION

ARCHirES OF IXTERXAL MEDIC

temperature of 10 C. the \elocity of chemical reactions increases
between two and three times." In other words, tlie temperature coef-
ficient is usually between 2 and 3. This means an increase of from
30 to 60 per cent, for the three degree rise from 37 C. to 40 C.
Virtually all of the fever experiments are withm these limits and the
average line shows a temperature coefficient of 2.3.

Van't Hoff and Kanitz -= give the temperature coefficients which
show the rate of increase in a number of chemical reactions with an
equal rise in teniperature. If we plot these in e.^actly the style of the

TEMP
41"

AO

39

38

37

36

till

TEMP. COEFFICIENT.

c\i 1 c^ 1 CO \ ^

/

//

^^

^

^

^

A

^

^^

^^

/^
/^^

^^p

90 100 no 120 130 140
RATF. OF CHEMICAL REACTIONS

150 1 60

170

rcaction.s taken from van

tliis chart represent a nmnlier of typical chemical
Hoff and Kanitz. The slope of the lines shows the
increase in the rate of the reactions as the temperature is raised. Note that
the lines correspond closely to those which represent the total oxidations i:i
the human body.

fever patients (P^ig. 12) we note that the lines have approximately
the same slope. In other words, the reactions in a fever patient respond
to a rise in temperature in a manner which resembles closely the
chemical reactions in the test tube suspended in a water bath.

There is a tremendous gap between tlie simple reactions in the
test tube and the complex oxidations in the diseased human body and
we should hesitate to compare them were it not for the large number

23. Kantiz. .A.: Tcmperatur und I.ehensvorgan.ue in Biochcr
stellungen: Heft 1. Gebriider Borntrager, 1915, Berlin.

in Einzeldar-

BARR ET AL.-TEMPERATURE REGULATIOX 633

of biologic reactions which show temperature coefficients between
2 and 3. \'an't Hoff calls attention to the rate of carbon dioxid elimina-
tion in plants which show a coefficient of 2.5. Kanitz gives a long
list of similar coefficients for plant respiration, rate of isolated hearts,
contraction of smooth muscle and the metabohsm in cold blooded
animals.

SU.MM.ARV .\N-D CONCLUSIONS

1. Eight calorimeter experiments have been made on subjects after
the intravenous injection of proteose and of typhoid vaccine. In
five of these it was possible to observe the phenomena of chills.

2. With the onset of a chill there is a sudden increase of from 75
to 200 per cent, in heat production due, in part, to the shivering. At
the same time, there is almost no rise in the heat elimination. This
discrepancy between production and elimination causes the storage
of a large amount of heat within the body. After the chill there is a
short period of level temperature when the heat production and heat
elimination are equal and both are from 20 to 40 per cent, above the
basal level. Following this, as the temperature falls, there is usually
a steady decrease in heat production until it reaches the normal level.
The heat elimination, on the other hand, increases still farther, thus
getting rid of the stored heat. During the falling temperature there
is never as large a discrepancv between elimination and production
as during the chill.

3. The respiratory quotient tends to be high during the chill, indicat-
ing the rapid combustion of the glycogen stores of the body. Aft°'-
the chili, the quotient falls steadily.

4. By means of a comparison of the heat production and heat
elimination, it is possible to determine the temperature changes of the
body as a whole and compare them with the changes in rectal temper-
ature. The rectal temperature indicates, in a general way, changes
in average body temperature, but it is possible to have a rise in
rectal temperature while there is a fall in the average body temperature.
The opposite is also true.

5. The heat lost in the vaporization of water from the skin and
lungs bears a fairly constant relationship to the total heat elimination
but has no relationship to the heat production during rapid changes in
temperature. Study of the water elimination in fever affords no
evidence that the body is unable to mobilize water for heat elimination.
Fever should not lie attributed to failure in the function of water
elimination.

6. Observations in this and in other fevers has demonstrated that
rise in body temperature is accompanied by increased heat production
the amount of which corresponds to the degree of fever. It is found

634 ARCHII-ES OF IXTEKXAL MEDICIXE

that this increase follows van't Hoff's law, which may be stated as
follows: "With a rise in temperature of 10 C, the velocity of chemical
reactions increases between two and three times."

7. The phenomena of the chill following intravenous injection of
proteose or vaccine are strikingly similar to those of the malarial
paroxysm, the method of temperature regulation being almost identical.

A CORRELATED STUDY OF THE INDICATIONS FOR

TONSILLECTOMY AND OF THE PATHOLOGY AND

BACTERIOLOGY OF THE EXCISED TONSILS*

LEOXORA HAMBRECHT, B.S., and FRANKLIN R. NUZU>r, M.D.

SANTA BARBARA. CALIF.

Careful studies of the flora of the tonsils and of the nasopharyngeal
cavities have been made. The discovery of hemolytic streptococci
is stressed as the most important bacteriologic finding. They are reported
as present in approximately 50 per cent, of throats with extreme
variations of from 10 to 100 per cent. In the nasopharynx they have
been found less frequently and in smaller numbers. The importance
of virulent hemolytic streptococci being the causative agent of many
clinical conditions is well established. The role of these organisms
as secondary invaders in postinfluenzal pneumonia, empyema, etc., is
recognized. Also the ability of these bacteria to produce lesions in
the stomach, gallbladder, appendix and endocardium; these lesions
being secofidary to primary foci of infection in the tonsils, teeth, etc.,
is stoutly affirmed by many. Therefore, the role played by streptococci
in disease processes is an important one.

The steady progress that has been made in recent years in classifying
the large and heterogeneous group of streptococci has given an added
interest to the relation of the various strains of these organisms to
clinical conditions. It has been demonstrated that several strains are
homogeneous, that they have definite cultural characteristics, sugar
reactions and agglutinin and precipitin properties. The association of
certain strains with specific clinical conditions is now proved. Of
the eight groups of hemolytic streptococci in Holman's classification,^
four are known to be of frequent occurrence and closely related to
septic disease processes and the remaining four are apparently nonpatho-
genic and of infrequent occurrence. Progress has also been made con-
cerning the relation and the clinical significance of Holman's eight
subgroups of nonhemolytic streptococci.

Likewise, the pathology of excised tonsils has been studied carefully
and the changes which occur as the result of local infection are well
known.

But while each phase of the relationship between the clinical
indication for the removal of the tonsils, the pathology of the excised
tonsils, and the predominating organism recovered, has been recorded

* From the Laboratories of the Santa Barliara Cottage Hospital.
L Holman, W. L. : Classification of Streptococci, J. M. Rcsearcli 28:, 37/
1916.

636 ARCHirES OF IXTERXAL MEDIC I XE

many times by independent workers, the three phases have not been
frequently worked otit on the same inaterial.

With this in mind we have tabulated the clinical indications for
tonsillectomy in 218 patients. We have noted the pathologic changes
in the excised tonsils. In studying the bacterial flora we have compared
the results obtained by making cultures from the surface of the
tonsil and, after cutting the tonsil with a sterile knife, with the cultures
made from the depths of crypts. In tabulating the streptococci found,
we have used the classification of Smith and Brown,- which divides
these organisms into the hemolytic alpha, alpha prime, and beta groups
and into the nonhemolytic gamma group. In deep colonies in poured
blood agar plates which are not too thickly seeded the alpha group has
a green area of discolored corpuscles immediately about the colony,
outside of which there may or may not appear a partially hemolyzed
zone. If the petri dish containing such colonies is placed in the icebox
for twenty- four hours, the hemolyzed zone about the colony may
become more clearly defined. Outside of this zone a second zone of
greenish discoloration may develop, and beyond this another zone of
hemolysis. If such a culture is alternately incubated and refrigerated
as many as four zones may develop. Pneumococci resemble this group
so closely that we have determined their solubility in bile, in order
to accurately identify the colony. The alpha prime group of hemolytic
streptococci have in deep cultures a faint haze immediately about the
colony, due to incomplete hemolysis. The colonies of the beta group
are surrounded by a sharply defined zone of complete hemolysis which
varies from 1 to 5 mm. in diameter. The gamma group produces no
hemolysis.

Method.— In order to obtain comparable results, we have followed
the cultural methods suggested by Brown,^ using beef extract and
Digestive Ferments Co. peptone, the agar having a reaction of 0.5
per cent, normal acid to phenolphthalein. Twelve c.c. of this was
mixed with 0.5 c.c. of human blood. Such a medium is very suitable
for studying the various types of hemolysis. With a sterile loop material
from the crypts was placed in about 2 c.c. of sterile physiologic solu-
tion of sodium chlorid from which emulsion the blood agar was
inoculated, poured into Petri dishes and incubated. The plates were
read at the end of twenty-four and forty-eight hours, then they were
l^laced in the icebox for twenty-four hours and again read. The low
power of the microscope was used in studying the type of heniolv-is
about the colon\-.

2. Smith, T.. and Brown. J. H. : A Study of Streptococci Isolated from Cer-
tain Presumably Milk-Borne Epidemics of Tonsillitis Occurring in Massa-
chusetts in 1913 and 1914, J. M. Research 31:445. 191S.

3. Brown, J. H.: The Use of Blood Agar for the Study of Streptococci.
Monograph, Rockefeller Inst. M. Research, \o. 9. p. 40.

HAMBRECHT-XUZLM—TOXSILLECTOMV 63/

Single colonies were fished from Petri dishes, streaked on blood
agar slants and the latter incubated. On obtaining a pure culture, the
bile solubility of the organism and its sugar reactions were determined
in those instances in which it Was necessary.

Results. — Of the patients in this group of 218, 60 per cent, were
children, 40 per cent, adults. Sixty-three per cent, gave a history of
repeated sore throat. Frequent colds, togethei with enlargement of
the tonsils were recorded in 14 per cent. Eight per cent, complained
of mouth breathing. Eight per cent, had a history of one or more
attacks of rheumatism. Seven per cent, had otitis media, and tonsil-
lectomy was done in an endeavor to improve this condition.

.Microscopic study of the entire number of tonsils removed revealed
definite pathologic changes in 93 per cent. Seven per cent, were
regarded as normal. In classifying the type of change we have followed
the p'an used by Moore,^ i. e., chronic lacunar tonsillitis, chronic inter-
stitial tonsillitis, and chronic peritonsillitis. We have classified separately
tho?e instances that presented gross or microscopic abscesses and those
in which lymphatic hyperplasia was the only change noted (Table 1 )

T.^BLE 1.— Perce.n-t.\ce of Ixst.ances in Which the V.-^rious
Ch.^xces Were Noted

Pathologic Condition Pfrceniagc

Chronic lacunar tonsillitis H

Chronic interstitial tonsillitis 21

Chronic peritonsillitis 6

Abscesses 10

Lymphatic hyperplasia as the only change 14

Mucous glands in capsule "!

Cartilage in capsule '• S

Forty-two per cent., or nearly one-half of the entire number, presented
changes in the lining epithelium of the crypts and its immediate
neighborhood. In many, these crypts were filled with debris. The epi-
thelium itself was quite frequently necrotic and microscopic abscesses
just beneath the crypt epithelium were common. In the 21 per cent,
of instances in which chronic interstitial changes were noted, there was
a marked increase in the connective tissue .stroma of the tonsil, repre-
senting a chronic inflammatory reaction. In the five per cent, of
tonsils that had peritonsillar involvement the changes as described about
the crypts and the connective tissue frame work were usually present
to some degree but the evidence of infection and of repeated inflamma-
tion were especially pronounced in the thickening and hyperplasia of
the capsular tissue.

The tonsils of the patients with a history of many attacks of tonsil-
litis, or of rheumatism, or of frequent colds, of mouth breathing, of
otitis media, presented certain pathologic changes.

4. Moore. J. J.: Chronic Tonsillar Infections. J. Lab. & Clin. Med. 3:2,«.3. 1917.

638 ARCHIVES OF IXTERXAL MEDICIXE

In Table 2 it will be noted tbat in 53 per cent, of the patients who
had attacks of tonsillitis, the pathologic changes in the tonsil were most
marked about the crypts, that is, the so-called chronic lacunar tonsil-
litis. Fourteen and four tenths per cent, of these had an associated
lymphatic hyperplasia. The next most frequent change in this type of
tonsil was the chronic interstitial tonsillitis present in 17.7 per cent.,,
while peritonsillitis and abscesses were present in 8 and 4 per cent.,
respectively. The tonsils of the patients with a rheumatic history pre-
sented chronic interstitial change in 66.6 per cent, and peritonsillar
change in 33 per cent. In 16 per cent, of these tonsils abscesses were
found. Of the patients complaining of frequent colds, 58 per cent,
had a lymphatic hyperplasia, and of those complaining of mouth breath-
ing 100 per cent, presented this change. In these patients the adenoid
tissue was also increased in amount.

TABLE 2. — Pathologic Ch.^nges in Tonsils in Various Conditions

Attacks of Mouth Otitis

Tonsillitis, Eheumatism, Colds, Breatliing. Media.

Patliology per Cent. per Cent. per Cent. per Cent, per Cent.

Chronic lacunar tonsillitis 53.0 16.6 33.3 ... 50

Chronic interstitial tonsillitis. IT.T 66.6 16.6

Chronic peritonsillitis S.O 33.0

Abscesses ot tonsils 4.0 16.0

Lymphatic hyperplasia 11.4* .... 58.3 lOO 50

* In association with other cluniges.

The bacteriologic findings, using Smith and Brown's classification,
are interesting from many standpoints. The results of other workers
using either the streaked plate or the poured plate method are so
variant that we have used both methods simultaneously. We have thus
been able to contrast the findings and to judge fairly of the worth of
each method (Table 3). Without specifically subdividing the hemolytic
streptococci into alpha, alpha prime or beta groups we have the
following :

T.^BLE 3.— Bacteriologic Findings

Streaked Plate Poured

Method, Method,

per Cent. per Cent.

Hemolytic streptococci, all types 45.0 96.1

Nonliemolytic streptococci ' 3.5 2.6

Staphylococcus w.o 3r.o

Pneumococcus i°» 'V-*

Bacilli 7.0 1.9

Comparing these tables the great diversity in the results obtained
is apparent. Using jwured plates, hemolytic streptococci and pneumo-
cocci were found twice as frequently, whereas staphylococci in report-
able numbers were found less than one-half as frequently as in streaked
flatcs. Bv the latter metliod certain rajiidly growing organisms, such

HAMBRECHT-XUZUM— TONSILLECTOMY 639

as the staphylococcus, obscured other and often more important organ-
isms. Hemolytic staphylococci prohibited the recognition of other
hemolytic organisms. The various groups of hemolytic streptococci
were recognized with difficulty and uncertainty by the streaked method
and were frequently missed altogether. We have also noticed that on
a streaked plate the hemolytic beta group often completely obliterated
the alpha, the alpha prime, or the gamma type even when the latter
are present in numbers.

The findings explain the wide difterence in the percentage of various
organisms in tonsil reports by other workers, some of whom used one
method and some the other. To us it is sufficient evidence that for
accurate results the poured plate must be used.

The hemolytic streptococci which were present in 96.1 per cent, of
the 218 tonsil examinations were classified according to the type of
hemolysis. Colonies of the alpha type were present in notable numbers
in 25 per cent, of instances, alpha prime in 32 per cent, and beta in
86.1 per cent.; all three of them frequently occurring in a single culture
from the same tonsil. The beta group was further subdivided into
those of wide and of narrow zone of hemolysis. Those classified as
of the narrow zone had a diameter of 1 mm. ; the wide zone had an
average diameter of 4 mm. Since most beta hemolytic streptococci
fall into one of the two groups, the narrow zone group representing
36.8 per cent, and the wide zone 73.4 per cent., we attempted to demon-
strate that this small or wide zone characteristic was a fixed thing and
warranted a further subdivision of the beta group.

Both groups were found to be pathogenic for rabbits (1 c.c. of a
twenty-four-hour bouillon culture). We then routinely inoculated the
test sugars with wide and narrow zone cultures and fermentation has
uniformly resulted as shown in Table 4.

TABLE 4.— DlFFEREXTI.\TION OF STREPTOCOCCI ACCORDING TO

Zone Production in Culture

Saccharine Lactose Eafflnose Salicin Mannite Iniilln

Since there had been no diflference in the reactions of the sugars
and since both were pathogenic for rabbits, the permanency of the
characteristic of a wide or narrow zone of hemolysis became of special
interest. Clawson found hemolysis constant in a series of 134 strains
even after two years. Brown ' recorded some loss of hemolytic activity
for certain strains. This loss always took the form of slightly smaller

5. Brown. J. H. : The Use of Blood Agar for the Study of Streptococci,
Monograph, Rockefeller Inst. M. Research, No. 9, p. 81.

640 ARCHIVES OF IXTERXAL MEDIC IS E

zones after a period of from fourteen months to two years. He never
noted a transition from the beta to the alpha types or vice versa and con-
ckided that the permanence of the apparently minor characteristics of
all the strains studied is surprising. Anthony '' also found only a slight
variability in the hemolytic power of streptococci. It being the con-
sensus of opinion that hemolytic characteristics of streptococci are
constant, we felt that very possibly beta hemolytic streptococci could
be di\ided into wide and narrow zone groups. However, this character-
istic of a wide or narrow zone of hemolysis was found to vary
frequently on subculture. A culture from a single colony of a wide or
a narrow zone organism often gave colonies with both wide and narrow
zones of hemolysis. This occurred so frequently that we felt that the
evidence was not at hand to warrant a subdivision of this group. It
would seem that the variation in the extent of hemolysis has to do
with an e.xtra cellular streptolysin, a substance which has been demon-
strated in cultures of hemolytic streptococci and which is a common
property of the members of this group.' The amount of hemolysis
apparently depends on the amount of streptolysin produced. -

A detailed study of the organisms isolated from the excised tonsils
of patients with various complaints such as tonsillitis, rheumatism, etc.,
was made. The findings are recorded in Table 5.

TABLE 5. — Org.^xisms Isol.'^ted from Excised Tonsils

.Alpha,

per Cent.

5.0

11.0

S.O

t.5
HO

-Alpha

Prime,

per Cent.

2.3

i6!6

Beta,
per Ceni
T6 0
Sl.O
61.0
100.0
54.5

14.0

.411 Types

Hemolytic

Streptococci,

t. percent.

83.0

92.0

79.0

100.0

73.0

S.i.4

Gamma,

per Cent.

1.5

2.1

Pneumo-
per Cent.

5.0

S.O
. 10.0

'oio

6.4
14.0

aphylococci
and Oiher

Oiganisms,

per Cent.

10.0

Mouth breathing...
Frequent eolds

ii!6

Rheumatism

i)0

Pathologic tonsils..
Normal tonsils

6.0

,56,0

Following tonsillitis the predominating organism was a hemolvtic
streptococcus in 83 per cent., the pneumococcus in 5 per cent, and staphy-
lococci and other heterogenous organisms in 10 per cent. Xon-
hemolytic organisms predominated in only 1.5 per cent. Of the
hemolytic organisms, the beta group were present in 76.3 per cent., the
alpha group in 5 per cent, and the alpha prime in 2.5 per cent. As
a contrast, tonsils from patients with rheumatism had as the pre-
dominating organism beta hemolytic streptococci in only 54.5 per cent.,
the a'pha or Streptococcus :-.ndans group in 18.5 per cent, and non-

6. .AiT.honv : Some Characteristics of the Streptococci Found in Scarlet
Fever. J. Infect. Dis. 6:332, 1909.

7. Ruediger. G. F. : The Production and Nature of Streptocolysin. I. A.
M. A. 41:962 (May 12) 1903.

%. Braun, H. : Ueber da"; Streptolysin. CentralW. f. Baktcriol. 62:383. 1912.

HAMBRECHT-XUZUM—TOXSILLECTOMY 641

hemolytic streptococci, pneumococci and staphylococci in 9 per cent.,
each. The alpha or Streptococcus z'irid<ins group was present in a
higher percentage in these tonsils than in any other. This coincides
well with clinical findings, this type of organism having a close relation-
ship to rheumatism. Patients with otitis media had beta streptococci
as a predominating organism in 100 per cent. The ear discharge had
the same organism. Mouth breathing patients whose tonsils presented
lymphatic hyperplasia harbored hemolytic streptococci in 92 per cent,
and pneumococci in 8 per cent. Patients complaining of frequent colds
and whose tonsils also presented lymphatic hyperplasia, had as the pre-
dominating organism beta hemolytic streptococci in 61 per cent., alpha
streptococci in 8 per cent., alpha prime «n 10 per cent., pneumococci in
10 per cent, and staphylococcus and other incidental organisms in 11
per cent.

Taking all of the pathologic tonsils as a groujx hemolytic strepto-
cocci were present in 96.1 per cent. In 85.4 per ctnt. they were present
as the predominating organism. Taking the bacteriologic findings of
the 7 per cent, of normal tonsils in this series hemolytic streptococci of
all types were present in only 28 per cent., pneumococci in only 14 per
cent, and staphylococci and heterogeneous organisms in 58 per cent.
Granting that this is too small a number of normal tonsils on which to
bafe conclusions, yet it is apparent that hemolytic streptococci are
present in a higher percentage in pathologic tonsils. This high incidence
of hemolytic streptococci is in keeping with the findings of Davis,''
Pilot and Pearlman "' and others, though it is considerably higher than
many other investigators have recorded.. The latter, however, in many
instances did not use the poured plate method.

W't have made cultures from the crypts of excised adenoid tissue
in eighty-four instances. The bacteriologic findings followed so closely
the findings of the tonsil culture in the same patient that a separate
table did not seem warranted. As a rule, the number of organisms
in the adenoid tissue was much smaller than in the tonsil.

SUMM.^RN'

'If 218 persons said clinically to need a tonsillectomy, microscopic
examination of these tonsils gave evidence of pathology in 93 per cent. :
7 per cent, were normal.

Of these 218 persons, 63 per cent, gave a history of repeated "sore
throats." 14 per cent, of frequent colds and were told that they had
enlarged tonsils. 8 per cent, complained of mouth breathing. 8 per cent,
of rheumatism and 7 per cent, of otitis media.

9 r-v:s. D. ].: Hemolytic Streptococci. .1. .\. M. .-X. 72:319 (Feb. 1) 1919
10. Pilrt. 1.. Hnd Pearlman. S. J.: Bacteriologic Studies of the Upper Res-
piratory Passages. J. Infect. Dis. 29:47. 1921.

642 ARCHIVES OF IXTERXAL MEDICIXE

Chronic lacunar (crypt) tonsillitis was the most frequent pathologic
condition found. It occurred in 42 per cent. Chronic interstitial tonsil-
litis was present in 21 per cent., chronic peritonsillitis in 6 per cent.,
gross or microscopic abscess in 10 per cent, and lymphatic hyperplasia
as the only change in 14 per cent.

Following ' repeated attacks of tonsillitis, changes in the tonsils
occurred most often about the crypts (42 per cent.). In tonsils from
patients with a history of rheumatism, chronic interstitial tonsillitis
was present in 66 per cent., chronic peritonsillitis in 23 per cent.

The organisms most frequently isolated from these tonsils were
hemoljliic streptococci. They were present in 96.1 per cent, of all
tonsils and were the predomingj:ing organism in 85.4 per cent.

The hemolytic streptococci were further subdivided into alpha,
alpha-prime, and beta groups and were present in 25 per cent., 32 per
cent, and 86.1 per cent., respectively.

Hemolytic streptococci were present in a much higher percentage
(96.1 per cent.) in the pathologic tonsils of our group than in the
normal tonsils (28 per cent.).

In nearly every instance the same organisms were isolated from
the adenoid tissue as from the tonsils of that patient.

Ring formation, as described by Brown, occurred only with the
alpha group of hemolytic streptococci and in 5 per cent, of the total
number of the alpha cultures.

.'\virulent diphtheria bacilli were isolated but three times in the
entire series.

By using both the streaked plate and the poured plate method of
culture and comparing the resiilts, we have found the latter much more
accurate and satisfactory.

BLOOD PIGMENT METABOLISM AND ITS RELA-
TION TO LIVER FUNCTION*

CHESTER M. JOXES. M.D.

BOSTON

The exact mechanism involved in the metabohsm of the blood
pigments, and the precise relation of the liver to these processes, are
still but imperfectly understood. Certain theories, however, concerning
blood pigment metabolism are very generally accepted. The pigments
of the bile have long been believed to be derivatives, in part at least, of
hemoglobin. The work of Eppinger and Charnas,^ Wilbur and Addis,-
Robertson,^ Schneider,^ Hansmann and Howard," Giffin, Sanford and
Szlapka," and others has shown that excessive degrees of red cell
destruction are accompanied by an increased elimination of bile pig-
ments. Most observers agree that the liver is the main agent concerned
in these metabolic changes. However, the lower bile pigments, princi-
pally urobilin and urobilinogen, have been supposed to be formed
independently of the liver, by the action of bacteria in the lower intes-
tine, on the bilirubin of the bile. The recent work of Hooper and
Whipple ' on dogs with biliary fistulae has made necessary a modifica-
tion of previous theories. These investigators question the intestinal
production of urobilinogen and urobilin, and the absorption of these
pigments from the portal circulation. They suggest that the liver itself
is capable of forming these substances. They also prove that bilirubin
can be formed in various parts of the body without the intervention of
the liver, and conclude that normally the liver may be only one of
several agents in the process of hemoglobin metabolism.^ Further-
more, they produce evidence that red cell destruction, with the conse-
quent liberation of hemoglobin, is not the only factor in the production

* From the Medical Services of the Massacliusetts General Hospital.

* This paper is No. 25 of a series of articles on the physiology and pathology
of the blood from the Harvard Medical School and allied hospitals, a part of
the expense of which has been defrayed from a grant from the Proctor Fund
of the Harvard Medical School for the study of chronic diseases.

1. Eppinger and Charnas : Arch. f. klin. Med. 78:387, 1913.
• 2. Wilbur and Addis: Arch. Int. Med. 13:325 (March) 1914.

3. Robertson, O. W.: Arch. Int. Med. 15:1072 (June) 1915.

4. Schneider, J. P.: Arch. Int. Med. 17:32 (Jan.) 1916.

5. Hansmann and Howard: J. A. M. A. 73:1262 (Oct. 25) 1919.

6. GifTin, Sanford and Szlapka : Am. J. M. Sc. 182:562, 1918.

7. Hooper and Whipple: Am. J. Physiol. 40:.1!2. 1916.

8. Whipple and Hooper: J. F.xper. M. 17:612. 1913.

644 ARCHIVES OF INTERNAL MEDICIXE

of bile pigments. Dietary changes ^ and various drugs ^" are able to
cause marked alterations in the elimination of pigments in the bile.

The resynthesis of the bile pigments to hemoglobin is a further
property usually attributed to the liver. At present it is generally
believed that the liver is able in some way to build up the lower pigment
fractions into the more complex molecule of hemoglobin. However,
evidence is lacking on this point, as well as to the manner in which
hemoglobin becomes incorporated in the red corpuscles.

It is the purpose of this paper to present certain evidence regarding
the normal and abnormal physiology of blood pigment metabolism in
man. and to demonstrate, if possible, a further relation between the
liver and such processes.

METHOD

The introduction of "biliary drainage" by Lyon,^^ in 1919, has pro-
vided a method by means of which a more systematic clinical study of
the duodenum' and biliary tract is permitted than was previously
possible. Lyon claims that a solution of magnesium sulphate intro-
duced through the duodenal tube relaxes the sphincter of Oddi, and
thus a free flow of bile into the duodenum is obtained. Furthermore,
Lyon believes, following Meltzer's '- theory of contrary innervation,
that the magnesium sulphate causes a contraction of the gallbladder
musculature, with a resulting flow of gallbladder contents into the
duodenum. Following the flow of dark gallbladder bile Lyon obtains a
flow of lighter colored bile which he believes is derived from the upper
biliary radicles and liver. Lyon thus attempts, after the use of mag-
nesium sulphate, to divide the bile drainage into three fractions, "A,"
"B" and "C," which are supposed to contain respectively bile from the
common duct, gallbladder and liver. By a study of the gross color,
certain other physical characteristics, the sediment and the bacteriology
of these fractions, Lyon believes it possible to diagnose and localize
pathology existing in the duodenum and various portions of the biliary
tract.

Although Lyon presents clinical data which are quite consistent
with his assumptions, nevertheless definite experimental proof that
magnesium sulphate, when introduced into the duodenum, causes a
contraction of the gallbladder is conspicuously lacking. The exact
action of the salt in the duodenum has yet to be determined. There is
a certain amount of evidence that the relaxation of the sphincter of
the common bile duct may not be accompanied by. contraction of the

9. Hooper and Whipple: J. Exper. M. 23:137, 1916.

10. Bauer and Spiegel: Deutsch. .^rcli. f. klin. Med. 1:129, 1919.

11. Lyon, B. B. v.: 1. A. M. A. 73:980 (Sept. 27) 1919.

12. Meltzer. S. J.: Am. T. M. Sc. 153:469, 1917.

JOXES— BLOOD PIGMEXT METABOLISM (,45

gallbladder walls. Chrohn, Reiss and Radin " were unable to prove
experimentally the existence of such a contrary innervation, although
they apparently believe that the so-called "B" bile contains gallbladder
contents. Einhorn ^* claims that various salts produce a flow of '"B"
bile into the duodenum. This assumption is undoubtedly correct,
although magnesium sulphate produces more constant and better results
than other salts. Einhorn further concludes from his experiments that
the flow of "B" bile is not due to a flow of bile from the gallbladder
into the duodenum, but that it is due merely to stimulation of liver
cells to increased activity, with a resulting excretion of bile pigment in
increased concentration. This conclusion, however, is based on crude
quantitative estimations of pigment values, and is probably incorrect.
Careful determinations of the bile pigments in a series of fractions
taken before and after the use of magnesium sulphate, in a number of
cases in which there was known to be no flow of bile possible from the
gallbladder, either on account of a previous obstruction of the cystic
duct, or on account of a previous cholecystectomy, tend, by comparison
with a series of normal cases, to disprove Einhorn's conclusions. Fur-
ther reference will be made to these determinations in a later portion
of this paper.

For practical consideration, in spite of the fact that experimental
work is still lacking as to the exact source of the "B'" bile, it seems
expedient to assume that it is made up, in part, of bile from the gall-
bladder. It is highly probable that a solution of magnesium sulphate,
when instilled into the duodenum, accomplishes two things. First, it
relaxes the sphincter of the common bile duct and causes a free flow
of bile into the duodenum. Second, it probably causes a slight contrac-
tion of the gallbladder musculature, with the result that some bile from
that organ is mixed in with the bile proceeding down the common duct.
The result is a mixture of duct, liver and gallbladder bile.

Bile pigments in the duodenal contents have received but slight
attention. Schneider,'' in 1916, and subsequently others, have made
quantitative estimates of the bile pigments of the duodenal contents,
using a spectroscopic method. These observers concluded that in those
cases in which it is generally considered that increased blood destruction
is taking place the excretion of bile pigments is also increased.
Schneider attempted to show a definite relation between the level of
the bile pigments in the duodenum and the actual degree of hemolysis
obtaining in any given case. Eppinger, Wilbur and Addis, Robertson,
and others had previously obtained high pigment value in similar
instances, by making bile pigment determinations of the stools. Hans-

13. Crohn, Reiss and Radin: J. A. M. A. 76:1567 (Junt 4)

14. Einhorn, M.: Xew York M. J. 113:313. 1921.

646 ARCHIVES OF IXTERXAL MEDICIXE

mann and Howard compared the method of estimating the pigments
in the stools with the estimations based on the duodenal contents.
Figures obtained by either method gave relatively high pigment values
in those cases in which increased hemolysis was apparently taking
place. Their findings were confirmatory of results obtained by Wilbur
and Addis, but Hansmann and Howard do not believe that, estimates
based on duodenal contents run exactly parallel to those obtained from
stool examination. Hansmann and Howard, however, believe the stool
method to be more correct. Examination of duodenal contents seems,
nevertheless, the more logical method of study. Such a method allows
a study of the bile before the pigments have become diminished or
altered by action of the intestinal bacteria. Furthennore, analyses
based on estimation of bile pigments in the duodenal contents are
performed more easily than similar determinations on the stools and
are not subject to errors due to such variable factors as constipation,
diarrhea, etc.

Lyon's method of obtaining a continuous flow of bile offers a dis-
tinct advantage over the method employed by Schneider and others, in
which determinations were based entirely on single specimens. Single
specimens, in the present studies, were subject to the greatest varia-
tions, on account of the intermittent flow of bile from the common bile
duct, and on account of various other factors such as salivary, gastric
and pancreatic secretions, which introduced errors by causing a dilution
of the pigment content in the duodenum.

The technic used in this series of cases consisted in the introduction
of the duodenal tube, and the collection by siphonage of duodenal con-
tents in six fractions. These six fractions were collected over fifteen
minute intervals, two fractions being taken from the fasting duodenum
jjrior to the introduction of a 33 per cent, solution of magnesium
sulphate, and four immediately following the use of the salt. The
entire collection of duodenal contents thus covered a period of about
one hour and a half.

The duodenal tube was retained over a period of from two to three
hours in the majority of cases, depending on the length of time neces-
sary for the tip to reach the duodenum. The exact location of the tube
in the duodenum was determined by fluoroscopic examination in the
majority of cases. Atropin sulphate, given before the introduction of
the tube, practically eliminated any undue flow of saliva. The use of
magnesium sulphate provided a nearly continuous and concentrated
flow of bile into the duodenum and minimized the errors caused by
the flow of gastric and pancreatic secretions. The objection might be
raised that atropin might of itself introduce an error, by causing
individual variations in the output of bile. Atropin does cause a slight

JOKES— BLOOD PIGMEXT METABOLISM 647

diminution in the excretion of bile by the liver cells.'" This diminu-
tion is, however, very slight and in the cases studied the administration
of the drug caused no appreciable effect in the flow of bile into the
duodenum.

Bile pigments were estimated by \\'ilbur and Addis' method of
spectroscopic examination, for each of the six fractions. This method
consists essentially in dissolving the urobilinogen, urobilin and other
lower bile pigments in a saturated alcoholic solution of zinc acetate,
and then determining the pigment content by the spectroscope. The
number of dilutions necessary to cause the disappearance of the char-
acteristic absorption bands of the individual pigment was taken as the
reading for any particular fraction, and a curve was plotted from the
values obtained. Values of urobilinogen and urobilin were added
together, and the total taken as the pigment value of the fraction. An
attempt was also made to quantitate the bilirubin values of the duodenal
contents, by the method described by Hooper and \\'hipple in their
work on dogs, but it was found impossible to obtain consistent readings
on human bile owing to the conversion of bilirubin in some of the
fractions into bilicyanin. The color obtained by this method, by treat-
ing the bile with acid alcohol, was in some instances the characteristic
blue-green desired and could be read against a standard solution of
copper sulphate as described by these authors. In the majority of
cases, however, the color ranged from a decided green to a dark blue,
and occasionally the entire series of fractions was intensely purple
owing to the presence of bilicyanin. Similar observations on animals
have been made recently by Rous and McMaster.'" Bilirubin figures,
when obtained, ran approximately parallel to those of urobilin and
urobilinogen. The actual dilution figures obtained from spectroscopic
examination were not multiplied by a constant, as done by Wilbur and
Addis in their original work, and later by Schneider, as there seemed
no advantage to be gained by this purely artificial procedure. The
curves shown on the accompanying charts, therefore, represent actual
dilution values of urobilinogen plus urobilin.

The method of fractional analysis, I believe, offers distinct advan-
tages over the method of studying only a single specimen. It provides
a free flow of bile into the duodenum over a considerable period of
time, and permits the taking of an average figure as well as the value
of individual fractions. In this way it is possible to make a comparative
study of the dififerent fractions, and to obtain a much more exact
picture of the level of bile pigments than can be gained from any
single observation. Even such a method, however, is open to error.

15. Rous and Mc.Master : J. E.xper. M. 34:47, 1921.

648 ARCHIl'ES OF INTERNAL MEDICIXE

and I wish only to point out its advantages and to emphasize its relative
accuracy.

In addition to an estimation of the bile pigment in the duodenal
contents, Blankenhorn's ^'"' method for studying the bilirubin content
of the blood plasma was employed. This method consists essentially
in a comparison of the yellow color of oxalated plasma with distilled
water. Dilutions of the plasma with water are made until the yellow
color has disappeared. The number of dilutions necessary to remove
the yellow color of the plasma are taken as the approximate bilirubin
content of the specimen. Normally between fifteen and twenty dilu-
tions give the desired end-point.

Bile
Pigments

125

100

75

50

25

Fractionsl 2 • 3 4 5 6

Chart 1.— Duodenal pigments in normal individuals. In this, and subsequent
charts, points -along the ordinates represent dilution values of the bile pigments,
urobilin and urobilinogen. Points on the abscissae represent separate fractions
of duodenal contents collected over fifteen minute intervals. The above curve
is identical with the "normal"' curves given in Charts 2, 3, 4 and 5, although in
each case the scale varies. The arrow indicates the administration of 50 c.c. of
a 33 per cent, solution of magnesium sulphate.

Bile Pigments in the Duodenal Contents in Normals.— .\^ a basis
lor comparison with pathologic cases, observations on the pigment
values in the duodenal contents of eight normal individuals were made

16. Blankcnhorn, M. S.: .-Krch.

Med. 19:344 (March) 1917

JOXES— BLOOD PICMEXT METABOLISM 649

and an average curve drawn from the results obtained. As shown in
Chart 1, the average reading of all the six fractions in this series of
normal cases was forty dilution units, with a maximum variation from
this figure of ten units. The peak of the curve came shortly after the
administration of magnesium sulphate and represents Lyon's "B" bile.
The average pigment value of the peak of the curve was one hundred
dilution units, with an individual variation up to fifty units. It will
be seen that the variation from the average figures is a wide one, both
in individual fractions and in the case of the general averages obtained
from the total fractional estimations. These variations occurred in
spite of the fact that duodenal contents in all cases were taken under
similar conditions as regards the fasting state, the time at which the
duodenal contents were collected and general freedom from symptoms.
This point should be emphasized, inasmuch as previous investigators
have inferred that the individual variation among normal persons is
only a slight one. Furthermore, it is noticeable that there was a wide
variation between individual fractions in the same normal person, even
before the administration of magnesium sulphate. Bauer and Spiegel "
have noticed similar variations in normal individuals in estimating the
bilirubin content of the blood plasma. A further discrepancy may be
observed between the pigment values in normals as given by Schneider
and the values obtained before the use of magnesium sulphate in this
series of determinations. ]\Iy figures for normal individuals are rela-
tively higher than those of Schneider. His average normal figure is
about five dilution units. The results obtained from my series of
normals, in the fractions that are comparable to his analyses, average
about 8.5 dilution units. The difl:'erence between the two figures may
be explained (1) by individual differences in obtaining end-points by
spectroscopic examination, or (2) by the fact that specimens of duo-
denal contents were taken in this series after waiting a relatively long
time following the introduction of the tube. Such a wait would insure
a better flow of bile. In either event the dififerences are purely relative,
and conclusions based on examinations of similar cases in both series
are in the main identical.

In cases such as gastric ulcer, and so forth, in which there was no
apparent cause for abnormal pigment \ahies, there was essentially no
deviation from the normal range.

Evidence of a Floiv of Gallbladder Bile Follozving the Use of Mag-
nesium Sulphate. — Following the establishment of the normal figures
a series of cases was studied in which there was absolute obstruction
of the cystic duct, as proved at operation, or in which the gallbladder
had previously been removed. Obviously, there could be no flow of
gallbladder or "B" bile in these cases and a comparison of results

650

ARCHIVES OF IXTERX.-iL MEDICIXE

obtained in these cases with the normal figures already given showed
no characteristic peak after the administration of magnesium sulphate.
On the other hand, there was only a moderate rise in pigment values
after giving the salt, as shown in Chart 2, This rise can be explained
entirely by a relaxation of the sphincter of Oddi, with a resulting free
flow of undiluted bile into the duodenum.

In another series of cases in which there was definite gallbladder
pathology without obstruction, the fractions taken immediately after
the administration of magnesium sulphate, or in other words, those
fractions taken at a time corresponding to the peak of the pigment
curve, were the only ones to show certain cellular and crystalline ele-

5ile
Pigments

200

150

100

50

1

\

\

^

A

^ \

^

\

\

Chart 2. — Duodenal pigments in normal individuals as compared with pig-
ments in patients with no flow of gallbladder bile. Normal pigment values
Values with no gallbladder (low • — • — • — •

ments believed to be characteristic of gallbladder contents. The similarity
of the sediments in these fractions to those actually obtained from the
gallbladders at the time of operation was striking. The close corre-
spondence between these sediment findings, their occurrence coincident
with the peak of the pigment curves, and the absence of a characteristic
pigment curve in cases where there was known to be no flow of bile
from the gallbladders all confirm the assumption that the so-called "B"
bile consists, at least in part, of actual gallbladder contents. The
importance of this assumption will be discussed in a later portion of
the paper in a study of cases in which there was known gallbladder
pathology.

JOKES— BLOOD PIGMEXT METABOLISM 6j1

Relation of Blood Destruction to Bile Pigment Elimination. — It was
desirable to establish, if possible, further definite evidence that there
was a distinct relation between the amount of hemolysis going on in
the body, and the level of the bile pigments. Observations were made
in two cases of paroxysmal hemoglobinuria which have been reported
in a separate paper. ^' In these cases immersion of the extremities in
icewater caused immediate and marked intravascular hemolysis. The
plasma, duodenal contents and urine were examined for changes in
pigment values during the course of the observations, which covered
a period of about twenty-two hours. There was no important change
in urinary pigments inasmuch as the attacks produced were not severe
enough to cause any but the slightest traces of hemoglobin to appear
in the urine. The hemoglobinemia was immediate and intense and was
accompanied in one case by a drop in the red count of over 800,000
cells per c.mm. Subsequently the hemoglobin content of the plasma
rapidly diminished, with an accompanying marked increase in the
bilirubin content. This increase in bile pigment in the plasma con-
tinued until it reached its height at a point coinciding with the disap-
pearance of hemoglobin from the plasma. It then gradually dropped,
reaching normal at the end of about eighteen hours. Coincident with
the peak of the bilirubin content of the plasma, the duodenal pigments
rose rapidly, reaching a level about six to eight times the normal level
in about three hours. These pigments did not return to the normal
level until after eighteen to twenty hours. These results rather defi-
nitely confirmed the generally accepted theory that increased blood
destruction is accompanied by increased elimination of bile pigments.
Furthermore, it seemed safe to assume from the above observations
that the bulk of the hemoglobin liberated into the circulation as a result
of any hemolytic process is rapidly taken care of within the liver and
broken down into lower bile pigments. Although other organs and
tissues possess a similar property of carn'ing on the metabolism of
blood pigments, under normal conditions the liver probably carries
on the greater part of this important chemical process.

Bile Pigments in Various Types of Anemia. — With the clear recog-
nition that increased blood destruction is accompanied by an increased
elimination of bile pigments in the plasma and bile, as demonstrated by
the above observations on paroxysmal hemoglobinuria, and as brought
out by numerous investigators, a series of cases of various types of
anemia was studied. This series included cases of anemia due to severe
hemorrhage, lowered bone-marrow activity, pernicious anemia, hemo-
lytic jaundice, malaria, and so forth. A somewhat similar series had
been studied by Schneider, and later by Giffin, Sanford and Szlapka.

17. Jones, C. M., and Jones, B. B. : .Arch. Int. Med. 29:669 (May) 1922.

652 ARCHIVES OF IXTERXAL MEDICI XE

1 wished, however, to obtain a comparative set of figures by the
fractional method of duodenal analysis, and to attempt a more detailed
study of the abnormal physiology occurring in these diseases. The
cases studied fell roughly into two groups: (1) cases in which increased
blood destruction is believed not to be present, or at least is not an
important feature, and (2) cases in which it is generally believed that
abnormal blood destruction is an important feature of the disease
process.

As examples of the first type of cases a group of patients was
studied in which the anemia was due entirely to blood loss. The anemia
was due in two cases to hemorrhage from duodenal ulcers, in one to
renal hemorrhage, in one to a series of attacks of paroxysmal hemo-
globinuria, and in one to prolonged menorrhagia. The case of
paroxysmal hemoglobinuria had been free from attacks for more than
a week, so that there was no complicating factor of recent hemolysis.
In none of these cases was there any evidence of abnormal red cell
destruction. As was to be expected, the actual pigment values were
all under the normal average (Table 1), indicating possibly an attempt

T.\BLE 1. — Bile Pigmexts ix Axemi.\ from Blood Loss

Bile Pigments "Relative" Plasma Hemo-

in Duodenal Duodenal Bilirubin globin.

Contents Pigments Content per Cent

on the part of the body to conserve hemoglobin. '"Relatixe" figures,
based on the actual pigment readings in the duodenal contents and the
percentage of red cells in relation to normal, with one exception (Case
22), were also within or below the normal range. The single case
referred to, with high "relative" figures, had a profound anemia, and
the explanation for the high figures may lie in the fact that the liver
was improperly functioning on account of the anemia itself.

"Relative" figures were obtained on the following assumption : The
pigment values in the duodenal contents are in a sense absolute values,
in that these values do not take into consideration the amount of
circulating hemoglobin. Obviously, even if the pigment values in the
duodenal contents are the same, there is greater relative blood destruc-
tion in a case with a low red count and hemoglobin than in a case
with a normal red count and hemoglobin. It is interesting, therefore,
to attempt roughly to correct these figures of pigment values to the
same standard of circulating hemoglobin. Thus it is possible to ascer-
tain the relative inteiisitv of the blood destruction. It does not of

JOXES— BLOOD PIGMEXT METABOLISM 653

course necessarily follow that the same relative intensity of blood
destruction would obtain if the red corpuscles and hemoglobin were at
the normal level. Because it was simpler to carry out this correction
on the basis of the numerical differences of red corpuscles, this pro-
cedure was adopted, rather than correction by utilization of hemoglobin
variations, which theoretically is more logical. Relative figures were
obtained by dividing the pigment values in the duodenal contents by
the percentage of normal which the red count of the individual case
■':.howed, and then multiplying by 100. Thus, for example, a patient
with a count of 3,000,000 red corpuscles per c.mm., and a pigment aver-
age of 100 units, other things being equal, would theoretically be
destroying one-half the percentage of total red cells destroyed by a
patient with a count of 1,500,000 red corpuscles per c.mm., and a pig-
ment average also of 100 dilution units.

T.ABLE 2. — Bile Pigments in a C.^.se of .^pl.nstic .Anemi.a

Average

Bile Pigments "Relative" Plasma Hemo- Reil

in Duodenal Duodenal Bilirubin globin. Blood Cells

Case Contents Pigments Content per Cent. (Millions)

A single case of true aplastic anemia (Table 2) was studied, which
also showed actual bile pigment values well below the normal. How-
ever, the "relative" figures were high. Inasmuch as in this case also
the anemia was extreme, the explanation of the high "relative" figures
is possibly the same as that given for Case 22 of the preceding series,
namely, the effect on liver function of the profound anemia. A more
logical explanation may possibly be that, with an extremely low level
of red corpuscles, and with practically no new blood formation, the
few cells in the circulation undergo more rapid dissolution than normal
on account of the undue work nut on them.

T.\BLE 3.-

-Bile Pigments in a Case

OF POLVC

ytHEMiA Vera

Case

.Average

Bile Pigments "Relative"

in Duodena! Duodenal

Contents Pigments

54 32

Plasma

Bilirubin

Content

24

Hemo- Red
globin. Blood CelU
per Cent. (Millions)

irs 8.0

One case of polycythemia vera was studied. The patient had a red
cell count of 8,000,000 cells per c.mm., and a hemoglobin content of
133 per cent. The actual pigment values averaged only slightly above
normal (Table 3), but the "relative" figures were below the normal
average. In spite of the enormous increase in the number of red cor-

654

ARCHIVES OF IXTERXAL MEDICIXE

puscles the process of blood destruction in this case was apparently
normal, or even relatively below normal.

In contrast to the above cases, and as an instance of disease in
which it is generally conceded that there exists an apparently high
degree of blood destruction, a series of nineteen cases of pernicious
anemia was studied. Other observers have pointed out that in per-
nicious anemia there is a marked increase in the bile pigments in the
blood, duodenal contents, stools and urine.

Chart 3. — Duodenal pigments in pernicious :

values — . Actual pigment values

values o-o-o-o-o-o.

"Relative" pigment
- . Normal pigment

Observations made in this series of cases entirely confirmed the
results of previous investigators. There was a high level of bile pig-
ments both in the blood plasma and in the duodenal contents. The
average increase in duodenal pigments over normal was more than
500 per cent.; the highest averages were nine times, and the lowest
twice the normal figures. Plasma bilirubin was, in the average, about
four times normal ; the highest figure was about eight times the normal,
while the lowest was about twice normal. Based on average figures,
therefore, the bile pigment content of the blood and duodenal contents
ran about parallel, but in individual cases there was marked discrep-

JONES— BLOOD PIGMEST METABOLISM 655

ancy between the two values. This difference, in individual cases,
between the pigment content of the plasma and the pigment values of
the duodenal contents, is perhaps significant, and suggests some inter-
ference with liver function — a point that will be discussed later.

A question of considerable interest to be determined in studying
these cases was whether the level of bile pigments in the duodenal
contents corresponded with the actual clinical condition of the patients.
Weakness, elevation of temperature, icterus, level of hemoglobin and
red cells, etc., supply the clinician with evidence for comparison between
individual patients. The condition of the individual patient theoreti-
cally depends, in large measure, on the relative severity of the hemolytic
process and on the relative degree of blood-forming activity. Thus, a
patient who is subjectively sick, and who presents the typical features
of a relapse, usually gives evidence of a marked predominance of blood
destruction over blood formation. A patient, one the other hand, with
few subjective symptoms, a high hemoglobin content, usually shows
evidence of little blood destruction and, on the contrary, a satisfactory
blood formation. In spite of the theory that varying degrees of blood
destruction are accompanied by corresponding variations in the level of
the bile pigments, it is evident that changes in blood formation may
modify the clinical picture to such an extent that the level of the bile
pigments in the duodenal contents, although measuring the amount of
blood destruction taking place, will not reflect the patient's clinical
condition. High pigment values might, therefore, be obtained, even in
the presence of a severe degree of blood destruction, without corre-
spondingly se^vere clinical symptoms. Furthermore, any aheration of
liver activity should modify the bile pigment excretion, both in the bile
and in the blood plasma.

These theoretical considerations were well sustained by the findings.
Examination of the figures obtained in the nineteen cases of pernicious
anemia showed that the actual pigment values corresponded only in a
very rough way to the clinical condition of the patient. Patients whose
duodenal piginents were very high frequently were clinically less sick
than those who showed relatively low bile pigment estimations, and
vice versa. A second set of figures, however, did correspond closely to
the condition of the individual patient, both as regarded his clinical
condition and as concerned the actual physiologic processes taking
place. These second figures are the "relative" figures already referred
to, and were obtained by dividing the actual pigment readings by the
percentage of red cells of the particular case. Such a modification of
the actual pigment readings gave a close approximation to the clinical
state of the patient, and in addition appeared to serve as a much clearer
index of the relation of blood destruction to blood formation. Actually,

656 archu'es of ixterxal medicixe

patients with high "relative" figures were sick, and presented the chnical
findings of severe blood destruction — elevation of temperature, jaun-
dice, etc., that are typical of a severe relapse. Those patients whose
"relative" figures were only moderately high, on the contrary, were
free from the more marked symptoms, while patients with "relative"
pigment values at a still lower level were in a well marked remission.
The average level of the "relative" figures, however, was about three
times that of the actual readings, and indicates clearly the severity of
the disease process. Furthermore, a comparison of the actual and
"relative" values in any individual case provided a fair estimate of
the balance between blood destruction and blood formation. \\'hen
the two sets of figures were not far apart it would appear that the two
processes were taking place at about equal rates ; when the "relative"

TABLE 4. — Bile Pigments in Nineteen Cases of Pernicious Anemi.\

Average

Bile Pigments "Relative" Plasma Hemo- Red

in Duodenal Duodenal Bilirubin globin. Blood Cells

Case Contents Pigments Content per Cent. (Millions)

figure was much higher than the actual reading it would seem that
blood destruction was exceedingly active, and vice versa. Table 4
illustrates these points. The actual readings in the first column and
the "relative" figures in the second column are both obtained by averag-
ing the total pigment values of the six fractions obtained during duo-
denal drainage. The cases are arranged in order of magnitude of the
"relative" figures and, as noted above, this order closely approximated
the severity of the patient's clinical condition. Case 1, for example,
was a patient in a very severe relapse, while Case 19 was a patient in a
well marked remission with almost complete freedom from symptoms.
Chart 3 illustrates the marked increase in big pigment elimination in
these cases over the normal level, and further emphasizes the difference
between the actual and "relative" findings.

JOXES— BLOOD PIGMEXT METABOLISM 657

Mention has already been made that in these cases of pernicious
anemia the bile pigment in the plasma did not always exactly parallel
the pigment content in the bile, although it was always well above nor-
mal, as previously shown by Blankenhorn. Such a finding suggests
that this failure of the plasma bilirubin to parallel the bile pigments in
the duodenal contents may be due to an alteration in hepatic function.
It is pertinent, at this point, to call attention to a question only briefly
noted in the literature. The average level of bile pigments in the duo-
denal contents in this series of cases of pernicious anemia is by actual
reading 194 dilution units, or about five times the normal values.

Brie

Pigments
200

150

100

50

0

Fractions

/

/

'\

//
//
/ / ,

\

\
\ \

\ /

Ay

\

^

.— — '

'

— -

2t

0-0-0-0-0-0.

-Duodenal pigments i
. .'Actual pigment

•Relative" pigment values
-. Normal pigment values

Individual cases went as high as eight to nine times the normal level.
Such an increase over normal has previously been considered as due
entirely to increased blood destruction, with excessive liberation of
hemoglobin. Undoubtedly some part of the excess of bile pigments is
due to the products of increased red cell destruction, but there is in
addition a second factor which must be taken into account. Compari-
son with the observations on paroxysmal hemoglobinuria already men-
tioned makes this clear. One of the artificially produced attacks of
hemoglobinemia was accompanied by a lowering of the red cell count
by as much as 850,000 cells per c.mm. The hemoglobin liberated into
the circulation was presumably derived solely from the destruction of

658 ARCHIVES OF IXTERXAL MEDICIXE

red cells, a destruction- which approximated more than one-tenth of the
total number of red cells, as measured by a routine red count. This
excessive destruction of blood was followed shortly afterward by a rise
in the bile pigments in the duodenum to a level of about 300 dilution
units, or between six and eight times the normal figures. In parox-
ysmal hemoglobinuria there is no known evidence of any liver injury.
Deranged liver function, therefore, need not be considered in parox-
ysmal hemoglobinuria, and any increase in bile pigments can safely be
attributed essentially to increased blood destruction. In this example
of pure hemolysis, uncomplicated by any other factors, a drop in the
red cell count of 850,000 was accompanied by a rise in duodenal pig-
ments to a level of about 300 dilution units. This level of bile pigments
was well above the average of the entire series, and was but little under
the level found in the most severe cases of pernicious anemia. It is
difficult to conceive, even in the most severe cases of pernicious anemia,
or in any other so-called hemolytic disease, that there is a constant rate

TABLE 5. — Bile Pigments in Other "Hemolytic" Dise.^ses

Average
Bile Pigments "Relative"

in Duodenal Duodenal

Contents Pigments Content

Plasma Hemo- R<'i1

irubin globin. Blooil Cells

per Cent. (MiUioas)

Hemolytic jaundice:

of blood destruction going on so rapidly as to cause in a few minutes
the dissolution of more than one-tenth of the total blood corpuscles
in the body. A process causing such a degree of blood destruction, in
the absence of a correspondingly rapid degree of blood formation,
ought to result in complete exsanguination in a very short space of
time. A second factor seems necessary to help explain the high level
of bile pigments found in pernicious anemia. Ashby ^* has recently
reached a somewhat similar concltision. This second factor, I believe,
lies in a marked impairment of liver function. Such an assumption,
although suggested by Hooper and Whipple as a result of their work
on dogs, has not been made as a result of observations in man.

In addition to the above cases of pernicious anemia, a number of
cases were studied in which it is also usually agreed that there exists
an abnormally high degree of blood destruction. This series contained
two cases of malaria and four cases of acquired hemolytic jaundice.
One of the latter had previously had his spleen removed. All of these

18. Ashby, W.: J. E.xper. M. 34:147, 1921.

JOXES— BLOOD PIGMEST METABOLISM

659

cases showed high pigment values (Table 5) entirely comparable with
those observed in pernicious anemia. As in the former cases, the
"relative" figures gave the more accurate picture, and closely paralleled
the clinical condition of the patients. In these cases, also, as in perni-
cious anemia, it seems reasonable to assume that there must be a
second factor to account for the extremely high bile pigment values
obtained. Blood destruction alone could hardly account for the
increased bile pigment elimination. In two of the cases of acquired
hemolytic jaundice in this series (Table 5, Cases 30 and 31) the liver
enlargement was so marked indeed as to dominate the entire clinical

Bfle
Pigments

200

/
/

\
\

/

\

150
100

/
/

\

/
/
/

\

\

\

/' //

//
//

\

'\

\

50

.-'^^

/

\

0

^

Chart 5. — Duodenal pigments in gallbladder disease. Pigment values in

cholecystitis ■ . Pigment values in cholelithiasis . Normal

pigment values o-o-o-o-o-o-o.

picture and to suggest definitely that a hemolytic process was not
responsible for the entire condition. Impairment of liver function in
these cases would also seem to be the logical additional factor to be
considered, and such an assumption would seem to be even more
logical in the cases of hemolytic jaundice than in the cases of perni-
cious anemia.

Bile Pigments and Impaired Liver Function. — In the type of cases
already studied, in which the process of abnormal blood destruction

660 ARCHIJ'ES OF IXTERXAL MEDICIXE

has been of long duration, it is well recognized that clinically, and at
post-mortem examination, it is commonly possible to demonstrate liver
pathology. Forty per cent of cases of pernicious anemia have during
life palpable livers,^" which at autopsy show a certain amount of
fatty infiltration and deposits of iron containing pigment. Cases of
hemolytic jaundice usually, and malaria not infrequently, present the
clinical evidence of liver enlargement. The suggestion has already
been made that in the above type of disease the abnormally high
elimination of bile pigments may be due in part to an alteration of
liver function. It is furthermore reasonable to assume, even in the
absence of any clinical signs of liver derangement, that there exists a
marked alteration in liver function, due merely to the presence of a
severe anemia. Such an assumption finds support in the well-recog-
nized fact that in cases of severe anemia the kidneys may show
evidence of definite alteration of function by renal function tests and
by the presence of albumin in the urine. With improvement in the
anemia, the renal function also improves. It is, therefore, appropriate
to turn from a consideration of the so-called hemolytic diseases, in
which abnormal blood destruction and liver damage may be accom-
panying factors, to the study of a series of cases in which it is evident
that the liver is the primary seat of disease, and in which there is
considered to be little or no question of abnormal hemolysis.

A series of eighteen cases was examined, all of which presented
clinical evidence of moderate to severe liver damage. This group
included cases of carcinoma of the liver, cirrhosis, either alcoholic or
syphilitic, hepatitis or cirrhosis due to a local or general infectious
process, and Banti's disease. These types of cases are not usually con-
sidered to have any important degree of increased blood destruction,
with the possible exception of Banti's disease. These cases, therefore,
were of interest as a basis for studying the functional capacity of the
liver, in terms of hemoglobin and bile pigment metal)olism. The find-
ings are tabulated and shown in Table 6 and Chart 4.

Six of the eighteen patients were clinically jaundiced, and nine
more showed a "potential" jaundice. By the term "potential" jaundice
is meant a condition in which the bilirubin content of the blood plasma
is abnormally high but not sufficiently high to cause tissue icterus. As
has been shown by Blankenhorn,^'* in certain chronic diseases in which
there is a continual abnormal elimination of bilirubin, the concentra-
tion of this pigment in the blood plasma may be many times normal
without causing tissue icterus. In spite of the presence of jaundice,
either actual or "potential," in fifteen out of these eighteen cases the
output of bile pigments into the duodenum was above normal. In the

19. Minot, G. R.: O.xford Medicine, 2:623.

JOSES— BLOOD PIGMEXT METABOLISM 661

remaining three cases, the concentration of the bile pigments in the
bile was within normal limits. Although some clinical abnormality in
the liver was demonstrable, the abnormality may not have progressed
far enough to overstep the large factor of safety present in this organ.

In fifteen out of eighteen cases, therefore, the liver, even when
almost entirely invaded by foreign tissue, tended not to eliminate a
decreased amount of bile pigment into the duodenum, but on the con-
trary to put out highly concentrated bile, as measured by pigment
content. The actual pigment readings in the series of eighteen cases
of liver disease showed an average curve (Chart 4), the values of
which were over twice the normal level. From these facts alone, there-
fore, it is evident that the jaundice present in these cases was not at
all obstructive in nature, at least in the usual interpretation, namely
a diminished output of bile into the duodenum with accumulation of
the residue in the blood and tissues.

Neither was the jaundice strictly hematogenous. The abnormal
elimination of bile pigments, both in the bile and plasma, can not be
said to be due to an increased amount of blood destruction. The
average red cell cotmt of the entire series was 3,'900,000 corpuscles per
c.mm., and one of, the more severe of the cases, one of advanced car-
cinoma of the liver, showed a red count as high as 6,000,000 per c.mm.
The red cells were markedly achromic, and the average hemoglobin
content of the blood was approximately 60 per cent. The average color
index of the cases was only 0.75, in contrast to the average index of 1.5
observed in the cases of pernicious anemia. In the stained smear the
red cells showed only achromia, with slight variations in size and
shape. There was lacking microscopic evidence of increased blood
destruction; namely, the presence of microcytes or fragmentation of
the red cells. Active blood formation, commonly present in the face
of active blood destruction, was not seen, at least as represented by
noteworthy changes in the number of young red cells. The usual con-
ception of these types of liver disturbance, furthermore, does not
associate them with increased blood destruction. Banti's disease, by
some clinicians, is occasionally associated with abnormal destruction of
the red corpuscles. The commonly accepted view, however, is that
expressed by Krumbhaar,=" who states that blood destruction is not the
important element in this disease. It seems fair to assume, therefore,
from the generally accepted views, that in these cases under discussion
blood destruction was not importantly increased, and that the jaundice
is not hematogenous in nature.

In the literature scant reference has been made to the bile pigment
excretion in the above type of case. The general statement has usually

20. Krumbhaar, E. B. : Nelson's Loose Leaf Living Medicine, 4:37,

662 ARCHIVES OF IXTERXAL MEDICINE

been made that secondary anemias, in contrast to primary anemias, are
accompanied by a diminished elimination of bile pigments. It is
clearly .evident, however, from the results charted in Table 6, that
cases of anemia associated with liver disturbance are usually accom-
panied by increased bile pigment excretion. The so-called secondary
anemias, if uncomplicated by liver disturbance, undoubtedly yield low
pigment readings. When, however, in the course of such a secondary
anemia the underlying cause afifects the liver the entire picture of bile
pigment excretion is changed, and instead of a low pigment elimina-
tion in the bile and plasma, there follows a complete reversion of the
physiological processes involved, and the bile pigments reach a new
and abnormally high level.

TABLE 6.— Bile Pigments in C.\ses of Liver Disease

Case
Cancer:

Average

Bile Pigments

in Duodenal

Contents

"Relative"
Duodenal
Pigments

144
20O
213

189

m

96
51

174

87
96
35

248

55

280

114

Plasma
Bilirubin
Content

40

80
15
40

30
50

140
66

28

25
]0O

10

15
250

30
l.W
69

Hemo-
globin,
per Cent.

85
63

25
80

66

SO
75

65
40

75
45
70

67
72

80
66

6.0

144

3.6

2.3

Cirrhosis (alcoholicl :

54-

151

4.0
5.3

36

Cirrhosis (syphilitic):

77

3.9

41

Hepatitis (infectious):

115

... 78

63 (typhoid)

64

.. 30

30

4.3

4.2

Hepatitis (toxic):

58

6.3

Banti's disease:

4.1

Avera -e

101

4.1

• Clinically Jaundiced.

It has already been shown that the jaundice occurring in such cases
is neither strictly obstructive nor hematogenous in nature, in spite of
the fact that in all the cases the liver parenchyma was severely
damaged. It is well known that the liver, like the other organs, has a
large factor of safety, as regards all of its functions. Exact informa-
tion as to the extent of this measure of safety has never been deter-
mined in man. McMaster and Rous "^ have recently shown that the
bile ducts from three-quarters of the liver substance can be obstructed
in dogs and monkeys without the development of any clinical evidence
of pigment or chelate accumulation in the organism. They also

21. McMaster and Rous: J. E.xpcr. M. 33:7,31, I92L

JOKES— BLOOD PIGMEXT METABOLISM 663

showed that in the dog nineteen-twentieths of the Hver substance can
be placed in a condition of stasis, without the occurrence of tissue
icterus such as regularly follows total obstruction in this animal. In
their experiments, they found that invariably a local obstruction
resulted sooner or later in atrophy of the affected tissue, with com-
pensatory hypertrophy elsewhere. Their conclusions are of particular
interest in the present discussion: "the clinical jaundice encountered in
association with local liver lesions should be viewed, not as the result
of local bile absorption, but as due to a general injury to the hepatic
parenchyma or ducts, or to blood destruction." Such injury with its
resulting hypertrophy, would accordingly result in functional changes,
and bile pigment excretion would accordingly be modified. The nature
of the cause of this functional disturbance is apparently not specific.
An examination of the accompanying table will show that in no
particular group of liver conditions was there any predominance of
high pigment values. Bile pigment excretion was apparently iniluenced
neither by the nature of the process, nor by the amount of the anemia.
The mechanism is probably similar in all the cases, and the degree of
derangement of hepatic function is solely dependent on the extent and
rapidity of the disease process.

There remains, then, to discuss the actual nature of this alteration
in liver function. As already noted, there was a marked increase in the
actual amount of bile pigments eliminated by the liver. In the absence
of any abnormal process of blood destruction the source of the exces-
sive amounts of bile pigments is still to be determined. Changes in
diet, according to Hooper and Whipple,^ can cause marked alterations
in bile pigment elimination in animals. Such a factor, however, can
readily be excluded in the present series. The most logical explanation
seems to be the following: Under normal conditions the liver is the
principal agent in the metabolism of hemoglobin set free during the
normal processes of red cell destruction. This pigment metabolism
involves the breakdown into less complex molecules, through bilirubin
and biliverdin, to the lower derivatives, urobilinogen and urobilin. The
formation of bilirubin from hemoglobin may take place in the blood
vessels and tissues without any intervention on the part of the liver,
and similarly, urobilin is undoubtedly formed in the intestine by the
action of bacteria on bilirubin. It is highly probable, however, that
the liver itself is capable of breaking down the bilirubin into urobilin,
without the intervention of the intestine. The observations already
mentioned made on cases of paroxysmal hemoglobinuria suggest such
a possibility. Furthermore, the liver has long been thought capable
of resynthesizing hemoglobin from the lower bile pigments by building
them up to more complex molecules and combining them with the iron

664 ARCHIl-ES OF IXTERXAL MEDICI \E

known to be retained by the liver. Such a process of resynthesis is
entirely analagous to the general physiologic properties of all human
cells and is not necessarily much more complicated than the formation
of urea or glycogen from lower chemical constituents. The process
of breaking down hemoglobin into its lower derivatives is, however,
probably a less difficult matter than the subsequent resynthesis of
hemoglobin from bile pigments. The latter function would perhaps
logically be the first to be altered or lost. With the failure of the
normal resynthesis of hemoglobin from bile pigments the unaltered bile
pigments would then form an excess and would be eliminated as such
in the bile.

The findings in this group of cases seem to confirm this supposition.
The loss of resynthesizing power in a damaged liver would, of course,
be only partial. The lowered formation of hemoglobin ought even-
tually to be reflected in a diminished hemoglobin content of the red
cells with resulting low color index and achromia. In all these cases
marked achromia of the red cells and a low color index occurred. The
hemoglobin averaged 66 per cent., and the red count averaged 4,100,000
per c.mm. This slight diminution was possibly the result of a gradual
slowing up of bone marrow activity. Such findings may be regarded
as probable evidence of a diminished production of hemoglobin. That
portion of the bile pigments not resynthesized into hemoglobin would
be excreted as such, and would account for the increased elimination
of bile pigments, even in the face of normal blood destruction. That
such a theory further corresponds with the actual findings in the
individual cases is attested by the fact that in the majority of cases
showing the greatest reduction of hemoglobin content there was a i)ro-
portionally high level of bile pigments in the duodenum. One case, for
example, with a hemoglobin content of 40 per cent, and a color index
of 0.35, showed a bile pigment elimination in the bile of over four
times the normal.

The above theory would satisfactorily account for the appearance
of jaundice and lowered hemoglobin content so frequently noticed in
the course of acute infections such as pneumonia, typhoid, scarlet
fever, septicemias, etc. In such conditions the infection, or the accom-
paning toxemia, may be assumed to cause a temporary alteration of
the liver function, with resulting alterations in hemoglobin metabolism.
The icterus frequently accompanying severely decompensated heart
disease may also be explained on the basis of altered liver function.

In a severely damaged liver not only should there be an increase in
the actual amount of bile pigments eliminated, but the relation of the
various pigment elements in the bile .should be distinctly altered. Those
[Mgments most easily formed ought to be excreted at once instead of

JOXES— BLOOD PIGMEXT METABOLISM 665

being completely broken down to the lower forms. In confirmation is
the frequent occurrence of excessive amounts of intermediate bile pig-
ments— cholecyanin and urobilinogen — found in the duodenal contents
in this series of cases of excessive red cell destruction, and those with
liver disease. Schneider and others have already noted the presence
of large amounts of urobilinogen in the duodenal contents in severe
cases of blood destruction. This excess of .urobilinogen was noted,
not only in the present series of cases with liver disease, but also in
those cases in which there was pathologic blood destruction. In addi-
tion, in those cases of severe anemia, the presence of cholecyanin was
observed, frequently in large amounts. This latter pigment, as well as
urobilinogen, is intermediate between bilirubin and urobilin, and its
presence would seem to indicate very rapid and incomplete metabolism
of hemoglobin derivatives by the liver. The presence, therefore, of
these intermediate pigments in excess in cases of pernicious anemia
would seem further evidence of liver damage in this disease.

It is therefore, tempting to assume that a disturbed function of the
liver in the disease pernicious anemia is a considerable factor in creat-
ing abnormal pigment values in the plasma and in the bile. It is not
as easy, howevef, to apply this explanation to pernicious anemia as to
the other anemias in which abnormal pigment values are found. In
pernicious anemia there is a relative increase in hemoglobin and iron
pigment is found in various organs. The irregular and biazarre course
of pernicious anemia, the attractive assumption that the red corpuscles
in pernicious anemia are not only abnormal but different from the red
corpuscles in other conditions, may account in part for the seeming
discrepancy. The existence of dift'erence between the red corpuscles
in pernicious anemia and in other conditions has been indicated by
work recently done on various types of anemia by Buckman -^ at the
Boston City Hospital. In any event, while a part of the increase in
bile pigment elimination in pernicious anemia may be attributed to
excessive blood destruction, the remainder may perhaps be laid to a
damaged liver. Inasmuch as continued attacks of hemolysis per se
cause liver damage, the two factors are really related.

Furthermore, as pointed out by Brule,-" the proper conception of
such a disease as catarrhal jaundice should locate the primary pathol-
ogy not in the biliary passages but in the hepatic cell itself. Such a
disease is primarily an infection of the liver parenchyma, and the
pathology and abnormal physiology should be centered in the degree
of actual parenchymal damage and disturbance of liver function. Such
a conception of catarrhal jaundice offers the logical explanation of the

22. Buckman, T. E. : Persoi

23. Brule, M. : Bull. med. 8:279, 1920.

666 ARCHIVES OF IXTERXAL MEDICIXE

diminished hemoglobin frequently foimd after an attack of even mod-
erate severity, and accounts for the increased amounts of bile pigment
eliminated in the bile after the flow is reestablished.

Findings in Cases of Gallbladder Disease. — In view of the frequent
association of liver disturbance with chronic disease of the gallbladder
it is pertinent at this point to examine briefly the results obtained in a
series of cases of cholelithiasis and cholecystitis. Observations were
made on ten patients suffering from typical gallstone attacks and on
six patients with typical symptoms of chronic cholecystitis. In a majority
of the cases the preliminary diagnosis was confirmed by subsequent
operation. The pigment curves, as shown in Chart 5, are easily
explained. They did not vary from the normal curve in their general
contour. The actual level of the bile pigments, however, was dis-
tinctly higher than normal. (This series, of course, did not include
cases of cystic or common duct obstruction.) The point of interest in
these cases is that there was a distinct difference between the pigment
values obtained from patients with stones and those obtained from
patients with only cholecystitis. Those patients with cholelithiasis gave
an average pigment curve approximately 75 per cent, above the normal
level, although in individual cases the average figure 'was as high as
three times normal. The cases of cholecystitis, on the contrary, gave
a distinctly higher average. The average pigment values in this group
were nearly twice those observed in cases with stone formation, and
and were three times the normal figure. Individual cases of this group
went as high as eight times normal. Furthermore, the peak of the
pigment cure representing the greatest concentration of gallbladder
bile was on the average more than twice that found in the group of
cases with stones. According to the present conception of gallbladder
disease, blood destruction does not pl&y an important part in the dis-
ease process. The high pigment values, therefore, were due either to
abnormal stasis or to liver pathology. The work of Rous and of
McMaster,^* recently published, indicates that in stasis the gallbladder
has a great power of concentration, with the result that any bile con-
tained in it, even for short periods of time, becomes abnormally high
in pigment and other constituents. They show that this power of con-
centration diminishes in the face of a pathological process such as the
presence of stone formation with partial or complete obstruction to the
normal entrance of bile into the gallbladder. The high pigment content
in cases of cholecystitis may thus be partially explained as well as the
dift'erence between those cases with stone formation and those with
only low grade gallbladder inflammation leading merely to stasis. It
is of especial interest in this consideration to note the findings reported

24. Rous and McMaster: J. Expcr. M. 34:47, 75, 1921.

JOXES— BLOOD PICMEXT METABOLISM 667

in a personal communication from Fitz ^^ from the Mayo Clinic. He
reports finding highly pigmented bile in most cases in which operation
was performed for cholecystitis, whereas the bile in those cases show-
ing calculus formation was also dark but less highly pigmented.
Furthermore, he was able to demonstrate that the specific gravity and
the nitrogenous content of those cases with only cholecystitis tended
to be much higher than in cases of cholelithiasis. Such findings are
strongly confirmatory of the concentrating ability of the gallbladder
in cholecystitis and help to explain the pigment values found in this
group of cases.

While the high pigment values in these cases are undoubtedly due
in part to gallbladder concentration, it is also highly probable that they
may be due in part to an accompanying cholangeitis and hepatitis. The
recent paper by Judd,^" emphasizing the common association of gall-
bladder and liver infection, is also confirmatory. Such a conception
would also explain the low hemoglobin content frequently found in
connection with long standing cases of cholecystitis, and occasionally
persisting even after cholecystectomy. The high pigment values in
such cases are probably due both to abnormal gallbladder concentra-
tion and to an alteration in liver function.

CONCLUSIONS

1. Increased blood destruction is accompanied by an increase of
the bile pigments in the blood plasma and bile.

2. Alterations in liver function, due to infection, new growth,
cirrhosis, or even a profound anemia per se, are also accompanied by
marked increases in bile pigment, both in the bile and plasma.

3. Jaundice, in cases with liver damage, may be entirely due to an
alteration in bile pigment metabolism, without the necessity of any
accompanying obstructive process or increase in the normal process of
blood destruction.

4. The high level of the bile pigment in pernicious anemia can not
be due solely to a process of increased blood destruction. A second
factor is necessary to explain the increased pigment elimination. This
second factor may well be an alteration in hepatic function.

5. In gallbladder disease the bile pigments in the duodenal con-
tents are abnormally high, especially in those fractions containing the
greatest concentration of bile from the gallbladder.

6. Cases of uncomplicated cholecystitis show a greater concentra-
tion of bile pigments than cases of cholelithiasis.

25. Fitz, R. : Personal communication.

26. Judd, E. S.: J. A. M. A. 77:197 (July 16) 1921.

668 ARCHll'ES OF IXTERXAL MEDICIXE

7. A presumable functional incapacity of the liver properly to
metabolize hemoglobin, due to any cause resulting in liver damage, is
accompanied by a lowered hemoglobin content of the blood. Such
cases also show high bile pigment values in the plasma and bile.

8. Owing to the frequent association of hepatitis with cholecystitis
it is probable that the frequent accompaniment of a low hemoglobin
content and an apparent anemia in chronic gallbladder diseases is due
to an alteration in liver function.

For numerous valuable suggestions and criticisms offered during the course
of this work I am greatly indebted to Dr. Roger I. Lee and Dr. George R.
Minot. I also wish to acknowledge the cooperation of Dr. Thomas E. Buckman
in allowing me access to the clinical material of the Boston City Hospital.

A STUDY OF HEMOGLOBIN METABOLISM IN
PAROXYSMAL HEMOGLOBINURIA

WITH OBSERVATIONS OX THE EXTRAHEPATIC FORMATION OF
BILE PIGMENTS IX MAX *

CHESTER M. JONES, M.D.

AND

BASIL B. JOXES, M.D.

BOSTON

The close relationship of the coloring matter of the red blood cor-
puscles to the pigments of the bile is now generally accepted. Proofs
for the existence of this relationship are numerous, and have been
obtained by many investigators. Hematoidin, chemically isomeric with
bilirubin, has long been known to occur in old extravasations of blood.
Stadelmann,^ working on dogs with biliary fistulas, showed that free
hemoglobin in the plasma, produced by the artificial destruction of red
cells, or by the injection of hemoglobin into the circulation, caused an
increase in the quantity of bilirubin in the bile. Stadelmann and Goro-
decki,^ by injecting a solution of hemoglobin either subcutaneously or
intraperitoneally, also caused in dogs a marked and prolonged rise in
bile pigment in the fistula bile. The work of Brusch and Yoshimoto,^
on dogs with biliary fistulas and ligated bile ducts, showed that intra-
venous injections of hematin caused increased amounts of bilirubin
and urobilin to appear in the bile. In man the excretion of bile pig-
ments in various so-called hemolytic conditions has been studied by
numerous investigators. These conditions are found in pernicious
anemia, hemolytic jaundice, malaria, etc. Hoppe-Seyler,* Gerhardt and
von Miiller,"' Eppinger and Charnas,''' de Jonge," Simpson,* Robertson,"

* From the Medical Service of the Massachusetts General Hospital.

* This paper is No. 23 of a series of articles on the physiology and pathology
of the blood from the Harvard Medical School and allied hospitals, a part
of the expense of which has been defrayed from a grant from the Proctor
Fund of the Harvard Medical School for the study of chronic di.sease.

1. Stadelmann: Der Icterus und seine verschiedene Formen. Stuttgart, 1891.

2. Stadelmann and Gorodecki : Ibid.

3. Brusch and Yoshimoto: Ztschr. f. exper. Path. u. Therap. 8:639, 1910.

4. Hoppe-Seyler : Virchows Arch. f. path. .'Xnat. 30:124, 1891.

5. Gerhardt and von Miiller: Ztschr. f. klin. Med. 32:. 303, 1897.

6. Eppinger and Charnas: Arch. f. klin. Med. 78:387, 1913.

7. de Jonge: Geneesk. Tijdschr. v. Nederl. Ind. 44:433, 1904.

8. Simpson, G. C. E. : Biochem. J. 5:.378. 1911.

9. Robertson, O. : Arch. Int. Med. 15:1072 (June) 1915.

670 ARCHIVES OF IXTERXAL MEDICIXE

Wilbur and Addis/" and others have investigated the urine and feces
for urobilin in human beings in normal and diseased conditions. In
cases in which there was apparently a rapid rate of blood destruction
these observers consistently noted abnormally large amounts of urobilin
in the feces, and usually in the urine. Simpson reported one case of
malaria which showed hematoporphyrin, as well as large amounts of
urobilin in the stool examinations. Schneider," Giffin, Sanford and
Szlapka,^^ and Jones ^^ and subsequent workers have demonstrated
abnormally large amounts of bilirubin, urobilin and urobilinogen in the
duodenal contents, particularly in cases of pernicious anemia, and other
conditions in which blood destruction is considered to be excessive.
The bilirubin content of the blood plasma has been shown by Blanken-
horn,^* among others, to be definitely increased in diseases in which it
is usually considered that pathologic blood destruction is taking place.
From the literature it is clear that a large amount of evidence has
been obtained as proof of the close relationship between liberated
hemoglobin and the bile pigments in the blood, bile and excreta. It is
true, however, that the relationship between the amount of hemoglobin
liberated during the process of blood destruction at any given time
and the bile pigments in the blood and bile, as measured by present
laboratory methods, may not be strictly a quantitative one. On the
contrary, although the level of bile pigments in the blood and bile
undoubtedly may serve as an index or measure of blood destruction,
it is, nevertheless, only a relative measure of hemolysis, and may be
influenced by various factors. Hooper and Whipple, ^° have shown in
dogs, that mere changes in diet can produce large variations in bile
pigment excretion. In dogs with biliary fistulas, a high carbohydrate
intake resulted, and in some cases an increase in fistula bilirubin as
high as 100 per cent, above normal, while a meat diet alone gave a
slight decrease in pigment values. These observers also showed that
the functional capacity of the liver, in dogs, in large measure influenced
the output of bile pigments. Dogs with Eck fistulas, and accordingly
with damaged livers, showed a marked diminution in bile pigment
elimination. On the contrary, Jones " was able to demonstrate, in a
series of patients suffering from various types of liver disease, a
marked increase in the output of the bile pigments, both in the blood
and the bile. The use of certain drugs is also known to modify bile
pigment excretion.

10. Wilbur and Addis: .-Vrch. Int. Med. 13:325 (March) 1914.

11. Schneider, J. P.: Arch. Int. Med. 17:32 (Jan.) 1916.

12. Giffin, Sanford, and Szlapka: Am. J. M. Sc. 182:562, 1918.

13. Tones, C. M. : This issue, p. 643.

14. Blankenhorn, M. A.: Arch. Int. Med. 19:344 (March) 1917.
15 Hooper and Whipple: Am. J. Phvsiol. 40:332, 349. 1916.

JOyES-JOXES—PAROYXSMAL HEMOGLOBIXURIA 671

It has also been shown by recent investigators that the early views
regarding the mechanism of hemoglobin metabolism must be modified
considerably. The liver, and to some extent the spleen, were originally
held to be the sole agents concerned with the breakdown of hemoglobin.
The liver alone was held responsible for the reconstruction of hemo-
globin from iron and the lower pigment molecules. Recent work, how-
ever, has shown that, whereas the liver undoubtedly plays the most
important role in hemoglobin metabolism under normal conditions, this
function, in animals, can be assumed, in part, if not entirely, by other
organs and tissues of the body. Hooper and Whipple '^ were able to
show in dogs with bile fistulas that the bile pigments could be formed
just as readily in animals in which the circulation of the liver was
greatly diminished by an Eck fistula, or with such a fistula and the
hepatic artery ligated, as in normal animals. Injection of hemolyzed
red blood corpuscles into a head-thorax circulation resulted in the
appearance of increased bile pigments in the blood. Bile pigments
were also found in increased amounts in the blood when free
hemoglobin was injected into the pleural or abdominal cavities.
Van de Bergh ^^ has also sliown in man, in four cases subjected to
operation, and in two necropsies, that the blood serum from the splenic
vein contained much more bilirubin than did the peripheral blood.
These patients had pernicious anemia, Banti's disease, hemolytic jaun-
dice, etc. He concludes that the excess bilirubin in the splenic blood
was formed by the spleen. The views recently expressed by Pearce,
Krumbhaar and Frazier," that the spleen is an important factor in
the process of blood destruction, are very generally accepted. In fetal
life the spleen has the power of extensive blood formation. In the
adult the spleen may undergo, in the presence of an injury to the bone
marrow, a myeloid metaplasia; i. e., it can regain its fetal function
under pathologic conditions. Whether the spleen may exert this power
of blood formation in adult life under normal conditions is doubtful,
though it is still an open question.

It thus has been shown in animals that hemoglobin katabolism can
take place without the intervention of the liver, and that the spleen, in
man, may carry on this function. In addition, it has been shown by
experimental work on animals, that there is a close relation between the
excretion of the bile pigments and the liberation of hemoglobin in the
blood. The investigations carried out in man in such diseases as per-
nicious anemia, hemolytic jaundice, malaria, etc., have produced fur-

16. van de Bergh, A. A. H. : Nederlandsch. Tijdschr. v. Geneesk. 1:1160
191S.

17. Pearce, Krumbhaar and Frazier: The Spleen in Anemia, Philadelphia
1918.

672 ARCHirES OF IXTERXAL MEDICIXE

ther evidence of the derivation of the bile pigments from the products
of hemolysis. Experimental evidence in man, however, in which mea-
surable hemolysis has been produced in the circulation, with an imme-
diate and continuous examination of the bile, and blood, is conspicuously
lacking.

It is our purpose, in this paper, to present further evidence regarding
the katabolism of hemoglobin in man. It will be shown, first, that there
is a direct relation between the liberation of hemoglobin into the plasma
and the excretion of the bile piginents ; second, that the liver is stimu-
lated to increased functional activity by abnormal intravascular hemol-
ysis ; and third, that hemoglobin can be broken down in the blood
vessels, capillaries and tissue spaces without the intervention of the
liver or other organs.

The experiments which form the basis of this paper were conducted
on two patients sufifering from paroxysmal hemoglobinuria. It will be
pertinent to insert at this point the following definition of this rare
disease : "Paroxysmal hemoglobintiria is a chronic disease, due to
syphilitic infection, manifesting itself in recurrent paroxysms of hemo-
globinuria, and in characteristic constitutional symptoms. The blood
of patients who suft'er with this disease contains in latent form a specific
hemolysin which becomes active when the blood is chilled, and produces
the attacks." ^'* Chilling of the blood to a temperature below 15 C.
causes this specific hemolysin to become attached to the red cells.
During subsequent warming at body temperature {37.?' C.) the
hemolysin becomes active through the influence of the complement
normally present in the blood, and hemolysis ensues. The amount of
hemolysis is readily measured by the amount of free hemoglobin in
the plasma, and is dependent on the length and severity of the chilling
to which the blood is exposed.

One of the two cases (G. L. T) exhibited all the characteristic
features of this striking disease. The patient was a congenital syphi-
litic. He showed in his blood the presence of the specific hemolysin
peculiar to the disease, and was subject to attacks of hemoglobinuria
on exposure to chilling. Antisyphilitic treatment had modified the
course of the disease process, but at the time of our experiments the
treatment had not been sufficient to free the patient from attacks. The
second patient (A. L.) was also a congenital syphilitic. His sister had
paroxysmal hemoglobinuria. Unlike his sister, however, he had never
had an attack at all suggestive of this disease. His blood, nevertheless,
showed the presence of the characteristic hemolysin, and both in vivo
and in vitro, it was possible, by exposure to chilling, to demonstrate

18. Jones, B. B., and Jones, C. M. : Nelson's Loose Leaf Living Medii
Washington, 1920.

JOXES-JOXES—PAROYXSMAL HEMOGLOBIXURIA 673

definite hemolysis. This second case may thus be classified as a
"potential" case of paroxysmal hemoglobinuria. It is of interest to
note in this connection, that, in the blood of forty-five syphilitic patients
who had no clinical .symptoms of paroxysmal hemoglobinuria, we were
able to demonstrate a similar hemolysin definitely in 6.6 per cent., thus
approximately confirming the observations of Donath and Land-
steiner," Kumagai and Inoue,-" and others.

The experiments conducted on these two patients consisted in the
production of attacks of hemoglobinemia, with a subsequent study of
the pigments in the blood, duodenal contents and urine. The condition
studied was thus essentially intravascular hemolysis in man, uncom-
plicated by any other factor, such as liver damage, trauma, or by the
introduction of any foreign substance into the circulation. Attacks of
hemoglobinemia were brought about by immersing one or both of the
patient's hands in ice water for several minutes, and then warming the
chilled members. In none of the experiments was the chilling severe
enough to cause more than a trace of hemoglobin to appear in the urine,
and in two of the e.xperiments hemoglobinemia only was. produced.

Blood for examination was taken from an arm vein, either into a
small amount of potassium oxalate solution, to prevent clotting, or
allowed to clot at body temperature over a water bath. Special care
was taken to prevent the occurrence of mechanical hemolysis.

The presence and amount of free hemoglobin in the blood was
determined by spectroscopic examination of the oxalated plasma. The
The number of dilutions with distilled water necessary to cause the dis-
appearance of the characteristic absorption bands of hemoglobin from
the spectrum was taken as the amount of free hemoglobin present in
any given specimen of plasma. Inasmuch as bilirubin gives no charac-
teristic absorption band in the spectroscope, its concentration in the
blood plasma was determined by the methed described by Blanken-
horn,'^ which consists merely in the dilution of the plasma or serum
with distilled water until the yellow color disappears on comparison
with a tube of distilled water. Here also the amount of bilirubin
present was taken as the number of dilutions necessary to cause the
disappearance of the yellow color. As a further test for bilirubin in
the plasma, the Gmelin test with nitric acid was used. This test gives
a positive reaction in the presence of relatively large amounts of bili-
rubin, but gives no characteristic color changes when hemoglobin alone
is present. The Gmelin test, therefore, was used as a second method
for determining the presence of abnormal amounts of bilirubin in the
plasma or serum, especially when the characteristic yellow color was

19. Donath and Landsteiner : Ztsclir. f. klin. Med. 58:173, 1905.

20. Kumagai and Inoue: Miinch. nied. Wchnschr. 38:361, 1912.

674 ARCHIVES OF IXTERXAL MEDICIXE

obscured by the additional presence of free hemoglobin. A positive
Gmehn test is not obtained with specimens of normal blood. The
presence of any quantity of free hemoglobin in the serum or plasma
made it impossible to estimate accurately the amount of yellow pigment
present. In such cases, when the yellow color was obscured by the
presence of free hemoglobin, dilutions were carried out with distilled
water until no color was left. A sample of the same plasma was also
examined spectroscopically for free hemoglobin. Comparison between
the amount of hemoglobin present, and the number of dilutions neces-
sary to remove all traces of pigment, gave a rough estimation of the
amount of bile pigment present. Inasmuch as the estimation of the
amount of free hemoglobin and bilirubin in the plasma or serum were
made by essentially different methods, it is obvious that any compari-
son between such estimations must be only an approximate one. The
same holds true of any comparison made between the bilirubin and
urobilin concentration in the duodenal contents, as will be pointed out.
Other pigments than hemoglobin and bilirubin were not found in the
blood.

The duodenal contents were obtained through an Einhorn tube,
and were examined for the presence of the various bile pigments —
bilirubin, urobilin, urobilinogen, cholecyanin, etc. Only the first three
pigments were found. The position of the tip of the tube in the duo-
denum was confirmed by fluoroscope, and a free flow of bile was
obtained by the use of a solution of magnesium sulphate, as described
by Lyon.^^ Bilirubin values were obtained by a colorimetric method
described by Hooper and Whipple ^^ in their work on dogs with biliarj'
fistulas. Briefly stated, their method consists in the treatment of the
bile drainage with acid alcohol, and reading the resulting blue-green
solution against a standard solution of copper sulphate and India ink
in a Duboscq colorimeter. Urobilin and urobilinogen values were
estimated by the method used by Wilbur and A.ddis, Schneider, and
others. The method consists in treating a given quantity of duodenal
contents (or feces) with an equal quantity of a saturated alcoholic solu-
tion of zinc acetate (Schlesinger's reagent), filtering, acidifying the
filtrate with Erlich's soltftion (paradimethylaminobenzaldehydro-
chlorid), and after allowing this to stand in the dark for fifteen
minutes, reading in the spectroscope. The number of dilutions with
ethyl alcohol necessary to cause the disappearance of the characteristic
absorption bands in the spectrum is taken as the value of the individual
pigments. Urobilin and urobilinogen values were added together and
expressed as one figure. As previously mentioned, the difference
between the methods of estimating bilinibin and the other bile pig-

21. Lyon, B. B. V.: J. A. M. A. 73:980 (Sept. 27) 1919.

lOXES-JOXES—PAROyXSMAL HEMOGLOBIXURIA 675

ments made any comparison between the different estimates purely an
approximate one. This method of studying the duodenal pigments is
fully described in a separate paper.'^

The presence of hemoglobin or urobilin in the urine was determined
by the use of a spectroscope. Other bile pigments were tested for but
none were found.

Measurement of the blood loss taking place following an attack of
hemoglobinuria was determined by a series of red cell counts taken
before and after the production of attacks. Such estimations were of
course somewhat inaccurate, inasmuch as they did not take into account
alterations in the peripheral circulation and the blood volume. For
practical considerations, however, such determinations indicated the
severity of an attack, as did the presence or absence of free hemoglobin
in the urine.

Three separate experiments were performed on the two patients.

PROTOCOLS OF EXPERIMENTS

Experiment 1. — Patient G. L. T.

The results of this experiment are clearly shown in Chart 1. Samples of
blood and urine were taken as normal controls before inducing an attack
of hemoglobinemia. There was no free hemoglobin in either the blood or the
urine, and the bilirubin content of the blood plasma was normal. Duodenal
contents were taken at fifteen minute intervals for an hour and a half, in order
to establish a normal pigment curve. The first peak shown in the chart came
after the introduction of a 33 per cent, solution of magnesium sulphate into the
duodenum, and was due partly to an increased flow of bile into the duodenum
through the relaxed sphincter of Oddi, and partly to the addition of some of
the concentrated bile in the gallbladder to the general flow, .-^fter the control
figures had been obtained both the patient's arms were immersed in ice-water
for five minutes, in order to produce intravascular hemolysis. Duodenal
contents were then collected at intervals during the next three hours and a half.
Samples of blood were also taken from an arm vein at varying intervals until
there was no further hemoglobin in the plasma, and until the bilirubin contents
of the plasma had practically returned to the normal level. Samples of urine
taken at different times during the experiment failed to show any increase
in bile pigment content, and at no time was there any evidence of hemoglobin.

Examination of Chart 1 shows clearly that there was an immediate and
sharp rise in the hemoglobin content of the blood plasma, although prior to
the attack there was no free hemoglobin present. Furthermore, before the
attack there was a normal amount of bilirubin in the plasma as measured by
its yellow color, and the Gmelin test was negative. The free hemoglobin
observed shortly after the arm had been chilled indicated the immediate destruc-
tion of the patient's red cells within the blood vessels. The amount of
free hemoglobin reached its peak within about four of .five minutes after the
attack had been produced. The plasma was bright red and no tests could be
obtained for the presence of bilirubin. Twelve minutes after the patient's arms
had been removed from the ice-water, a specimen of blood showed less than
one-fourth the amount of free hemoglobin observed in the previous sample.
The plasma was only slightly red, and there was a positive reaction to the
Gmelin test for bile pigment. Furthermore, the color dilution of the plasma
was increased over the previous specimen by over SO per cent. Blood taken
forty-five minutes after the attack showed a still further reduction in free

676

ARCHIVES OF IXTERXAL MEDICI XE

hemoglobin content, and the plasma was deep golden in color. The Gmelin
test was strongly positive, and color dilution values were more than four times
the original normal figure. From this point the pigments in the blood gradually
returned to normal, so that in one hour and twenty minutes after the produc-
tion of the attack of hemoglobinemia there was no free hemoglobin in the
plasma. The bilirubin content had returned so near to normal that the Gmelin
test was negative.

A study of the pigments in the duodenal contents showed no increase over
the peak of the normal curve, obtained prior to the attack, until one hour after
the production of the attack. At this time there was a marked rise in the
bilirubin values. It will be noticed that this rise in bilirubin content of the bile

50

25
0

PLASMA

\

PI6MENTS

Br.j
Hgb./

\

\

DUODENAL
PIGMENTS

1

X

/
/

/

^-~

/

V

hrs/;^ 1 (

)* y.

* y\

'- h

> 3^

225

1 50

75

Chart 1. — The above curves of plasma and duodenal pigments are super-
imposed and have identical time relations. Points along the ordinates represent
values of bilirubin (Br.), free hemoglobin (Hgb.). and combined urobilin and
urobilinogen (Ub., Ubg.). Points along the abscis.sae represent specimens of
plasma or duodenal contents obtained at the time intervals indicated. An
attack of hemoglobinemia was produced at the point marked "Hemolysis."
Magnesium sulphate solution (33 per cent.) was given at various times as
indicated by arrows.

started at the time when the free hemoglobin had entirely disappeared from
the blood, and when the bilirubin content of the plasma had already be,gun to
return to the normal level. The bilirubin content of the bile continued to
increase until it reached its highest concentration at a point two hours following
the attack. From this point it gradually diminshcd in airiount, but three and
one-half hours after hemolysis was produced the bilirubin concentration in the
bile was still nearly twice the greatest amount obtained before hemoglobin
was liberated into the circulation. Urobilin and tirobilinogen values also

JO\'ES-JONES—PAROVXSMAL HEMOGLOBIXURIA 677

showed a marked increase, but the time of appearance of these pigments was
later than that of bilirubin. At the end of two hours, when the bilirubin
concentration had reached its peak, urobilin and urobilinogen values began
to show a definite increase over the normal. From this point, however, they
continued to increase until the greatest concentration of these pigments in the
bile was reached about three hours after hemolysis had been produced. These
pigments then gradually diminished in amount, although three and one-half
hours after the induction of hemolysis their concentration in the bile was still
about double the highest point reached in the control readings.

Repeated examinations of the urine revealed at no time the presence of any
abnormal pigments.

Analysis of the above findings brings out the following points. An attack
was produced in a patient, in which there was an immediate liberation of a
large amount of hemoglobin into the circulation. .'Mmost immediately following
the attack the hemoglobin freed from the patient's red corpuscles reached
its highest concentration in the circulation. At the end of twelve minutes the
larger part of the free hemoglobin had been removed from the circulation and
had been replaced by bilirubin, which appeared in amounts greatly increased
over normal. As the hemoglobin diminished, the bilirubin content increased,
until the latter reached its peak in about forty-five minutes. Both pigments
gradually returned to normal, about one-half hour later, and as they approached
normal the bilirubin content of the bile showed a marked increase. Accompany-
ing this increase of bilirubin in the duodenal contents there was a similar rise
in the lower bile pigments, urobilin and urobilinogen. This latter increase,
however, proceeded at a much slower rate, and did not reach its peak until
the bilirubin had already begun to return toward normal.

Here, then, is an orderly sequence of events, and the following
deductions may reasonably be drawn. Following an uncomplicated
attack of red cell destruction in man the hemoglobin liberated into the
general circulation was rapidly changed to, or replaced by, an increased
amount of bilirubin in the plasma. Inasmuch as there was no restric-
tion on the circulation, it is evident that the liver, or any other organ,
might have participated in this transformation of hemoglobin into bile
pigments. As a matter of fact, however, there was no demonstrable
response in the liver excretion until all the heinoglobin, and the greater
part of the excess bilirubin had disappeared from the plasma. Such a
finding seems to indicate that the change from hemoglobin to bilirubin
might well have taken place merely in the blood vessels and capillaries.
That there was a marked response on the part of the liver to the
sudden excess of hemoglobin in the circulation is indicated by the
elimination in the bile of a tremendous amount of bilirubin. This
increase amounted to as much as three times the normal concentration
of this pigment in the bile, and is clear evidence of a definite relation
between the liberation of hemoglobin into the circulation and the excre-
tion of bilirubin in the bile. Such an increase is also definite evidence
of a stimulation of the liver to increased activity by the products of
red cell destruction. The accompanying rise in concentration of
urobilin and urobilinogen in the bile following the production of an
attack of hemolysis indicates the close relation of these pigments to
bilirubin, and thus indirectly to hemoglobin.

678 ARCH WES OF IXTERXAL MEDICIXE

After a careful examination of the time relation between the
appearance of an excess of bilirubin in the bile and the subsequent
increase of urobilin and urobilinogen another point of interest is
brought out concerning the origin of these last two pigments.^- It will
be noticed on Chart 1 that the peak of a bilirubin curve in the duodenal
contents came approximately one hour before urobilin and urobiHnogen
were found in their greatest concentration. Furthermore, the peak of
the latter pigments came within three hours of the beginning of the
experiment. This rise in urobilin and urobilinogen values was clearly
the result of a liberation of hemoglobin into the general circulation, and
followed the increase of bilirubin in the bile by a definite but relatively
short interval. The time interval elapsing between the appearance of
excess bilirubin was obviously too short to allow of intestinal action
upon the bilirubin excreted in the bile, and subsequent reabsorption of
pigment via the portal circulation. Bacterial action, especially on
protein molecules, is not marked until the lower portion of the small
intestine and the upper portion of the large intestine is reached. In
this experiment it is obvious that bacterial reduction of the excess
bilirubin in the intestine could in no way account for the increased
amounts of urobilin and urobilinogen in the bile, and is direct evidence
against the early theory of the intestinal formation of urobilin and
urobilinogen. The time element alone appears to exclude such a possi-
bility. The logical assumption is, therefore, that these lower pigments
were formed by the liver directly from the bilirubin which was a result
of increased hemoglobin katabolism. The power of the liver to form
not only bilirubin, but urobilin and urobilinogen, is clearly indicated.
The results suggest that the formation of bilirubin may take place out-
side of the liver, but do not prove this conclusively, inasmuch as the
blood determinations were all taken from the general circulation, from
which the liver and other organs could not be excluded. With these con-
siderations in view a second experiment was tried with two main
objects: (1) to check the results of the first experiment, and (2) to
produce evidence of the extrahepatic formation of bile pigment.

Experiment 2. — Patient, G. L. T.

The results of this experiment are shown in Chart 2. Control specimens of
blood, duodenal contents, and urine were taken and examined as outlined in
Experiment 1. An attack of hcmoglobinemia was produced this time by
immersing only the patient's left arm in ice-water, in order to study the
changes taking place in the blood vessels, w-ithout the intervention of the
liver or other organs, a tourniquet was applied above the elbow of the left
arm before immersion, and was kept on this arm for twenty-five minutes. In
this way the blood in the vessels, capillaries and tissue spaces in the lower

22. Urobilinogen, although present in increased amounts,
found in great concentration.

JOXES-JOXES—PAROyXSMAL HEMOGLOBIXURIA

679

left arm could be subjected to chilling and subsequent warming without
mingling with the general circulation. Samples of blood were taken from the
left arm several times before the removal of the tourniquet, in order to observe
any changes taking place in the pigments while the local circulation was thus
isolated. Specimens of blood were also taken from the right arm at similar
intervals in order to ascertain whether there was any leakage of blood from
the immersed arm past tlie tourniquet into the general circulation. Unfor-
tunately, the tourniquet pressure was not sufficient to prevent some leakage, and
the results, while more suggestive than in the first experiment, were still not
conclusive. The chart showed a similar curve for the pigments in the blood
plasma to that obtained in the previous experiment. It will be noted, how-
ever, that both the free hemoglobin and the bilirubin contents of the plasma
reached a greater concentration than that obtained before. It also required a
much longer period for these pigments to return to their original concentration
in the blood. The higher values may be readily accounted for by the fact
that the attack of hemolysis produced in this experiment was more severe than

300

2 00

100

Chart 2. — Constructed in the same manner as Chart 1. The significance of
the various pigments and the method of their determination is fully outlined
in the text.

the previous one, as evidenced by the finding of traces of free hemoglobin in the
first specimen of urine voided after the attack. Furthermore a large amount
of the free hemoglobin in the vessels of the left arm was not released into the
general circulation until the removal of the tourniquet, some twenty-five minutes
after the attack of hemolysis was produced. Such a delay would naturally
retard the complete katabolism of the various pigments. In this case the bili-
rubin content of the plasma did not return to normal until about twenty-one
hours later.

A.S a means of determining the relative amount of blood destruction pro-
duced in this experiment duplicate red counts were taken from the peripheral

680 ARCHIVES OF IXTERXAL MEDICI XE

and general circulation before the immersion of the arn. in ice-water, and
shortly after the removal of the tourniquet. Before hemolysis was produced
the red cell count was 5,394,000 per c. mm. Similar counts made almost immedi-
ately after the tourniquet had been removed averaged 4,528,0CO per c. mm. In
other words, the number of red corpuscles destroyed was roughly about 850,000
per c. mm. Such a figure does not take into account changes in the peripheral
circulation or in blood volume, but permits at least a rough estimate of the
severity of the attack.

The duodenal contents were collected as before, and determinations of
urobilin and urobilinogen were made. The curve of bilirubin elimination was
not estimated. This time, however, specimens were collected for four hours
after the initiation of hemoglobinemia. The duodenal tube was then removed,
but was reintroduced the following morning in order to determine approxi-
mately the length of time necessary for the duodenal pigments to return to a
normal level. Blood e.xaminations were made at the same time. Estimations of
the duodenal pigments showed that they reached their peak in about three
hours, but that the\' did not return to normal until about some twenty-one hours.

The results of this experiment confirmed the findings and conclu-
sions of the first. In addition they afforded a rough estimate of the
amount of blood destruction necessary to produce a given level of pig-
ment values, both in the blood plasma and in the bile. In an attack of
hemoglobinemia just sufficient to produce a trace of hemoglobin in the
urine it apparently takes about twenty hours for the bile pigments in
the plasma and bile to return to their former normal level. The experi-
ment failed to give conclusive evidence of the extrahepatic formation
of bile pigment.

Experiment 3. — Patient A. L.

The results are shown in the accompanying table. The purpose of the
experiment was to obtain conclusive evidence that the bile pigments, or at
least bilirubin, could be formed in the peripheral blood vessels and other
tissues that occur in an extremity, without the intervention of the liver. Such
a probability was suggested in both the previous experiments, but failure
properly to isolate the peripheral arm circulation from the general circulation
made any evidence obtained inconclusive. In this experiment the rubber arm
band of a sphygmomanometer was applied to the patient's left arm above the
bend of the elbow, and was used as a tourniquet. The systolic blood pressure
of this patient was 118. Throughout the experiment the pressure on the arm
was maintained above this systolic pressure with the result that the blood
below the tourniquet was completely isolated from the rest of the general circu- '
lation. After a specimen of blood had been obtained as a normal control, the
patient's left arm, with the tourniquet applied, was immersed in ice-water
(temperature, 5 C). In order to cause only a moderate amount of hemolysis
the arm was kept in the water for a period of only two and one-half minutes.
It was then withdrawn and covered with a warm blanket to supply the
deficiency of body heat caused by the stasis of the peripheral blood. Specimens
of blood were withdrawn from the arm circulation at intervals of three, twenty
and thirty-three minutes after hemolysis had been produced. A specimen was
then taken from the gneeral circulation before the removal of the tourniquet
in order to prove that there had been no communication between the arm and
the general circulation. The tourniquet was then removed and the blood from
the arm circulation allowed to mingle with that of the general circulation.
A final specimen was taken from the general circulation twenty minutes after
the removal of the tourniquet.

JONES-JOXES—PAROyXSMAL HEMOGLOBIXURIA 681

The first specimen of blood taken after immersion of the arm (Specimen 2)
showed a definite trace of free hemoglobin. The next sample of blood, taken
twenty minutes after the beginning of the experiment, showed a large amount
of free hemoglobin, so that the plasma was reddish yellow in color. The
Gmehn test was negative. The delay in the appearance of the hemoglobin may
be explained by the fact that, owing to the complete isolation of the arm
from the general circulation, the temperature after immersion was well below
that of the rest of the body. Inasmuch as the hemolytic reaction in paroxysmal
hemoglobinuria takes place completely only after the blood has been returned to
body temperature, it becomes evident that the slow return of the arm to body
temperature greatly delayed the process of hemolysis. Specimen 4 was taken
thirteen minutes later. The plasma was dark yellow in color, instead of having
the strong reddish tinge observed in the previous specimen. Simple color dilu-
tion showed the same pigment content as in Specimen 3. The hemoglobin con-
tent, however, was less than half that observed in the previous sample.
Furthermore, there was a definite positive Gmehn test. The tourniquet was
then removed as the arm was very cold and cyanotic. Specimen S, taken from
the general circulation just before the tourniquet was removed. sho>ved that
no hemoglobin had entered the general circulation from the left arm. Specimen
6. taken from the general circulation about an hour after the attack of hemolysis
was produced, and twenty minutes after Specimen 5. showed a trace of free
hemoglobin, and in addition gave a doubtful Gmehn reaction.

The above findings may be summed up as follows: Intravascular
hemolysis was produced in a very restricted portion of the circulation.
Tluis the products of hemolysis were completely isolated from the
influence of the general circulation, or of any of the organs of the body.
Free hemoglobin was hberated and reached its greatest concentration
in the plasma in twenty minutes. Although still completely isolated,
the free hemoglobin rapidly diminished in concentration. Accompany-
ing this reduction in hemoglobin concentration there was a marked
increase in the bilirubin content of the plasma, as indicated by a posi-
tive Gmelin test, and by a change in color of the plasma from a reddish
yellow to a dark golden yellow. Inasmuch as no bile pigment had been
introduced into the blood of the isolated arm circulation at any time it
is evident that the free hemoglobin had undergone a transformation in
the bilirubin in the local blood vessels, capillaries and tissue spaces.
The absence of any trace of free hemoglobin in the general circulation
before the removal of the tourniquet indicates that there was no com-
munication between the process taking place in the arm and the rest of
the body.

It will be noticed in the table that the hemoglobin content of the
plasma in Specimen 6, taken from the general circulation twenty
minutes after removal of the tourniquet from the left arm, is still
about one-half the concentration in Specimen 4. Obviously the libera-
tion of the free hemoglobin contained in the lower left arm into the
blood of the general circulation, other things being equal, should result
in a much greater dilution than that observed. .\ similar discrepancy

682 .-IRCHirES OF IXTERXAL MEDICIXE

may also be noticed in a careful comparison of Charts 1 and 2. In
Experiment 1 the free hemoglobin of the plasma obtained from the
general circulation shortly after hemoglobinemia was induced was
twenty-two dilutions. In Experiment 2 the hemoglobin concentration
in the blood plasma taken from an almost completely isolated lower
arm circulation, following a more severe attack of hemolysis, and
resulting in the appearance of traces of hemoglobin in the urine, was
only thirty dilutions. Such an apparent discrepancy between the
amount of free hemoglobin observed in the general circulation and
that noted in a restricted part of the circulation may be readily
explained by a consideration of the mechanism taking place in parox-
ysmal hemoglobinuria. As has already been explained, complete hemol-
ysis occurs only when the temperature of the chilled blood has returned
to 37 C, In Experiments 2 and 3 the temperature of the arm in which
the local circulation had been restricted by the application of the
tourniquet did not return to normal for a considerable length of time.
In Experiment 3. for example, the arm temperature, thirty-tive minutes

Evidence of thf. Extr.-\hep.\tic Formatiox of Bile Pigments*

Specimen Time Interval Color of Color Gmelin Hemoglobin

of Plasma .Alter Attack Plasma Dilutions Test Dilutions

1 Control .Straw 12 0 0

■i 3 minutes Pink straw 20 0 2

3 20 minutes Reddish yellow 4.5 0 9

4 33 minutes Dark yellow 45 + 5

5 35minutes Straw 12 0 0

Pmk yellow 35 ± 2.5

'Explanation: Intravascular bemoly-i- ii.i- i.(..Ii,..i : i!>. i.-..I- ..f i!-. Im«,, ], fi .m.,

following the taking of Specimen I. ^; ■ •■ ■ • ' . ■ !,■ •■, • •' •;

arm. where the blood was cut off finn t

tourniquet with a pressure constantly ^-1- ' ! i ,.- , ^|<, . , ,, . i,, „

from the general circulation before the It'll' : M.M w:.- '.'ii.<'\''\ I'mi rlu l>i! :i!im ^pn mm :i '.

was taken from the general circulation twiity uiiiintis aftir tlic toniiii.|ii.t ikuI h.-i ii taken
off the left arm.

after the induction of hemolysis, was still well below normal. .As a
result complete hemolysis did not occur in either Experiment 2 or 3 as
long as the blood in the arm was isolated from the rest of the circula-
tion. Only when the blood contained in the arm vessels was allowed to
enter into the general circulation did the hemolytic complex become
completely activated. Thus hemoglobin determinations in specimens
of blood taken from the local arm circulation were relatively low, while
those specimens taken from the general circulation after removing the
tourniquet continued to show a relatively high hemoglobin concentra-
tion in the plasma because of a continuation of the hemolytic process.

Uctinite evidence is thus presented of the extrahepatic formation of
bilirubin from hemoglobin in man. The process in this instance took
place solely in the blood vessels, capillaries, and tissue spaces of the
patient's left lower arm. Inasmuch as the process was entirely con-
fined to these anatomic structures it seems logical to assume tliat the

JOXES-JOXES—PAROVXSM.^L HEMOGLOBINURIA 685

principal if not the sole agents concerned in the transformation of
hemoglobin into bilirubin were the cells of the vascular endothelium.
Such a conclusion has already been made by Hooper and Whipple ^' in
the case of animals. If such an assumption is true, it is logically sug-
gested that the normal activity of the liver in the process of hemoglobin
metabolism is due to the endothelial cells with which it is richly sup-
plied, both in the blood vessels and sinuses, and in the so-called stellate
cells of Kuppfer. While undoubtedly the transformation of bilirubin
into the lower bile pigments is due to the activity of the parenchymal
cells of the liver, it is evident that some, and possibly a large part of
the first steps of hemoglobin katabolism may be carried on without the
intervention of the hepatic parenchyma.

CONCLUSIONS

1. In the absence of complicating factors, varying degrees of blood
destruction in man are accompanied by corresponding variations in the
concentration of bile pigments in the blood plasma and bile.

2. In man, the liberation of excessive amounts of hemoglobin into
the circulation results, first, in the rapid elimination of hemoglobin
from the blood stream and its replacement by bilirubin ; second, in the
more gradual disappearance of excess bilirubin from the plasma ; third,
in the appearance of an increased amount of bilirubin in the bile;
fourth, in subsequent diminution of bilirubin in the bile, and its replace-
ment by increased amounts of the lower bile pigments, notably urobilin-
ogen and urobolin; and fifth, in the gradual elimination from the bile of
excessive amounts of the latter pigments. There is a definite time
interval between each of these 'phenomena. The above process is
probably only an exaggeration of the normal process involved in the
metabolism of hemoglobin.

3. Experimental evidence in man suggests strongly that the bulk
of this pigment elimination is normally carried on by the liver.

4. Under abnormal conditions, in man as well as in animals, free
hemoglobin can be broken down, at least as far as the bile pigment
bilirubin, in the blood vessels, capillaries, and tissue spaces, without the
intervention of the liver or any other organ.

5. These experiments suggest that a large part of the transforma-
tion of hemoglobin into bilirubin could occur normally in the blood
vessels. Possibly the greater portion of this change takes place in the
blood vessels of the liver because of the vascularity of this organ.

6. The low bile pigments urobilin and urobilinogen can be
formed in the liver, without the inter\enti(in of bacterial action in the
intestine.

The writers wish to express their appreciation to Dr. Roger I. Lee and
Dr. George R. Minot for extremely vahial)lc criticisms and suggestions.

STUDIES OF THE CAUSE OF PAIN IN GASTRIC
AND DUODENAL ULCERS II.

PERISTALSIS AS THE DIRECT CAUSE OF PAIN IN GASTRIC
ULCERS WITH ACHVLIA AND IN DUODENAL ULCERS

LEO L. J. HARDT, M.D.

ROCHESTER. MIXX.

For many years clinicians have been aware of the fact that not
infrequently the symptoms of infection of the gallbladder and appendix
and of achylia gastrica of unknown etiology may simulate those of
typical gastric or duodenal ulcer. Since reflex pain may be indistinguish-
able from the true pain in ulcer, to wdiat should this gastric pain, brought
on by extragastric lesions, be attributed? Is there direct damage to
the stomach and duodenum through this reflex mechanisin?

In the consideration of the cause of gastric pain from gastric and
duodenal ulcers or reflexly from an infected gallbladder, appendix, or
achylia gastrica, two factors are of importance. On the one hand, is
the question of acidity or hyperacidity and hypersecretion, and, on the
other hand, variations in tonicity, intragastric tension, and peristalsis
of the stomach and duodenum. Clinicians generally have been satis-
fied with the plausible explanation of hyperacidity and hypersecretion
being the most likely cause of pain. As proof they administer alkalies
which stop the pain, and quite logically they conclude that control of
the pain and subsequent healing of the ulcer are mainly questions of
neutralization. This view has been substantiated by Cannon ^ in his
classical work on the acid control of the pylorus. His theory does not
explain the emptying of the stomach in gastric achylia, the rapid exit
of water and egg albumin, nor the observations of Spencer, Meyer, ■
Rehfuss and Hawk,- that a 1 per cent, solution of sodium bicarbonate
hastens the discharge from the normal human stomach.

The experimental work of Luckhardt, Phillips and Carl.son ^ indi-
cates very conclusively that the pylorus opens for the ejection of
chyme when it is reached Ijv powerful ad\ancing rings of contractions
and when tonicitv of the stomach musculature is grcatlv increased.

1. Camion, W. B. : The .•\cid Control of the Pylorus, .•\in. J. Physiol. 20:
283. 1907.

2. Spencer. W. H. : Meyer. G. P. ; Rehfuss. M. E., and Hawk, P. B. : Gastro-
intestinal Studies. XII. Direct Evidence of Duodenal Regurgitation and Its
Influence nn the Chemistry and Function of the Normal Human Stomach. Am.
J. Physiol. 39:459. 1915.

3. Luckhardt, A. B.: Phillips. H. T.. and Carlson, A. J.: Contrihutions to
the Physiology of the Stonuich. LI. The Control of the Pvlorus. .A.m. J.
Physiol. 1:57, 1919.

HARDT—PAIX IX GASTRIC ULCER ()85

The gastric contents entering the duodenum are usually acid to phenol-
phthalein, but rarely show the presence of free acid to dimethyl-
amidoazobenzaldehyd.

Carlson,* Ginsburg, Tumpowsky and Hamburger' and Hardt,",
working independently, have demonstrated, by means of kymographic
records, various types of contractions and peristalsis of the stomach,
which they concluded are the main factors in the causation of the pain
of ulcer. New light was thrown on the etiology of pain, not only from
gastric or duodenal ulcers, but also from achylia gastrica.

INVESTIGATION

Method. — The relation of gastric motility to pain was studied in
twenty-tive patients by the kymographic method. All patients were
given a standard meal, consisting of two soft boiled eggs, two pieces of
toast, two glasses of milk, and the juice of a grapefruit. From one to
two hours later, two tubes were swallowed, a Rehfuss tube and a
small rubber tube with a fine rubber balloon attached at one end and a
chloroform manometer to the end which projected from the mouth.
The balloon, held as closely as possible to the cardiac end of the
stomach, was blown full of air ; it was compressed according to the
various types of contractions of the stomach, and thus air forced into
the manometer caused the rider to record the variations in tonicity
and the contractions on a slowly moving kymograph. These tracings
were continued for from one to three hours, and at intervals of one-
half hour from 15 to 30 c.c. of stomach contents was aspirated to
determine the presence or absence and percentage of free and total
acids. The twenty-five patients were divided into two groups :

Group 1. — This group comprises twenty patients with duodenal
ulcers, the majority of whom came to the clinic during a quiescent
period in their trouble. The diagnosis was based on the clinical history
and confirmed by roentgen-ray examination. All these patients were
treated medically by the Sippy method at the completion of the kymo-
graphic record. None of the patients whose kymogram showed only
Type I contractions (Fig. 1) experienced pain. All of these patients
during the course of experiment revealed adequate free acids ranging
from 20 to 90 (in terms of one-tenth normal hydrochloric acid). Six

4. Carlson, A. J.: Contributions to the Physiology of the Stomach. XLIV.
The Origin of the Epigastric Pains in Cases of Gastric and Duodenal Ulcer,
.\m. J. Physiol. 45:81. 1918.

5. Ginsburg, H. ; Tumpowsky, I., and Hamburger, W. W. : Contributions to
the Physiology of the Stomach, XXXV. The Newer Interpretation of the
Gastric Pain in Chronic Ulcer, J. A. M. A. 67:990 (Sept. 30) 1916.

6. Hardt, L. L. J. : Pain in Active Pathologic Processes in Stomach or
Duodenum. Gastric and Duodenal Contractions as the Direct Cause, J. A.
M. A. 70:837 (March 23) 1918.

686 ARCHIl'ES OF IXTERXAL MEDICIXE

of the patients in whom contractions of Type III (Fig. 2) or Type IV
(Fig. 3) were recorded, experienced burning or gnawing pain similar
to pain in ulcer, in practically every instance synchronous with the
peristalsis, but as long as Type I contractions were recorded they were
without pain.

The degree of acidity seemed to have little bearing on the pain.
The acidity on the whole was lower at the time of the pain and active
peristalsis than during the absence of pain and slight peristalsis.

Fig. 1 (Case 147.732). — Type I contraction. Tonus
peristalsis, without pain.

Group 2. — This group comprises fi\e patients with achylia gastrica.
Three patients were without pain and without any demonstrable patho-
logic condition, one patient had gastric ulcer and one had pain without
any demonstrable pathologic condition. The three patients without
pain had no definite epigastric distress other than a little bloat-
ing or a burning sensation. Two of them revealed only Type I contrac-
tions during the course of the experiment: but the third had definite

Fig. 2 (Case A35 1.625
ith pain.

beginning liunger peristals

T\pe TI (Fig. 4) and Type III peristalsis. These three patients clearly
show that normal tonus variations and peristalsis can be present in
achylia gastrica. The patient with gastric ulcer (Case A147,732) had
repeated gastric analyses which failed to reveal evidence of free acids.
The tyjiical epigastric pain of a gastric ulcer continued, ])ain coming on
from two to three hnm-s ;ifier eatiuij, with relief bv food, water, alkalis.

HARDT—PAIX IX GASTRIC ULCER 687

gastric lavage, and emesis. On several occasions during a period of dis-
tress, acid- free contents were washed out of the stomach and the patient
obtained prompt relief. Kymographic tracings begun two hours after
a test meal showed tonus variations without pain ( Fig. 4) ; but gradu-
ally as the tonus variations were replaced by more active contractions
of Type II and Type I\' intermittent epigastric pain was complained
of, which in almost every instance was synchronous with the peristalsis.
The fifth patient (Case A254,365) had pain typical of ulcer, differing
somewhat in that it continued for long periods without remission, but

Fig. .■? (Case .-\356.6(T4 i . — Type IV contraction. X'igorons hnnger peristalsis
i'hich is coincident with gnawing or burning pain.

Fig. 4 (Case A234.365). — Type II contraction. Exaggeration of tonus varia-
tion which precedes the vigorous peristalsis : frequently associated with a mod-
erate amount of pain.

no demonstrable pathologic condition. Several analyses of gastric con-
tents failed to reveal any evidence of free acidity. At the time the
patient experienced pain the kymograph recorded Type II (Fig. 4) and
Type III (Fig. 2) contractions, the pain being absent during the period
of slight tonus variations ( ¥\g. 1 ) .

REPORT OF CASES

C.ASF. 1 (.^47,732). — History. — S. .'\. M., aged 49 years, tirst came to the

Clinic, Dec. II, 1915, complaining of epigastric pain three to four hours after

meals. The attacks occurred in spells lasting from a few weeks to a month

and remissions lasted from three to five months. Belching, drinking hot water.

688 ARCHirES OF IXTERXAL MEDICIXE

milk, or cream had usually given relief. For three months beginning August,
1915, he had vomited nearly every night between 1 and 3 a. m.

Operation (Dec. 20, 1915). — This revealed a perforating duodenal ulcer
extending into the pylorus, with adhesions to the head of the pancreas and
with almost complete closure of the pylorus. Gastro-enterostomy and appen-
dectomy were performed. Gastric analysis at this time revealed total acids 42
and free acids 30, with laboratory findings pointing to pyloric obstruction. The
patient was free from symptoms for three months following operation.

Course. — Frotn 1916 until the patient's second admission to the Clinic. June
19, 1921, he had frequent attacks of epigastric pain two to three hours after
meals, which were relieved by emesis, food and alkalies. Roentgen-ray exam-
ination revealed a gastric ulcer.

At operation, June 27, 1921, scar tissue was found on the duodenum, but
no ulcer within. An ulcer 1.5 cm. in diameter was located on the lesser curva-
ture of the stomach, 3.75 cm. above the pylorus. The gastro-enterostomy was
found to be patent and functioning. The ulcer was excised and the diagnosis
confirmed by microscopic examination.

Casf: 2 (.A254.365).— //i.s/o;-y.— M. T., aged 25 years, first came to the Clinic
Dec. 26. 1918. He complained of epigastric distress of a burning character one
hour after meals and at midnight. The pains were usually relieved by eating.
In addition he complained of diarrhea which was closely associated with the
epigastric distress. Gastric analysis failed to reveal any free acidity. Eitdameba
histolytica was found in the stools. He was given emetin treatments for five
days and sent home with advice as to treatment of diarrhea.

Dec. 26. 1919, the patient returned, still complaining of frequent attacks of
burning pain in the epigastrium. The pain would be present for three or four
days and then disappear for a week. The diarrhea had cleared up. He was
given bromids three times daily after meals.

Dec. 27, 1920, the patient again returned with identical gastric complaints.
The bromids had given relief for six months. This time he was given dilute
hydrochloric acid, fifteen minutes after meals.

.April 19, 1921. the patient returned with epigastric pain which had not been
relieved by the previous treatment with hydrochloric acid.

The stomach had been examined with the roentgen ray on each admission
but no evidence of gastric or duodenal ulcer had been obtained. Repeated
gastric analysis had failed to reveal any free acids.

At the last visit kymographic records were taken which showed that the
pains were intermittent and coincident with the active peristalsis. .Alkalies and
tincture of belladonna gave relief.

COMMENT

The two groups of patients substantiate the theory of variations in
tonicity and peristalsis as the main factor in the cause of gastric pain ;
the acidity is considered a secondary and in some cases not even a
necessary finding, as in the two cases of achylia gastrica. That the
motility and tonicity of the stomach is quite independent of the acidity
is indicated by the cases of achylia in which all the normal contractions
were obtained in the absence of acidity. Furthermore, emptying of
the stomach was not interfered with, since in none of these cases were
there the slightest symptoms or signs of retention.

The quiescent period is probably the result of a diminution in the
degree and extent of the inflammatory process, together with a decrease
in the tonicity and contraction of the gastric and duodenal musculalure.
The administration of food, alkalies, water, emesis and gastric lavage

HARDT—PAIX IX GASTRIC ULCER 689

temporarily produces this quiescent state, mainly through the inhibi-
tion of the peristalsis which is replaced by die nonpainful digestive
peristalsis, described by Rogers and Hardt." The acid in all probability
merely exaggerates to some extent the pain resulting from the more
vigorous peristalsis and pyloroduodenal spasms. .

The logical therapy in cases of gastric and duodenal ulcer should,
it seems, tend primarily to inhibit peristalsis. About 85 per cent, of
the patients treated surgically in the Mayo Clinic have been cured oi
the ulcer or satisfactorily improved. It might be assumed from these
results that surgical procedures inhibit the vigorous peristalsis for a
period long enough to promote healing of the ulcer, or at least a sub-
sidence of the more acute inflammatory process. It is hoped thai in
the future this assumption can be demonstrated more conclusively by
the kymographic method.

CONCLUSIONS

1. Gastric ulcer may be present in patients with achylia and ni.iy
produce all the clinical symptoms characteristic of ulcer. The pain
is primarily due to the peristalsis acting on an irritable focus. .All
the medical measures by which gastric acidity is neutralized and sup-
pressed also inhibit the gastric peristalsis and thus relieve the pain.

2. Patients with gastric achylia in the absence of any demon-
strable organic lesion may reveal the normal tonus variations and
peristalsis.

3. Patients with uncomplicated duodenal ulcer do not experience
pain during the period of digestive peristalsis, even in the presence of
an adequate acidity. Active peristalsis of the "hunger type" (Types
III and IV) is essential in the production of pain.

7. Rogers. F. T.. and Hardt, L. L. J. : Contributions to the Physiology of
the Stomach. XXVI. The Relation Between the Digestion Contractions of the
Filled, and the Hunger Contractions of the "Empty" Stomach. Am. J. Physiol.
38:274, 1915.

THE SEAT OF THE EMETIC ACTION OF THE
DIGITALIS BODIES *

ROBERT A. HATCHER and SOMA WEISS

NEW YORK CITY

Emetics are divided commonly into two classes : (a) Those which
irritate certain nerve endings in the gastric mucous membrane; (b)
those which stimulate the vomiting center in the medulla directly.

Hatcher and Eggleston ' showed that emesis follows the intravenous
injection of any one of the several digitalis bodies in dogs, even after
the removal of the gastro-intestinal tract, and while they considered
it possible that vomiting is due to a reflex arising in some peripheral
structure, such as the esophagus, they came to the conclusion that all
the evidence available points to the vomiting center as the seat of the
emetic action of all of the digitalis bodies.

The fact that these bodies differ so widely in their chemical composi-
tion directed our attention to the coincidence that every member of the
group shows a greater or less parallelism between its cardiac and
emetic activities, though the several members of the group differ
widely one from another with reference to the intensity of both of
these actions, suggesting a common seat, and we have sought to
determine wliether the vomiting which digitalis bodies induce is indeed
of cardiac origin.

It is also of especial interest that emesis and cardiac standstill may
be induced within a few seconds after the intravenous injection of
a large dose of digitoxin but that neither of these effects can be
induced nearly so quickly by the largest doses of ouabain.

We have used crystalline ouabain (sometimes called crystalline
strophanthin) in the larger number of our experiments because it is
a typical digitalis body of uniform purity, is readily soluble in water,
and it lends itself to studies such as we planned. We have also used
amorphous strophanthin, digitoxin, digitalein and tincture of digitalis
in those experiments which we consider crucial.

The digitalis bodies leave the circulation rapidly after their intra-
venous injection, = and only traces of the poisons can be found in the

* An abstract of this paper wa.s presented before tbe Society for E.xpcrimcntal
Biology and Medicine Oct. 19, 1921.

* From the Laboratory of Pharniacology of Cornell University Medical
College.

* This work was carried out nnder the auspices of the Therapeutic Researcli
Committee of the Section on Pharmacy and Chemistry of the .\mcrican Medical
.\ssociation.

1. Hatcher and Eggleston: The Emetic Action of the Digitalis Bodies.
J. Pharmacol. & Exper. Therap. 4:113 (Nov.) 1912.

2. Hatcher and Eggleston: Studies in the Elimination of Certain of the
Digitalis Bodies from the Animal Organism, J. Pharmacol. & Exper. Therap.
13:4.11 1919.

HATCHER-WEISS— DIGITALIS BODIES 691

blood after a few minutes, a fact of which we made use in the pre-
Hminary experiments of this investigation. In these experiments we
lied or compressed the carotid and vertebral arteries, in order to
diminish the circulation in the medulla, after which an amount of a
digitalis body equal to a little more than the average emetic dose was
injected intramuscularly or into the femoral vein. Emesis resulted in
nearly every case though it is certain that less of the poison reached
the vomiting center than could have reached it had there been no inter-
ference with the circulation.

\'omiting occurred in one experiment within two minutes after
beginning the intravenous injection of ouabain and at the moment
when an amount equal to about twice the average emetic dose had
been injected, though emesis was delayed in the greater number of
the experiments. This delay was almost certainly due in part to
delayed absorption after intramuscular injection, and this in turn to
the fact that the animals were somewhat depressed.

The cerebral circulation of the dog is said to become nearly normal
soon after the carotid and vertebral arteries are tied, but this is not
true of the cat. The animals usually slept or remained drowsy after
the operation of tying the vessels, and further evidence that the circu-
lation in the brain of the cat is greatly diminished by tying these
arteries is afforded by the fact that the internal and external jugular
veins were severed in one of these experiments without the loss of a
drop of blood, and the blood escaped slowl\- when the carotid artery
was cut across.

The results of these experiments are not conclusive because we
cannot be absolutely certain that all of the poison had left the blood
stream before emesis occurred and the delay in the onset of vomiting
possibly may have permitted a fairly large amount to reach the medulla.
The onset of emesis within two minutes after beginning the injection
of ouabain in the first experiment, and at a time when the total amount
injected was equal to only about twice the average emetic dose, affords
strong evidence that vomiting was not due to a direct action on the
medulla, because the interference with the circulation prevented the
poison from reaching the medulla in an amount comparable to that
which would reach it in the normal animal after an emetic dose.

An interval of twenty-four hours had been allowed to elapse after
the operation and before the ouabain was injected in the first experi-
ment, however, and it is possible that the circulation was partially
reestablished during that interval, though the condition of the animal
did not indicate that the circulation was normal. In the remaining
experiments of this series the drugs were injected after the lapse of a

692 ARCHIVES OF JXTERXAL MEDICIXE

period which was considered necessary for complete recovery from the
effects of the anesthetic. The protocols in brief of two experiments
will be given.

PROTOCOLS OF EXPERIMENTS WITH OU.\B.\IN

Experiment 1. — Female cat, weight 3.8 kg.; anesthetized with ether; the
carotid arteries tied at a point about 3 cm. above the level of the sternum ; the
vertebral arteries tied near their origin ; wound closed with sutures.

4:40 p. m. : Animal released; sleeps at once.

4:50 p. m. : Lifts head; wakens; soon sleeps; rapid respiration.

S :40 p. m. : Asleep but moves frequently.

Following day :

9 :00 a. m. : Sits up ; disturbed equilibrium.

3:25 p. m. : Femoral vein connected with buret for injection.

3:30 p. m. : Start injection of ouabain 1:10.000.

3:32 p. m. : Emesis; 0.13 mg. ouabain per kilogram of weight injected.
Animal destroyed.

Experiment 2. — Male cat, weight 3.58 kg. Operation as in preceding experi-
ment except that chloroform was used to induce anesthesia.

3:10 p. m. : Animal released.

3 :35 p. m. : Sleeps quietly.

3:45 p. m. : 0.075 mg. ouabain per kg., 1: 10,000, injected intramuscularly.

4:08 p. m. : Nausea followed almost at once by vomiting.

Following day : .'\nimal in partial stupor ; killed with chloroform ; respira-
tion markedly diminished ; difficult to kill animal even with chloroform applied
to nose on cloth.

Subsequent experiments were designed to compare the effect of
permitting the drug to act on the vomiting center but not on the rest of
the body, with that seen when the poison was allowed to enter the
general circulation but not to reach the vomiting center.

In five experiments the carotid and vertebral arteries were tied or
compressed, and the brain, with the vomiting center, was perfused for
periods varying from five to fifteen minutes through the carotid
arteries and jugular veins with defibrinated blood to which ouabain had
been added in amounts varying from a small fraction of the average
emetic dose to one that caused convulsions. In no case did vomiting
occur nor were any symptoms of nausea obser\ed after the animal was
released.'

" Five experiments were performed in the manner just described,
except that the ouabain was injected into the femoral vein while the
brain was being perfused with unpoisoned defibrinated and diluted
blood, the perfusion being continued for periods of ten minutes, dur-
ing which time all but traces of the poison left the circulation. The
brain was perfused at a pressure greater than that of the general cir-

3. We employed a modified Langendorff apparatus; the temperature of the
perfused fluid was kept at 37 C. ; the pressure was regulated by means of
compressed o.xygen or air and a mercury valve.

HATCHER-JVEISS— DIGITALIS BODIES 693

(.Illation, and we are fairly certain that no more than traces of the
poison (if any) reached the medulla during the period of perfusion,
though we could not measure the pressure in the circle of Willis at
that time. Three of these five animals vomited ; one showed unmis-
takable signs of nausea, and one showed some signs of nausea but
these two did not vomit, because they were much depressed. The
vertebral arteries were intact in two of these experiments but the
results did not differ materially from those in two experiments in
which they were tied.

Every one who has perfused the brain of the li\ing animal appre-
ciates the difficulties which are involved in the attempt to secure an
approximately normal condition of the animal during the procedure,
and we shall not enter into a discussion of the details of these experi-
ments. We should be unwilling to base our conclusions concerning
the seat of the emetic action of the digitalis bodies on the results of the
perfusion experiments alone, but we can say that these results are in
harmony with those obtained in other experiments which afford con-
clusive evidence that the digitalis bodies do not induce emesis through
a direct action on the medulla. Those animals which were used in this
series of experiments and which did not vomit after the injection of
ouabain intramuscularly or into the femoral vein were so greatly
depressed that emesis could not be expected to occur.

It is well known that Thumas * described an area measuring about
5 mm. in length and about 2 mm. in width, situated in the floor of
the fourth ventricle and extending to a point about 2 mm. posterior to
the calamus scriptorius, which he called the vomiting center. Thumas
found that destruction of the tissue lying within this area inhibits
vomiting, and that the application of apomorphin hydrochlorid to this
area causes emesis in dogs. We have found that the application at
this point of as little as 0.000.1 mg. apomorphin hydrochlorid for
every kilogram of body weight causes emesis in dogs, and that larger
doses cause vomiting within a few seconds.

The experiment is performed in the following way: The animal is
anesthetized with chloroform and secured with the belly resting on
the operating board; the base of the skull is exposed by incision;' a
button of bone is removed by means of a trephine having a diameter
of six millimeters; the opening is enlarged slightly by chipping the
hone ; the dura mater is incised at the margin of the cerebellum ;
the wound is closed with a pledget of cotton soaked in phenol solution
in oil and the animal is released. When the animal has recovered from

4. Thumas, L. J. : Ueber das Brechcentrum iiiul iiber die W'irkung einiger
pharmakologische Mittel auf dasselbe, Virchows. .\rch. f. path. .\nat. 123:44,
1891.

694 ARCHIVES OF IXTERXAL MEDICI XE

the effects of the anesthetic, it is replaced on the board (or, in some
cases, it stands on the table) ; the tissues are pulled aside and any
blood or spinal fluid is removed with a pledget of cotton ; the solution
to be tested is dropped onto the area from a syringe graduated to
0.01 CO., or the solution is dropped onto a very small pointed camel's
hair brush with which it is painted onto the surface. It is important
that general anesthesia be avoided at this time, and the use of the
phenol in oil. applied to the edges of the wound, prevents pain.

Cats and dogs usually bear the operation well and appear normal
as soon as they recover from the effects of the anesthetic. Ether was
used in one of our experiments in which the drug was applied on the
day following the operation, but chloroform was used in the other
experiments because it is much less prone to cause vomiting. None
of our animals vomited from the effects of the operation or from those
of the chloroform. Troublesome hemorrhage from the diploic vessels
commonly follows the use of a trephine of greater diameter than that
mentioned. Some of our experiments were performed with practically
no loss of blood. Physiologic solution of sodium chlorid was dropped
onto this area in control experiments and was found to be without
perceptible effect.

We have sought to determine whether the digitalis bodies induce
vomiting after direct application to the vomiting center, and ouabain,
amorphous strophanthin, digitoxin and digitalein were applied to this
area, but in none of the experiments of this series were we able to
induce nausea or vomiting in this way, though widely varying amounts
of the poisons were used. When very large doses of these drugs are
applied to the center, they cause death without inducing emesis, and
the application of moderately large, but not fatal, doses also appears to
cause depression of the vomiting center, for emesis cannot then be
induced by the intramuscular injection of ouabain.

The direct application of small amounts of the digitalis bodies to
this area does not have any perceptible eft'ect on the vomiting reflex,
and the intramuscular injection of ouabain then causes emesis precisely
as it does in the normal animal. While we speak of the amounts thus
applied as small, since they are far less than the amounts required by
intravenous injection to cause vomiting in the normal animal, they
are actually much larger than those which can come into contact with
the tissues of this area after the intravenous injection of an emetic
dose and before vomiting takes place. The tissue embraced within the
area described by Thumas constitutes approximately 1/75,000 of the
total weight of the dog. and since vomiting frequently follows the
intravenous injection of ouabain or digitoxin within two or three

HATCHER-WEISS— DIGITALIS BODIES o95

minutes, it is evident that only a minute fraction of tlie total amount
injected can come into contact with the vomiting center before emesis
takes place.

The protocols in brief of four experiments are given. The first of
these shows that armorphous strophanthin is absorbed into the general
circulation, for the effects were typical except for the absence of
nausea and \omiting. and that it causes depression of the vomiting
mechanism. The second experiment shows that digitoxin also induces
depression of the vomiting mechanism, since the intramuscular injec-
tion of a fatal dose of ouabain then failed to cause emesis. The
third experiment shows that the application of a small amount of
ouabain is without influence on the vomiting reflex, and the fourth
shows that the application to this area of a moderate amount of
ouabain causes depression or paralysis of the vomiting mechanism,
since the intramuscular injection of very large doses of ouabain then
failed to induce emesis. The result in this experiment is of interest
in that the animal lived thirty-six minutes after the first intramu.scular
injection of an amount of ouabain equal to nearly three times the
average fatal dose, and that it lived twenty minutes after the second
injection, made sixteen minutes after the first injection, having received
a total of nearly eight times the average fatal dose. It would apjiear
that the depression induced delays the absorption of tiie drug from the
intramuscular tissues.

The results of these experiments in which the poison was applied
directly to the vomiting center point almost conclusively to the fact
that ouabain does not induce emesis through any direct action on the
medulla, and the absorption of a fatal dose after its application to the
floor of the fourth ventricle without the production of nausea is of
especial interest.

I'KOTOCOLS OF EXPERIMENTS

Experiment shozcing absorption of amorplwus strophanthin after its apfl'-
cation to the floor of the fourth ventricle. — Male cat, weight, 2.3 kg. Chloroform
administered for anesthesia.

11 :n8 a. m.: Operation completed, animal released.

1 :3n p. m. : .Animal ate meat.

2:30 p. m.: 0.5 mg. amorphous strophanthin per kg. in 20 parts of phy-
siologic solution of sodium chlorid applied to the floor of the
fourth ventricle; immediate depression ; animal unable to stand.

2 ;49 p. ni. : Respiration irregular.
3 :05 p. m. : \o nausea ; death.

Experiment shotcing depression of the vomiting mechanism following the
application of a moderate amount of digitoxin to the floor of the fourth ven-
tricle.—Ma.\e cat, weight 4.4 kg.; chloroform administered for anesthesia.
11:45 a. m. : Operation completed.

2:25 p. m.: Water administered through a stomach tube.
2:31 p. m. : 0.002 mg. digitoxin per kg. in l.OCO parts mucilage applied to
floor of fourth ventricle; animal released; condition excellent.

696 ARCHIVES OF IXTERXAL MEDICIXE

2;48 p. m. : Xo perceptible effect; 0.45 nig. ouabain per kg. of weight

injected intramuscularly.
3 :03 p. m. : No symptom of nausea ; convulsions and death.
Experiment showing that the application of a small amount of ouabain to
the floor of the fourth ventricle is without perceptible effect on the romiting
mechanism. — Female cat, weight, 2.24 kg. ; chloroform administered for anes-
thesia.

2:10 p. m. : Operation completed.

2:30 p. m.: 0.005 mg. ouabain per kg. in 10,000 parts physiologic solution
of sodium chlorid, applied to the floor of the fourth ventricle.
3:00 p. m. : No symptoms of nausea.

0.2 mg. ouabain per kg. in 10,000 parts of physiologic solution
of sodium chlorid injected intramuscularly.
3:15 p. m. : Emesis, repeated.
3:20 p. m. :Convulsion and death.

Experiment showing the depressant action of a moderate amount of ouabain
on the vomiting mechanism following its application to the floor of the fourth
ventricle. — Female cat. weight, 2.9 kg.; chloroform administered for anesthesia;
operation as preceding.

Animal appears normal.

0.01 mg. ouabain per kg. in 500 parts physiologic solution of

sodium applied to the floor of the fourth ventricle: respiration

rapid.

Respiration about normal.

No symptom of nausea.

0.33 mg. ouabain per kg. intramuscularly.

0.66 mg. ouabain per kg. intramuscularly.

Walks normally ; diarrhea.

No symptom of nausea ; convulsion and death.

It is well known, of course, that the heart is supplied not only with
nerve libers from the vagus, but also with fibers from the sympathetic
which pass through the stellate ganglia to the sympathetic chain of
ganglia, and we found that cutting the cord above the level of the
second thoracic vertebra (which prevents impulses from passing from
the heart to the medulla by way of the sympathetic) usually prevents
nausea and vomiting after the administration of digitalis bodies.

Section of the cord below the level of the fifth thoracic vertebra
has no perceptible efifect on the emetic action of the digitalis bodies.
This operation does not interfere with afferent impulses from the heart,
hence these results have a certain value only in connection with those
in which the cord was cut at a higher level.

Since the emetic action of the digitalis bodies is sometimes abolished
by section of the cord above the level at which the sympathetic fibers
from the heart enter, and since atropin does not affect this action, we did
not anticipate that vagotomy would influence it. nevertheless we under-
took to determine the efifect of this operation.

A cannula was placed in the trachea and the vagi were cut at the
level of the sixth cervical vertebra, after which the injection of ouabain
invariably caused vomiting in four experiments in the cat.

3:15 p.

^■.^7 .:

m.
m.

4:30 p.
4:49 p.

m

5 :0S p.
5:18 ?.
S :25 p.

m.
m.
m.

HATCHER-WEISS— DIGITALIS BODIES 697

Removal of the stellate ganglia '-" alone interfered with the emetic
action of ouabain in some of these experiments, and removal of the
ganglia together with cutting of the cardiac branches of the vagus
prevented the appearance of symptoms of nausea almost invariably,
after the administration of digitalis bodies, but mercuric chlorid still
caused emesis promptly. In one experiment of this type, however,
nausea followed the injection of digitalis, but a necropsy on the cat
showed that the sympathetic cardiac nerve gave off three small branches
to the sympathetic chain at a point between the heart and the stellate
ganglion, consequently the removal of the ganglia did not prevent the
passage of impulses from the heart through the sympathetic to the
medulla in this experiment. The sympathetic nerve shows many
irregularities in different individuals, and care is necessary in the
interpretation of the results of experiments in which it is involved.

We believe that the results of these experiments justify the conclu-
sion that when the nerve supply to the heart is intact the injection of a
digitalis body causes emesis if the animal is in good general condition,
but that the digitalis bodies are incapable of inducing nausea or vomit-
ing when all of the nervous connections between the heart and the
medulla are cut, though mercuric chlorid still causes vomiting exactly
as it does in the normal animal.

This would indicate that the emetic impulses to which ouabain
(or digitalis) gives rise do not traverse the same afferent path, or
paths, which the afferent emetic impulses resulting from the action of
mercuric chlorid traverse, or, to express it more accurately, mercuric
chlorid appears to give rise to emetic impulses which reach the medulla
through paths other than, or in addition to, those traveled by the emetic
impulses which digitalis bodies induce.

Poisons are widely distributed in the vegetable kingdom, and it is
evident that animals (as well as man) often take them with their food.
We are so accustomed to seeing vomiting and diarrhea result from the
irritant action of poisons (including under that term all harmful sub-
stances, such as indigestible food) that we are prone to lose sight of
the fact that the stomach and intestine are not the only organs of the
body which require protection from injury due to ingested poisons, and
the heart, liver, lungs, kidneys and nervous system are attacked by
certain poisons which have no injurious action on the stomach. The
latter probably has developed a greater range of tolerance than any of
the other organs, and it would be remarkable if Nature had provided
such a complex reflex as that necessary for vomiting for the protection

5. The operation which we employed for the removal of the stellate ganglia
is that described by E. Cyon " (6. E. Cyon : Methodik der physiologischen
Experimente and Vivisectionen, Giessen, 1876, p. 174) which does not involve
the opening of the chest.

698 ARCHIVES OF IXTEKXAL MEDICIXE

of the stomach while leaving other, and more vital, organs having a
similar innervation unprotected, and vomiting effectually protects the
heart against the further absorption of poisons no less than it protects
the stomach.

Impulses appear to pass upward from the heart to the medulla
chiefly by way of the sympathetic, and to a less, though probably vari-
able, extent, by way of the vagus. When the ■sympathetic alone is cut
the administration of ouabain usually fails to induce nausea or vomit-
ing. This may be due to the fact that the impulses passing upward
by way of the vagus are usually insufficient to set up the vomiting
reflex, or it may be that in those cases where vomiting is not elicited
by the digitalis bodies after the sympathetic has been cut the vagus
carries no fibers concerned with this vomiting reflex. It is significant,
at any rate, that vagotomy alone does not prevent eniesis after the
injection of the digitalis bodies and that Eggleston ' found that atropin
does not interfere with emesis induced by digitalis, though it does
inhibit that caused by pilocarpin.

In order to show that the cutting of the nerve paths from the
heart, and not the disturbance due to the operative procedure, inter-
fered with emesis, we conducted several experiments in which the celiac
plexus was removed and the vagi were cut at the level of the diaphragm
before the digitalis bodies were administered. This operation is more
severe than that involved in the removal of the stellate ganglia and
cutting the vagi in the neck, and some of the animals were so depressed
that one could say with confidence that vomiting could not be induced
by any digitalis body. Digitoxin was injected into two of these animals
despite their being greatly depressed, because it seemed possible that the
drug might induce emesis.

We repeated the experiment on seven cats which bore the operation
with less depression, and all of these vomited or showed unmistakable
signs of nausea, and we can say that the removal of the celiac plexus
and cutting the vagi at" the level of the diaphragm do not interfere with
the emetic action of the digitalis bodies except in those cases in which
the operation causes severe depression. With improved operative
technic there was less depression and three of the last four animals of
this series vomited while the fourth showed unmistakable signs of
nausea.

Inasmuch as the animals in which the vagi were cut and those used
in the last series of experiments behaved like nurnial animals (except
for the depression) toward the digitalis bodies, tlierc is no ob\ious need

7. Eggleston : The Antagonism between Atropin and Certain Central Emetics
J. Pharmacol. & E.xper. Therap. 9:11 (Oct.) 1916.

HATCHER-WEISS— DIGITALIS BODIES 699

of giving the protocols of the experiments, but condensed protocols of
experiments of the other types are given.

The tabulated results of all experiments show that thirty-five of
the animals which received digitalis bodies or mercuric chlorid vomited
or gave unmistakable signs of nausea, and that the result in one of
these was doubtful. Thirty of these actually vomited; four showed
unmistakable signs of nausea but did not vomit after the severe opera-
tion for removal of the celiac plexus and vagotomy. One of these
four, and the other one which failed to vomit, had only average emetic
doses of ouabain, and such doses sometimes fail to induce emesis in
normal animals. The typical signs of nausea — frequent chewing and
swallowing of saliva^are as unmistakable as vomiting itself. When
the animal licked its lips only infrequently, even though repeatedly, it
was counted as doubtful. In none of the forty-four experiments in
which the results are recorded as negative was there any symptom of
nausea.

Protocols of experiments shozfintj the effect of section of the cord on the
emetic action of ouabain in the cat. 1. Female cat, weight. 1.7 kg.: chloroform
administered for anesthesia.

2 :20 p. ni. : Spinal cord severed between tlie first and second thoracic

vertebra; no interruption of respiration; fore legs normal, hind

legs paralyzed.
2:30 p. m. : Condition fair.
4:15 p m. : 0.15 mg. ouabain per kg. in 10.000 parts of physiologic solution

of sodium chlorid injected intramuscular!},.
4:36 p. m. : 0.1 mg. ouabain per kg. in 10,000 parts of physiologic solution

of sodium chlorid injected intramuscularly.
5:20 p. m.: 0.05 mg. ouabain per kg. in 10.000 parts of physiologic solution

of sodium chlorid injected intramuscularly.
5 :25 p. m. : No evidence of nausea : convulsions and death.
2. Female cat. weight, 2.30 kg.; chloroform administered for anesthesia.
3:30 p. m.: Spinal cord severed between the sixth and seventh thoracic

vertebra ; slight hemorrhage.
4:10 p. m.: 0.2 mg. ouabain per kg. in 5,000 parts physiologic solution of

sodium chlorid, injected intramuscularly.
4 :20 p. m. : Vomiting.
4:25 p. m. : Convulsions and typical death.

Protocols of experiments sho-wing effect of extirpation of the stellate
ganglia and vagotomy. 1. Male cat, weight 1.12 kg.; chloroform administered
for anesthesia.

12:10 p. m. : Operation for removal of both stellate ganglia; cannula into the

trachea ; vagi cut.
12:25 p. m. : 1.0 mg. digito.xin per kg. in 2,000 parts physiologic solution

of sodium chlorid injected intramuscularly ; rapidly developing

depression.
1 :34 p. m. : Xo symptoms of nausea have developed : convulsion.
1 -.iS p. m. : Death ; necropsy shows complete extirpation of both ganglia.
2. Male cat, weight 1.64 kg.: chloroform administered for anesthesia.
3:10 p. m.: Completed operation as in preceding experiment.
3:18 p. m.: Condition excellent.

700 ARCHIVES OF IXTERXAL MEDICIXE

TABLE 1. — Showing the Effkct (if the Applicatuin of Digitalis Bodies to

THE Floor of the Fourth Vextricle_ (the Vomiting Center of

Thumas) IX the Cat

Substance

Ann. per Kg. in

Used

Fractions of a Mg.

ouabain

0.01

ouabain

0.01

ouabain

0.01

ouabain .

C0.05

0.1

ouabain

[0.3

ouabain

0.005

ouabain

0 005

ouabain

O.OOS

0.015

ouabain

0.001

ouabain

■^0.0065

ouabain

0.45

L strophantliin

0.000,67

1. strophantliin

0.01

1. strophanthin

0.5

(ligitoxin

0.01

digitoxm

0.002

digitalein

0.015

great depiession

no perceptible effect

depression of vomiting center *

no perceptible effect

no perceptible effect

death

no percepiible effect * *

no perceptible effect**

no perceptible effect * *

depression of vomiting center '

no perceptible effect t

no perceptible effect

death

no perceptible effect * *

no perceptible effect * *

death

depression of vomiting center '

depression of vomiting center '

no perceptible effect**

* The intramuscular injection of a digitalis body later failed to induce eraesis,
showing that the vomiting center was depressed.

* * The intramuscular injection of a digitalis boay later induced emesis,
showing that the vomiting center was not paralyzed.

t A dog was used in this experiment.

T.-XBLE 2. — Showing the Effect or Extirp.-vtion of the Stellate Ganglia

Alone, and with Vagotomy, ox the Emetic Action of the Digitalis

Bodies ix the Cat

Substance

.\m. per

Kg.

Mode of

Nausea or

Used

in Mg. -

Administration

Vomiting

i:XTIRl>.\

TION OF THE S

TELLATI

; GANGLIA ALONE t

ouabain

0..^

intramuscularlv

_

0.5

intramuscularly

digitoxin

1.5

intramuscularly

+

tinct. digitalis

\ 1000.0

intravenouslv

—

mercuric chlorid

1 50.0

bv stomach

+

tinct. digitalis

lOtKlO

intravenously

-^

extirpation

OF THE STKLL

.\TE GA

XGLIA WITH VAGOTOMY

digitoxin

1.0

intramuscularly

_

digitoxin

1.2

intramuscularly

—

strophanthin

0.4

intramuscularlv

—

strophanthin

0,5

intramuscularly

tinct. digitalis

j 1000.0

intravenously

—

( 50.0

bv stomach

+

mercuric chlorid

50.0

by stomach

+

* The + sign indicates that nausea or vomiting occurred ; the — sign
indicates that they were absent.

t There was incomplete extirpation of the stellate ganglia in three experi-
ments, and in these nausea and vomiting occurred. They are not tabulated here
but they are included in the table giving the summary of results of all of the
experiments.

HATCHER-JVEISS— DIGITALIS BODIES 701

3:25 p. m.: 50 mg. mercuric chloritl per kg. in 1.000 parts of water admin-
istered through stomach tube.

^ :28 p. m. : Retching.

3:46 p. m. : Typical nausea and vomiting; animal destroyed; necropsy
showed complete extirpation of both stellate ganglia.

Depression of the vomiting center could be determined only by the sub-
sequent intramuscular or intravenous injection of an emetic dose of ouabain
or other digitalis body. This test was not made in every experiment and it is
probable that the vomiting center was depressed in several of those cases
where no perceptible effect is recorded. General depression always tends to
inhibit emesis.

TABLE 3. — SuMM.ARY of the Rksl'lts ok ExPERiMKiNTS Designed to Determine
THE Se.\t of the Emetic .\ction of the Digitalis Bodies

Animal Xausea or
Vomiting

+ —
I. Carotid and l\-rlcbral Arteries Tied:

1. Ouabain injected cat 5 0

2. Digitoxin injected cat 2 2*

3. Controls ; no poison used cat 0 5
II. Perfusion of flic Brain and I'oiiiiting Center:

1. Ouabain added to perfused fluid cat 0 5

2 Ouabain injected into femoral vein cat 5 1?

3. Controls ; no poison used cat 0 3

III. Digitalis Bodies A f Hied to Vomiting Center Direetiv:

1. Ouabain cat 0 8

2. Ouabain dog 0 2

3. Digitoxin cat 0 2

4. Digitalein cat 0 1

5. Amorphous strophanthin cat 0 3

IV. Seetion of the Cord :

(a) above the level of the second thoracic vertebra:

1. Ouabain injected intramuscularly cat 0 2

2. Pilocarpin injected intramuscularly cat 1? 1

3. ^lercuric chlorid by stomach (control) cat 2 0

(b) section below the level of the fifth thoracic vert.:

1. Ouabain injected intramuscularly cat 2 0

V. I'ayi Cut About the Level of the Sixth Cervical Vertebra:

1. Ouabain injected cat

2. Mercuric chlorid by stomach (control) cat
VI. Extirpation of the Stellate Ganglia :

1. Ouabain injected cat

2. Digitoxin injected cat

3. Tincture digitalis intravenously cat

4. Mercuric chlorid by stomach (control) cat
VII. Extirpation of the Stellate Ganglia and Vagotomy:

1. Digitoxin intramuscularly cat

2. Amorphous strophanthin intramuscularly cat

3. Tincture digitalis intravenously cat

4. Mercuric chlorid by stomach (control) cat
VIII. Extirpation of the Celiac Ganglion with Vagotomy

1. Ouabain intramuscularly cat

2. Digitoxin intramuscularly cat

* These animals were much depressed.

t The extirpation was incomplete in two experiments.

t Tliere was an abnormality of the sympathetic nerve.

4

0

1

0

0

9

1

0

■n

1

1

u

0

->

0

0

It

1

1

0

9

0

5

2^

702 ARCHirES OF IXTERXAL MEDICI XE

While the present discussion is concerned primarily with the prob-
lem of the seat of the emetic action of the digitalis bodies, we wish to
offer certain suggestions relating to the physiology of vomiting, and to
state that we are now trying to secure evidence to determine whether
our view is correct, for this problem is intimately concerned with that
of the emetic action of the digitalis bodies.

^^"e believe that the vomiting center described by Thumas bears the
same relation to the act of vomiting (and possibly to other functions)
which the spinal cord bears to the many normal reflexes in which it is
known to be concerned. We believe that afferent impulses more or less
constantly, or, at least, frequently, pass from various peripheral organs,
including the stomach and the heart, through the sympathetic to the
center in the medulla of the normal animal, but that these normal
impulses are too feeble to set up the powerful reflex concerned in
vomiting which is accompanied by violent, and even convulsive, con-
tractions of the diaphragm and abdominal muscles.

It is well known, of course, that sensory impulses pass almost
constantly from various parts of the body to the cord and give rise to
slight reflex movements or none. If one scratches the skin gently there
is no perceptible reflex, but a violent scratch induces a prompt reflex
mo\enient. It is also well known, of course, that strychnin acts on the
cord in such a way that the passage of impulses is facilitated so that
gentle scratching of the skin then induces typical convulsions.

We believe that the direct action of apomorphin on the vomiting
center in the medulla is wholly analogous to that of strychnin on the
cord, and that when the reflex excitability of the center is increased by
apomorphin emesis results from normal afferent impulses.

We haxe recently obtained evidence which we believe lends some
support to the view that the vomiting center of Thumas is merely a
mechanism for the coordination of the reflexes concerned with nausea
and vomiting (and possibly with other functions) and while we do not
wish to enter into a discussion of the details of these experiments at
this time we wish to say that we have induced nausea in cats and dogs,
with actual vomiting in one, by applying strychnin to the vomiting
center. It is significant also that niorphin causes apomorphin-like
emesis in dogs and strychnin-like convulsions in frogs.

\'oniiting is known to be induced by the action of toxic substances
or by injuries aft'ecting many organs, including the stomach, intestines,
liver, uterus, kidneys, testicles, and brain, and the results of our work
point to the heart also as the seat of reflex vomiting.

Since the various reflex paths are always ready for instant service.
c\en in individuals who have never vomited, it seems reasonable to
sujipose that the tone of these paths is maintained in health by means

HATCHER-JVEISS— DIGITALIS BODIES 703

of impulses which traverse them constantly or frequently ; for example
those from the stomach when it contains food, those from the heart
when there is any minor disturbance or change in rate due to sudden
exertion, and it is well known that violent or prolonged exertion fre-
quently induces nausea of greater or less severity.**

If our views are correct, nausea and vomiting are of fundamental
importance for the protection of various organs and tissues against
poisoning (using that term in its broadest sense) and diiiferent organs
have developed this protective mechanism independently of the irritant
action which these substances exert on the gastric mucous membrane.
It is especially interesting in this connection to obser\-e that rodents,
which are incapable of vomiting, have developed several different, and
apparently independent, methods of protecting themselves against the
toxic action of digitalis bodies on the heart, and also against the
injurious actions of various other vegetable poisons.

SUMM.^RY

1. Several of the digitalis bodies, including ouabain, amorphous
strophanthin, digitoxin, digitalein and tincture of digitalis, were used
in about eighty experiments designed to determine the seat of their
emetic action in the cat and dog.

2. In one series of experiments the carotid and vertebral arteries
were tied, after which the intramuscular or intravenous injection of
ouabain or digitoxin caused nausea or vomiting. Two animals failed
to vomit owing to severe depression.

3. Nausea and vomiting could not be elicited in cats by perfusing
the brain and medulla with diluted defibrinated blood to which ouabain
had been added.

4. Nausea and \omiting were induced in cats by the injection of
ouabain into the femoral vein in experiments in which the poison was
prevented from reaching the medulla by perfusing that organ with
unpoisoned diluted defibrinated blood for a period of ten minutes,
during which all but traces of the poison left the circulation.

5. Nausea or vomiting could not be induced in any of the experi-
ments on fourteen cats and one dog in which ouabain, amorphous
strophanthin, digitoxin, and digitalein were applied to the floor of the
fourth ventricle — the vomiting center of Thumas — in widely van-ing
amounts.

6. The application of small amounts of digitalis bodies to the floor
of the fourth ventricle — the vomiting center — is without influence on
the vomiting reflex and the subsequent intramuscular injection of
ouabain or other digitalis body causes emesis in the same way as it

8. One of us is frequently trouliled with nausea following certain types of
aderate exertion that induce some cardiac irregularity and rapid pulse.

704 ARCH 11' ES OF IXTERXAL MEDICI XE

does in the normal animal. Large doses applied to this area depress
the center, and vomiting cannot then be elicited by the intramuscular
or intravenous injection of a digitalis body.

7. Apomorphin hydrochlorid causes emesis in dogs when it is
applied to the vomiting center in amounts corresponding to 0.0001 mg.
per kilogram of body weight

8. Cutting the vagi at the level of the sixth cervical vetrebra (with
tracheotomy) does not interfere with the emetic action of an intra-
venous injection of ouabain.

9. Section of the cord above the level at which the sympathetic
cardiac fibers enter it, or removal of the stellate ganglia, usually pre-
vents nausea and vomiting after the administration of digitalis bodies.

10. The severing of all nervous connections between the heart and
the medulla always prevents nausea and vomiting after the injection
of moderate doses of digitalis bodies.

11. Removal of the celiac plexus does not interfere with the emetic
action of the digitalis bodies except in so far as the operation causes
depression.

12. The administration of mercuric chlorid through a stomach tube
in doses of 50 mg. per kilogram of weight causes emesis in cats in
which the spinal cord has been cut at the level of the second thoracic
vertebra and in those in which the stellate ganglia have been removed
and the vagi have been cut.

CONCLUSIONS

Digitalis bodies cau^e reflex nausea and vomiting through their
direct action on the heart.

The afferent impulses pass from the heart to the vomiting center
in the medulla, by way of the sympathetic mainly, in |iart, by way of
the vagus, probably.

Nausea and vomiting accompanying \arious circulatory distur-
bances, and more particularh- those of cardiac origin, ac(|uire a new
interest for the clinician in the light of our results.

A theory relating to the physiology of nausea and vomiting is
submitted.

THE ALKALI RESER\E IX PULMONARY
TUBERCULOSIS *

DAVID S. HACHEN. B.S.. M.D.

CIXCl.VXATI

The question of acidosis in tuberculosis is still a debatable one.
Pottenger ^ states that "there are probably many factors present in
tuberculosis which have a tendency to increase acidosis, such as deficient
intake of oxygen, deficient excretion of carbon dioxid, which occurs
particularly in the disease, as a result of diminished pulmonary area,
etc. Klebs - takes the opposite view, that "the gaseous metabolism in
tuberculosis is but slightly, if at all, altered, the system accommodating
itself to the lessened lung area, and, as is usual in the body, accom-
plishing the same amount of work with the decreased amount of tissue."
A. Loewy, and Kraus and Chvostek " found a moderate increase in
oxygen intake and carbon dioxid excretion in cases of pulmonary tuber-
culosis. It will be one of the objects of this work to show that in far
advanced cases of pulmonary tuberculosis with considerable destruc-
tion of lung tissue by cavitation and caseous bronchopneumonia, there
is only a slight decrease in the blood alkali reserve, at no time approach-
ing a real acidosis.

The cases were carefully selected from 213 patients having every
variety of lesion, the far-advanced type predominating. The method
of Van Slyke * was used to determine the bicarbonate content of the
blood plasma in terms of the percentage by \olume of carbon dioxid.
Using the precautions outlined by Van Slyke, 10 c.c. of blood was
drawn from a median vein at the elbow, placed in a centrifuge tube
containing 5 drops of a 20 per cent, solution of potassium oxalate, and
covered with liquid petrolatum. The blood was centrifuged and the
carbon dioxid combining power of the plasma determined within three
hours in every instance. The temperature, pulse and respiration of
each patient was taken from fifteen minutes to half an hour after the
blood was drawn. The first specimen of urine passed by the patient
following withdrawal of blood was examined for reaction, and at the
same time tests for urpchromogen and diazo substances were made.
The reaction of the urine was determined by the use of a 0.2 per cent.

* From the Percy Shields Memorial Research Laboratory, Cincinnati Tuber-
culosis Sanatorium, and the Department of Bacteriology, University of Cincinnati.

1. Pottenger, F. M. : Clinical Tuberculosis, V. 1:456, 1917.

2. Klebs, A. C. : Tuberculosis, 1509, p. 296.

3. Loewy, A., Kraus & Chvostek : quoted from Arnold Klebs.=

4. Van Slyke. D. D. : Method of determining carbon dioxid and carbonates
in solution. /. Biol. Chcni. 30:.W (June) 1917.

706 ARCHIl'ES OF IXTERSAL MEDICI XE

solution of methyl red in alcohol. One drop of this reagent was added
to 5 c.c. of clear urine in a test tube, the contents shaken and the
color reading made by looking through the depth of the fluid. A
distinct canary yellow color imparted to the urine was indicative of a
urine alkaline to methyl red, i. e., having a hydrogen-ion concentration
less than pn 5 ; an orange color pointed to a urine neutral to methyl red ;
i. e., a hydrogen-ion concentration of />h 5 ; while a red color showed a
urine to be acid to methyl red or having a hydrogen-ion concentration
greater than pa 5.

An accurate check on the clinical conditions of the patients was
kept, and the cases were classified in four groups as follows :

Condition 1. — '"Up-patients," requiring only a minimum of rest
hours. Clinically, these patients were in good condition, and usually
did small chores in the ward kitchen or main dining room.

Condition 2. — Patients who were put to bed because they were
coughing a little, running a slight afternoon fever denoting some
active lung lesion. Clinically, these patients felt well and were kept
in bed with difficulty. They were allowed toilet privileges only.

Condition 3. — These w^ere bed-ridden patients, who ran a high
afternoon temperature, i. e., over 100 F., coughed a good deal, pro-
duced considerable sputum daily, sufifered with chills and sweats, and
were slightly dyspnoeic at times. Clinically, these patients were mani-
festly ill.

Condition 4. — The.se patients suffered from an accentuation of
symptoms outlined under Condition 3, and were considered in an
immediately dangerous state. They would sometimes become slightly
cyanotic with approaching death.

n.^T.A .\ND KESULTS

The cases selected were males and females, white and colored, old
and young, having every variety and severity of lesion. Sixty-seven
determinations were made on sixty-one cases. In six cases, two
determinations were made about a week apart in an endeavor to obtain
records as close to death as possible.

A change for the worse in the clinical condition of the patient was
always accompanied by a mild corresponding drop in the alkali reserve
(Table 1).

In five of the six cases in which two determinations were made
between five and sixteen days apart, a decrea.se of from two to eight
points in the alkali reserve was noted (Table 2).

Comparison of the alkali reserve of the blood with the reaction of
the urine in the bladder in a series of cases showed a tendency for the
former to decrease, as the reaction of the urine passed from alkaline
to acid (Table 3).

HACHEX—PULMOXAKV TUBERCULOSIS 707

A comparison of the alkali reserve with the temperature of patients
showed that an increase in temperature above 100 F. was usually
accompanied by a drop in the alkali reserve (Table 4).

TABLE 1.— Blood Alkali Reserve of Slxty-Oxe Patients*

Clinical Condition 1 2 3 4

Xumher of Determinations 22 23 13 9

Alkali Reserve Range 48-70 50-73 52-62 50-62

Average Alkali Reserve 61.1 59.9 56.4 54.4

* Xormal blood alkali reserve 53-78.

TABLE 2. — Blood Alkali Reserve ix Six Special Cases

Case Xinnbt-r Date

E-186 3/31/21

4/7/21
F-30 ■ 3/31/21

4/7/21
F-67 4/2/21

4/18/21
F-70 4/2/21

4/7/21
E-178 4/2/21

4/17/21
E-308 4/7/21

4/18/21

TABLE 3. — Comparison of Alkali Reserve with the Reaction of
Bladder Urine

Reaction of Urine * Alkaline Neutral Acid

Xumber of Determinations 19 23 18

Range of Alkali Reserve 52-73 50-68 48-70

Average Alkali Reserve 61.1 58.0 57.7

* Alkaline to methyl red; hydrogen ion concentration less than /)h5. Xeutra!
to methyl red; hydrogen ion concentration, />H 5. .\cu\ to methyl red; hydrogen
ion concentration greater than pH 5.

TABLE 4.— Comparison of Alkali Reserve with Temperature

Temperature 100 F and under Over 100 F

Xumber of Determinations 37 27

Range of Alkali Reserve 48-73 50-69

Average Alkali Reserve 60.4 56.6

There was apparently no correlation between variations in the
respiratory rate of patients and the blood alkali reserve. In fifteen
cases with a respiratory rate over 25, the average alkali reserve was
58.0, as compared with an average alkali reserve of 57.?i in fifty-two

Alkali

Clinical

'Reserve

Condition

56

111

58

III

54

III

52

IV

58

III

50

IV

62

IV

58

IV

60

III

52

IV

62

III

59

IV

708 ARCHIVES OF IXTERXAL MEDICIXE

cases with a respiratory rate under 25. There was no correlation
between variations in pulse rate and the blood alkali reserve. In seven
cases where blood had been drawn one to sixteen days previous to
death, the blood alkali reserve ranged between 50 and 58, the average
being 52.7. Two cases of acute miliary tuberculosis, diagnosis verified
at necropsy, had alkali reserves of 52 each. In one case the blood was
examined forty-one days, in the other case seven days previous to
death.

The existence of positive urochromogen or diazo substance in the
urine was indicative of a tendency for the blood alkali reserve to
diminish. Eleven cases giving positive reactions showed an average
alkali reserve of 56.3. It will be noted in Table 1 that this figure is
very close to the average found in patients designated Condition 3.
this latter average being 56'.4.

T.\BLE 5. — The .\lk.\i.i Rlskrve in C.\ses with Extensive Luxg Destruction

Showing Number of D.ays Before De.^th Following the L-\st

Alkali Reserve Determin.xtiox

No of days before

Case Xo.
F-I07
F-7n
F-30
F-186
F-165
F-107
F-178
F-141
F-118
F-191
F-1 72
F-161

A careful study of the lung necropsy findings was made in twelve
cases by Dr. J. B. Rogers. Every case revealed extensive lung destruc-
tion with fibrosis, cavitations and caseous bronchopneumonia, and yet
the alkali reserve ranged between 50 and 63, the average being 56.2.

With the permission of Dr. J. B. Rogers a detailed account of the
lung necropsy findings is given in the first three cases outlined in
Table 5. '

C.\SE 1. — .Alkali reserve 5.S, one day before deatb.

Right Lung. — Practically tbc entire upper lobe is occupied by active acute
interlocular cavities surrounded by a caseous gelatinous pneumonia. Tbe
lower border of the middle lobe is occupied by caseous pneumonia, while tbe
apex contains a cavity approximately one inch in diameter. The upper half
of the lower lobe contains numerous cavities, while the lower half is infiltrated
with caseous pneumonia.

Left Lung. — The outstandinu feature is the presence of an empyema. .A
cavity is found rupturing into tlie pleural sac which contains approximately

death that alkali

.Alkali reserve

reserve was done

55

1

58

9

52

16

58

8

52

n

50

46

52

56

55

45'

52

90

66

102

60

90

63

99

HACHEX—PULMOXARY TUBERCULOSIS 709

500 c.c. pus. The entire lung is collapsed and compressed against the hilum.
Both upper and lower lobes are excavated by multilocular cavities surrounded
by fibrous tissue.

Case 2. — Alkali reserve 58, nine days before death.

Right Lting. — The upper lobe is honeycombed by large multilocular
cavities surrounded by a limited amount of fibrous connective tissue. The middle
lobe contains numerous small cavities, one-half inch in diameter. The lower
lobe is completely consolidated as a result of confluent caseation and broncho-

Left Lung. — The upper one half of the upper lobe contains large multilocular
cavities, while the lower one half is infiltrated with caseous broncho-pneumonia.
The upper one half of the lower lobe contains confluent caseous broncho-
pneumonia.

Case 3. — Alkali reserve 52. sixteen days previous to death.

Right Lung. — Practically the entire upper lobe is hollowed out by a cavity,
which is surrounded by confluent caseous bronchopneumonia. The middle
lobe is infiltrated with confluent caseous bronchopneumonia. The lower lobe
contains scattered masses of caseous bronchopneumonia.

Left Lung. — .\t the apex there is a cavity about one inch in diameter, sur-
rounded by caseous pneumonia. The remainder of the upper lobe is occupied by
smaller cavities which are surrounded by caseous bronchopneumonia. The
lower lobe contains patches of caseous bronchopneumonia.

The necropsy findings in the remainder of the cases are very similar
to those described above, the alkaH reserve of the blood being sur-
prisingly high in some of the cases with extensive Iitng destruction.
This state of affairs is in sharp contrast with that found in influenza
and influenzal bronchopneumonia where the alkali reserve dropped
as low as 24 in a severe case (Hachen and Isaacs '").

SUMMARY AND CONCLUSIONS

1. In tuberculosis there is a moderate depletion in the blood alkali
reserve only after the lesion becomes far advanced and is accompanied
by rather severe clinical symptoms, such as increased fever, chills and
sweats, slight dyspnea and general malaise ( Table 1 ) .

2. The blood alkali reserve in an individual case continues to
decrease slowly with approaching death until a minimum of 50 is
reached (Table 2).

3. The blood alkali reserve was 3 points lower in cases where the
urine as voided was neutral or acid to methyl red. An acid urine is.
of course, not an indication that an "acidosis" exists (Table 3).

4. An increase in temperature above 100 F. was usually accom-
panied by a decrease in the alkali reserve (Table 4).

3. There was no correlation between the respiratory rate and the
blood alkali reserve.

6. There was apparently no correlation between the pulse rate and
tiie blood alkali reserve.

S Hachen D. S.. and Isaacs. R. : The .Mkali Reserve in Epidemic Influenza
d Broncho Pneumonia. J. A. M. A. 75:1624 (Dec. H 1920.

710 ARCHIVES OF IXTERXAL MEDICIXE

7. The alkali reserve was comparatively low (52) in two cases of
acute miliary tuberculosis.

8. Urochromogen or diazo substance in urine was frequently found
when the blood alkali reserve was relatively low.

9. In thirteen cases showing at necropsy extensive tuberculous
involvement of all lobes, the average alkali reserve was 56.

10. Although there is a decrease in the alkali reserve as the case
advances, at no time is there a marked "acidosis" in pulmonary tuber-
culosis.

Archives of Internal Medicine

PIGMENT METABOLISM AND REGENERATION
OF HEMOGLOBIN IN THE BODY*

G. H. WHIPPLE

ROCHESTER, N. Y.

The words "pigment metabolism" should mean the general
exchange and balance in the body of all pigment substances — the
income of pigment forming material, the story of pigment building
in the body, the disposition of recognized end products in the body and
the elimination of certain pigment complexes.

Any junior medical student can relate the true story of pigment
metabolism in the human body. The medical textbooks have long
retold the story as illustrated in Figure 1 ; and it seemed to be one of
the facts in physiology which stood firm in the midst of progress and
newer investigations. Its very age gave it respectability and true
academic security of tenure. The time honored story is as follows :
Certain food elements and iron are constructed in the bone marrow
into a complex substance, hemoglobin. When the red cells are worn
out or destroyed, the hemoglobin appears as bile pigment in a quantita-
tive ratio. This bile pigment secreted into the intestine is changed
to stercobilin and in large part excreted in the feces. Some of it may
be absorbed and re-excreted by the liver; but, given a Hver abnor-
mality, it may escape the portal blood stream and be excreted by the
kidneys as urobilin.

A number of recent investigations have modified this picture
somewhat, as illustrated by Figure 2. Wilbur and Addis ^ suggest that
with the absorption of stercobilin the "pyrrol complex" may be split
off and reconstructed into hemoglobin. This is a very interesting
hypothesis and would be an example of a very pretty conservation
on the part of the body, but we shall review experimental observations
which we believe rule out this suggestion.

Our conception of body pigment metabolism may, perhaps, be
expressed diagrammatically as shown in Figure 3. One point in
particular deserves notice in that we no not accept as proved that

* From the George Williams Hooper Foundation for Medical Research,
University of California, San Francisco; Harvey Society Lecture, Jan. 7, 1922.
I. Wilbur and Addis: Arch. Int. Med. 13:235 (Feb.) 1914.

712 ARCHIVES OF IXTERXAL MEDICIXE

there is any absorption of stercobilin from the intestine. Granting
that urobilin may be formed in the liver, there is not a shred of
evidence, clinical or experimental, that stercobilin is ever absorbed
from the intestine. Most observers admit that a times urobilin may
be formed in the liver (Wilbur, Addis and many others), but forget
this fact when absorbed in a discussion of the formation of stercobilin
in the lumen of the intestine. We believe that stercobilin in the
intestine is as little concerned with this question of pigment metabolism
as is the stercobilin in the feces or the urobilin in the bladder urine.
Therefore in Figure 3 we know of no evidence for a line between the
circles indicating urobilin and stercobilin.

We have published (Whipple & Hooper -) evidence that bile
pigment is not necessarily related directly to destruction of red cells
and hemoglobin. This is indicated in Figure 3 by a direct line from
"pigment complex" to "bile pigment" and illustrates the relation of

Fig. 1. — Pigment metabolism in the human body as formerly believed to occur.

bile pigment production to food and other factors which can modify
pigment production quite apart from hemoglobin destruction. Grant-
ing that bile pigment production may be influenced by other factors
than hemoglobin destruction, we see how absurd it is to draw con-
clusions unreservedly as to blood destruction from the analysis of
stercobilin — for example, in pernicious anemia. The life cycle of the
red cell has been established (Eppinger and Charnas^) on such flimsy
evidence as the analysis of stercobilin.

That body protein as well as food factors are concerned in the
production of bile pigment and hemoglobin is easily established by
fasting experiments. Bile pigment excretion will continue during fast-
ing periods, and, more than this, we have shown that hemoglobin will
be formed in fasting periods not only sufficient for red cell maintenance
but for actual increase above a moderate anemia level.' It is obvious

2. Whipple and Hooper: Am. J. Physiol. 40:349. 1916.

3. Eppinger and Charnas : Int. of klin. Med., 1913. p. 'j

4. Whipple and Hooper: -Am. J. Physiol. 45:576, 1918.

WHIPPLE— PICMEXT METABOLISM

713

that disintegration products of body cells are used in the upbuilding
of hemoglobin and are related to the output of bile pigment, urobilin,
stercobilin, and urochrome.

BILE PIGMENT

Much of the older work on the biliary pigments has been reviewed
recently by Hooper and Whipple^ and need not be discussed at this
time. It has been claimed by some that hemoglobin introduced into
the blood stream will be quantitatively excreted as bilirubin in the bile.
Whipple and Hooper " have been able to show that no such quantita-
tive relationship holds for hemoglobin and bile pigment. It seems very
probable that much of the hemoglobin set free in the blood stream
may be used in the body economy for a variety of purposes — among
others the construction of hemoglobin for new red cells. For example.

Fig.

-Pigment metabolism as modified by recent investigatii

it is sufficiently well established that intravenous injections of hemo-
globin or the destruction of red cells in the body will aid in the
recovery from simple anemia with consequent upbuilding of new
hemoglobin." It is very probable, however, that the hemoglobin in
the blood stream is not used direct but only after being broken down
to the unit structural factors — whatever these may be. This point
is graphically illustrated by the double arrows between the "pigment
complex" and "hemoglobin" in Figure 3.

It is now generally accepted that true bile pigment can be
formed from hemoglobin within the body or by other than liver

5. Hooper and Whipple: Am. J. Physiol. 40:332, 1916.

6. Whipple and Hooper: Am. J. Physiol. 43:258, 1917.

7. Hooper, Robscheit and Whipple: Am. J. Physiol. 53:263. 1920. Itami :
Arch. f. exper. Path. u. Pharmakol. 62:104, 1910. Itami and Pratt: Biochem.
Ztg. 18:302, 1909.

714 ARCHIVES OF IXTERXAL MEDICIXE

cells. Whipple and Hooper' showed that this transformation could
be effected within two hours in the blood stream of the head and
thorax with complete liver exclusion. The same workers ° showed
that hemoglobin can be transformed into bilirubin in the serous cavi-
ties within a period of twelve hours. McNee '" has confirmed a part
of this work. It is probable that the vessel endothelium and
Kupffer cells are concerned in the vascular reaction. We believe that
this reaction is not a physiologic curiosity but one of considerable
importance in all conditions associated with escape of hemoglobin into
the blood stream — for example, paroxysmal hemoglobinuria, malaria,
toxic anemias, etc. Under such conditions we believe there is good
evidence that much of the hemoglobin is changed to bile pigment
and other substances quite apart from essential liver cell activity.

Fig. 3. — Conception of body pigment metaboli

It is well established^ that the bile pigment elimination in dogs
can be increased by a change in diet — for example, a sudden change
from a meat to a carbohydrate diet may increase the bile pigment
elimination more than 50 per cent. This can be repeated time after
time and it seems at least improbable that this reaction is dependent
on blood destruction. We may explain this reaction, in part, as
follows. The meat diet is normal for the dog. The increase in bile
pigment excretion due to carbohydrate excess may represent an
abnormal or alternative reaction — a deviation of pigment elements
and construction into bile pigment for elimination. It is possible that
some of these pigment elements concerned in this reaction might be

8. Whipple and Hooper: J. Exper. M. 17:612, 1913.

9. Hooper and Whipple: J. Exper. M. 23:137, 1916.

10. McNee: J. Path. & Bacteriol. 18:325, 1913.

WHIPPLE— PIGMEXT METABOLISM 715

available under favorable conditions (anemia) for hemoglobin con-
struction or under usual conditions (meat diet) for elimination
elsewhere than in the bile. Urochrome is a possible end product of
pigment elements and deserves much more study in normal and
abnormal conditions.

The term "pigment complex" is used in this paper to indicate a
group of substances which are essential parts of the mature body
pigments. It is obvious that certain food factors contribute to this
"pigment complex" as foods are directly concerned in the production
of new hemoglobin and the formation of bile pigments and urochrome.
It is equally clear that the body protein and cells contribute to this
"pigment complex," as all body pigments are produced in measurable
amounts in fasting periods. We believe that the evidence is sufficient
to show that as hemoglobin disintegrates in the body it also contributes
to the "pigment complex" and so influences, in a measure, the new
formation of hemoglobin. It is probable that only a small amount of
the destroyed hemoglobin is conserved in this fashion. The pyrrol
nucleus seems to be one of the factors which must be concerned in this
"pigment complex," and it is probable that all facts related to pyrrol
metabolism will have a direct relation to the complicated body pigment
metabolism. These points are illustrated graphically in Figure 3.

Bile pigments in the bile fistula animal are not increased by the
feeding of fresh bile pigments or of fresh or cooked blood or of
digestion products obtained from blood. '^ This might be assumed to
be from lack of absorption. There is no evidence that bile pigment
or stercobilin are absorbed from the intestine. However, it has been
shown that the feeding of hemoglobin will influence the curve of new
hemoglobin construction after anemias.' This indicates an absorption
of substances which are concerned with the "pigment complex" but
it is clear that these same factors do not influence the output of bile
pigments in bile fistula dogs — at least under the conditions of our
experiments. These experiments are much against the suggestion of
Addis (Fig. 2) that there may be a conservation of bile pigment
factors which are absorbed from the intestine and reconstructed into
hemoglobin. Perhaps the strongest argument against the absorption
of stercobilin and its utilization in body pigment construction (Fig. 2)
is the fact that bile fistula dogs under observation continuously for
two years or longer show no evidence of pigment lack, no anemia, no
fall in pigment production and no reaction whatever to the feeding of
bile pigments.

A study of the bile pigment output of the Eck fistula liver fur-
nishes some interesting facts to consider at this time. Dogs with

11. Whipple and Hooper: Am. J. Physiol. 42:256, 1917.

716 ARCHIVES OF IXTERXAL MEDICIXE

combined Eck and bile fistulas eliminate less bile pigment than con-
trols— sometimes only from 30 to 50 per cent, of normal.^- The
Eck fistula liver is functionally inefficient and there is no direct
contact with the portal blood. Both these facts may well contribute
to this low pigment output, but the main point to emphasize is that
the pigment output is influenced by liver function rather than by the
amount of hemoglobin waste products formed in the body. We have
ample evidence that various liver injuries will likewise depress bile
pigment excretion — again, clear evidence that the liver has a con-
structive function in producing bile pigments rather than a simple
passive eHminative function.

Bile fistula dogs with anemia give very interesting and complex
reactions to hemoglobin injections." In general, we may say that the
elimination of bile pigments is not in any way parallel to the amount
of hemoglobin injected. The same is true for control dogs with bile
fistulas but no anemia. There is some evidence in these anemia experi-
ments for conservation of certain of these pigment factors within the
body. We may suspect a reconstruction of some of these factors into
hemoglobin because of the anemia needs. There are some experi-
mental data in favor of this explanation.

Certain of our bile fistula dogs have developed peculiar diseased
conditions which may or may not be concerned directly with pigment
metabolism. For example, certain dogs lose great amounts of inorganic
salts from the bones, so much so that the ribs show very many "green
stick fractures." These bones are reduced to mere elastic shells and
the heavy long bones are likewise depleted of lime salts. There are
many suggestive points in these experiments which are of peculiar
interest at this time (diet and bony changes) but a discussion of this
complex question must be postponed for the present.

Splenectomy in bile fistula dogs " gives reactions at times which
are of the greatest interest to the hematologist. These dogs may
show blood crises much like those observed in pernicious anemia.
The color index may be very high for considerable periods — a most
unusual condition in dogs. At times such dogs show maximum pig-
ment production, which cannot be explained by destruction of red cells
and hemoglobin. A specific experiment [Dogs 16-41, Table 69 '•''] shows
periods of great bile pigment increase — -even six times normal per kilo
body weight. To account for all this bile pigment as derived from
hemoglobin we must postulate a complete destruction of all circulating
hemoglobin every four or five days or less, or a daily red cell replace-
ment of from 20 to 25 per cent. This is unthinkable in the light of

12. Whipple and Hooper: Am. J. Physiol. 42:544, 1917.

13. Hooper and Whipple: .^m. J. Physiol. 43:275. 1917.

WHIPPLE— PIGMEXT METABOLISM 717

our knowledge of red blood cell regeneration in the dog, and we can-
not imagine that all this pigment had been built up to hemoglobin
before being broken down into bilirubin. Rather we wish to assume
that the body is stimulated to a maximum production of pigment
substances — in part, toward hemoglobin, in part, to bilirubin and, per-
haps, to other pigment substances. The high color index indicates a
maximum saturation of corpuscles wnth the pigment hemoglobin (refer
to pernicious anemia below).

Study of the bile pigments present in human serum in health and
disease has been reported by a number of investigators (\'an den
Bergh," Brule,^^ Blankenhorn '" and many others). It is evident that
there are various free and partially bound pigments in the blood plasma
or serum. Some of these pigments may dialyze through a parchment
membrane and others will not do so. The significance of these various
types of pigments has been discussed at considerable length by the
different workers, but as yet there does not seem to be complete accord
as to interpretations. Much important information will undoubtedly
come from this work but wherever possible a more extended study of
the body, feces and urine pigments should be made simultaneously.

UROBILIN

In discussing the pigment substances in the urine we wish to use
the term urobilin to include the closely related substance urobilinogen.
We use the term stercobilin to indicate the same substances in the
feces. It seems to be accepted generally that stercobilin is formed in
the intestine due to bacterial action on bile pigments. It is assumed
solely on indirect and incomplete evidence that stercobilin is, in part,
absorbed from the intestine. We believe there is no evidence for this
assumption and much against it. This absorption of stercobilin should
be discredited until such time as some positive evidence of intestinal
absorption is brought forward. It is admitted by most clinical observers
that urobilin is present at times in human bile in the liver ^ or gall-
bladder.^' We have observed the presence of urobilin frequently in
bile fistula dogs during fasting periods— in fact, this may seriously
interfere with experimental work on bilirubin.^" We believe that all
evidence favors the production of urobilin in the liver and bile pas-
sages (cholangitis), its absorption at times into the blood stream from
the liver and subsequent appearance in the urine. It is probable that

14. Van den Bergh : Dcr Gallenfarbstoff im Blute, 1918.

IS Brule: Recherches recentes sur les icterus, Paris, Masson et Cic, 1919.

\6. Blankenhorn: Arch. Int. Med. 27:131 (Jan.) 1921.

17. Straus and Hahn : Miinchcn. med. Wchnschr. 67:1286, 1920.

18. Whipple, Hooper and Robscheit: Am. J. Physiol. 53:151, 167, 1920.

718 ARCHIVES OF IXTERXAL MEDICIXE

urobilin, like bilirubin, at times is produced in tissues other than the
liver (blood extravasations, pneumonia, etc.).

We do not wish to deny the value of urobilin determination on the
urine, feces and bile, but the deductions drawn from such observations
should be made with a proper conception of the various reactions con-
cerned; and absorption of stercobilin is not one of the factors to be
accepted at present. Studies of urobilin in the urine may indicate
certain liver abnormalities or pigment disturbances in other parts of
the body. Analyses of stercobilin give figures of great interest but
high figures may indicate overactivity of the liver in pigment produc-
tion rather than excessive blood destruction with elimination through
the liver of bile pigments or urobilin. The analysis of fresh human
bile obtained by means of the duodenal tube will give facts of much
value (Lyon and others^''). It should be remembered that such sam-
ples are suitable only for qualitative analysis. The amount of dilution
can never be determined but the ratio of bile pigments to bile acids
may be of considerable significance, although difificulties in analysis
are present. The presence of urobilin and abnormal elements in the
duodenal bile may contribute facts of great value for a complete
understanding of a complicated disease picture.

The various methods for determination of urobilin or stercobilin
are admittedly unsatisfactory. Moreover, certain changes in the feces
may make analysis inaccurate or quite impossible (constipation, diar-
rhea, etc.^). Attempts aimed toward accurate quantitative extraction
are reported by Hausmarin,^" Goiffon.^i Baumann ^^ and others. When
a simple and accurate quantitative method is at hand the work on
these problems will "be greatly facilitated. A method to give complete
satisfaction in stercobilin estimations must give accurate figures not
only for stercobilin but its various related substances which confuse
the picture in simple complications like constipation and diarrhea.

UROCHROME

Urochrome is the stepchild of the pigment family. Numerous
theories as to its origin have been advanced but it can safely be
asserted that its parentage is doubtful. It has been suspected that
urochrome was related to urobilin, to blood pigments, to urea and to
food pigments. These views have recently been reviewed by Pelkan,^'
whose work indicates an important relationship of urochrome to food

19. Lyon: J. A. M. A. 73:980 (Sept. 27) 1919.

20. Hausmann: Ztschr. f. exper. Path. u. Therap. 13:373, 1913.

21. Goiffon: J. Pharmacol. Chem. 21:286, 1920.

22. Baumann: Arch. Int. Med. 28:475 (Oct.) 1921.

23. Pelkan: J. Biol. Chem. 43:237, 1920.

WHIPPLE— PIGMEXT METABOLISM 719

protein. He brings evidence that carotin is not concerned in the
urochrome excretion.

Urochrome contains the pyrrol group,-* which is always of interest
in any pigment work. There has been no extensive work done to study
the appearance and relationship of this pigment with other important
body pigments in health and disease. It is at least possible that uro-
chrome forms one of the avenues of disposal of pigment elements
("pigment complex") even before such substances are built into
hemoglobin or bile pigments. These possibilities do not preclude
fluctuations of urochrome due to large intake of foods rich in pigment
forming materials. This substance should be studied simultaneously
with other pigments (blood and bile) in health and disease, in clinical
and experimental conditions.

LIPOCHROME

Lipochrome is a peculiar pigment which is thought to have merely
a passive function in the body with no relationship to the urobilin,
urochrome or other pigments containing the pyrrol nucleus. Schulze -°
states that lipochromes are very closely related to the yellow radicle
of chlorophyll and members of the group of carotinoid pigments includ-
ing carotin and xanthophyll. The lipochromes are soluble in fat and
fat solvents. DoUey and Guthrie^' state that these lipochromes can
be removed from the body fat almost completely by diet periods free
from carotin intake. Van den Bergh '* states that the lipochrome
content of the blood varies with the diet intake of these food pigments.
It is well known that the serum of diabetics is rich in lipochrome
pigment, and he attributes this to diet factors. At present we have
no reason to suppose that the lipochromes have any direct relationship
to the other body pigments, but we should not close our minds to this
possibility, especially in disease conditions, for example, hemochroma-
tosis and pernicious anemia.

IIEMOGLOIJIN

Hemoglobin is without question the most important and interesting
of all the body pigments. Much of the interest in other body pigments
comes from a relationship known or assumed to exist between these
various pigments and hemoglobin. A very closely related or identical
substance, myohematin f muscle hemoglobin) exists in the striated
muscle tissue of the body and undoubtedly plays an important part in
the rapid exchange of oxygen and carbon dioxid between the function-
ing muscle protoplasm and the circulating hemoglobin. Myohematin will
not be considered in detail in this paper, but we should always keep this

24. Weiss: Med. Klin. 13:659, 1917.

25. Schulze: Sitz. Ges. Nat. Freunde, Berlin. 1914, p. 398.

26. Dolley and Guthrie: Science 50:191, 1919.

720 ARCHIVES OF INTERNAL MEDICINE

substance in mind when we study various gases (for example, carbon
monoxid) which are absorbed by hemoglobin in the body. The genesis
of myohematin is of fundamental importance, and we have no right
to assume that the bone marrow cells are concerned in its production.
Is this substance formed from its very elements by the activity of
the muscle tissue alone? Or is some partially built up "pigment com-
plex" utilized by the muscle tissue in building this most complex
substance? How rapidly is the myohematin used up in the body as
compared with the daily wastage and repair of circulating hemoglobin?
It will be of great importance to study various conditions in which
myohematin may be present in abnormal amounts. It has been pointed
out recently -^ that this substance usually amounts to about 10 per
cent, of the total body hemoglobin, but much more work is needed
to ascertain the conditions under which the myohematin content may
be found to depart from normal. Such studies will give information
of much value for the complete understanding of the body pigment
metabolism. Hoagland ^' has recently completed experiments to show
that aseptic anaerobic autolysis of beef muscle will produce measurable
amounts of hematoporphyrin. He suggests that this is a normal
reduction product of hemoglobin in the body and a substance inter-
mediary to the final end product, bilirubin.

In the adult human it is generally assumed that hemoglobin is
fabricated in the protoplasm of red cell groups within the bone mar-
row. We have suggested that other tissues (for example, the liver)
may be concerned in building up "parent pigment substances" which
are essential to the proper construction of mature red cells containing
hemoglobin. It is worth while reviewing a few facts concerning the
development of red cells in the embryo. Sabin ^' has pointed out that
blood cells in the second day chick embryo develop from the endo-
thelial cells and angioblasts. Hemoglobin is present at this time but
the liver is not functionally active. This seems at first sight to
indicate that endothelial cells can produce red cells and the pigment
hemoglobin. This may in fact be true but we must not forget the
yolk sac with its various storage factors developed by the mature hen
and drawn on continuously by the developing embryo. The endothelial
cell, however, may be active in a variety of the body pigment reactions
and can almost certainly produce bilirubin from hemoglobin.*

Therefore, it is possible that these cells have a capacity to build
up pigment substances and so take part in hemoglobin production. We
recall, too, the interesting relationship in early fetal life (human)

27. Smith, Arnold and Whipple: .'\m. J. Physiol. 56:.ii6, 1921

28. Hoagland: J. .Agricul. Res. 7:41. 1916.

29. Sabin: Anat. Rec. 13:199, 1917.

WHIPPLE— PIGMEST METABOLISM 721

between the liver cells and the islands of blood forming cells. This
would indicate a possible relationship between the liver cell, the Kupffer
cell and the developing red blood cell. It is, of course, possible that
this relationship continues through the agency of the circulation during
aduh hfe.

DIET FACTORS AND HEMOGLOBIX

W'e have long been interested in the influence of various diet factors
on the regeneration of red cells and hemoglobin during periods of
simple anemia in dogs experimentally produced by hemorrhage. This
work was begun by C. W. Hooper and me as an outgrowth of our
bile pigment investigations and later continued with the cooperation
of F. S. Robbins. The investigation is now being continued by Rob-
bins and Whipple. We may refer to certain publications " for much
of the experimental detail, methods, protocols, etc. At the outset it
seemed obvious from simple experiments that hemoglobin regenera-
tion could be influenced easily by a variety of diet factors. We
thought it highly desirable to work with animals of sufficient size
and suitable type so that blood could be obtained readily for various
analyses by venous puncture. Dogs were obviously best suited for
these experiments. The removal of small blood samples is very easy
and does not complicate the regeneration curve as the amounts removed
are so small as compared with the blood volume (from 800 to 1500
c.c). These dogs are omnivorous and will eat readily all types of
food mixtures. Our routine experiments included careful determina-
tions of blood volume, hemoglobin and red cell hematocrit, red cell
counts from blood drawn by venous puncture and body weight.

Our experimental data give numerous examples to show the neces-
sity of such complete determinations and the mistaken deductions which
may be derived from incomplete experiments. This applies particu-
larly to experiments of long duration where the differences between
any given group and the controls are but slight. The general pro-
cedure is as follows: A group (usually four) of normal dogs is
standardized (blood volume, red cell hematocrit, red cell count, etc.).
These dogs are then bled one- fourth of their total blood volume on
two successive days, at times a third bleeding is used. After a rest
of one day they are again standardized and placed on the experimental
food mixture. Complete determinations are done thereafter each week
until the end of the experiment.

Two characteristic tables (Tables 1 and 2) may be cited as exam-
ples of the wide differences in hemoglobin regeneration which may be
associated with sugar feeding (Table 1) and with meat feeding
(Table 2). We see that the blood regeneration is but slight— a slight
increase over and above the maintenance factor in red cells and an

ARCHirES OF IXTERXAL MEDICIXE

TABLE 1. — Blood Regener.\tion After Sugar Feeding
Dog 17-28. White bull, female, adult.

1

Pigment Volume =
Hb. per Cent. Times
Blood Volume

Ulood Volume

1
1

i
s

a

!

to

J

a
»
d

1

g

1

1

i

d

5

^

S
.1
1

Remarks

C.c.

c.c.

c.c.

%

%

Kg.

C.c,

1/19

1.620 1 1.500

90O

60.0

108

0.76

..1

7,4

11.60

129

Fasting

1/20

Bled 375 e.e.

1/22

Bled 270 c.c.

Bled 105 c.c.

1/24

3S8 1 900

603

m

34.0

61

0.80

1.79

3,8

7,8

10.40

87

1/24

Diet: 50 gm. cane

sugar

25 gm. glucose, 40O c.c. w

ater

1.121

m

4,^7

39.0

«

0.62

1.64

5,2

7,2

9.40

120

2/ 9

636

1.027

637

390

38.0

0.65

1.63

4,8

6,2

8.90

11.')

2/lfi

634

961

586

375

39.0

66

0.69

1.69

5,6

10,0

8.50

113

Diarrh.+

2/23

541

933

562

373

40.0

58

0.56

1.45

5,2

9,0

8.00

11/

* Anisocytosis ol red cells.

Blood volume with dry oxalate. Hemoglobin with Sahli tubes.

Hemoglobin index equals hemoglobin per cent, divided by red cell hematocrit per cent.

TABLE 2. — Blood Regeneration After Beef Hea
Dog 18-116. Bull mongrel, female, young adult.

Times

1

i

1

1
>

1

1
>

1

d
n

J2

a

1

1

1

a

1

s

d

d
a

i

1

i

Remarks

C.c.

0,0.

Or,

%

%

Kg.

c.c.

12/ 2

2,120 ' 1,720

805

907

0.56

2.32

11,4

1.14

14.90

H5

12/ 2

Diet: Craekermc

land

nilk

12/ 2

Bled 430 c.c.

12/ 4

Bled 430 c.c.

12/6

939 1 1.182

803

369

31.2

7!)

0.57

2.54

6,9

21,2

14 5,=)

SI 1

„, .

12/ ;

Bled 300

c.c.

12 9

776

1,260 1 930 1 323 | 25.6 | 62 | 0.S4 | 2.38 | 3,7 | 17,5 | 13.95 • 90 1

12/ 9

Diet: 2.56 gm. cooked beef heart;' 610 gm. cooked beef llver'-lOO calories per kilo

12/16
12/23
12/30

1,082
1,890
2,270

1..130

1,685
1,648

844
860
747

476
818
902

35.8
48.6
.54.7

81 0.88
112 0.57
138 ....

2.25
2.30
2.50

4,6
9,9

18.2
11,5

14.65 1 91 1

15.26 1 110 1

15,50 1 100 1

1 1

WHIPPLE— PIGMEXT METABOLISM 723

even smaller increase in hemoglobin within a period of four weeks.
The contrast with meat and liver feeding is striking (Table 2) which
shows complete return to normal in a period of three weeks. Between
these extremes are all types of reaction, and it is easy to understand
that a diet containing several food factors may give a complicated
reaction. It is easy to understand that one food factor may be inert
in this reaction and a second factor may likewise be inert but the two
together may have a distinct influence on the curve of red cell and
hemoglobin regeneration. For this reason it is necessary to test a
given factor under a variety of conditions (supplementary feeding,
etc.) before we can feel sure that we understand its reaction under
anemia conditions. This is one of several reasons why the accumula-
tion of convincing experimental data is so time consuming. We cannot
accept the experiments of Downs and Eddy,^" who report a positive
influence of secretin on the production of red cells in anemia. They
record only the red cell count with no figures for red cell hematocrit,
hemoglobin or blood volume observations. That secretin may influence
the red cell and hemoglobin production may be true but this is neither
proved nor' disproved by their experiments.

MEAT

Our experiments show that diets of cooked beef muscle or cooked
beef heart are ver>' favorable for a rapid regeneration of red cells and
hemoglobin (Table 2 and others^'). Cooked liver ranks with cooked
muscle, and these food factors will effect a prompt reconstruction of
the anemia picture to normal. These favorable diet factors are also
potent when given after long periods of anemia and unfavorable diet
intake. This is the severest test for any diet factor as to its influence
on red cell and hemoglobin reconstruction. Certain diet factors may
give a favorable reaction if given at once after the anemia is pro-
duced but may give an unfavorable reaction if given at the end of
a long period of anemia and unfavorable diet intake.'^ Meat extracts
(commercial) are inert and possess none of the factors which influence
red cell and hemoglobin regeneration.

FOOD GRAINS

The common food grains (wheat, barley, rice) in the form of
cooked bread or crackers do not furnish many factors which promote
red cell regeneration. Full diets of these materials with skim milk
may eflfect a slow rise in the level of hemoglobin and red cells which
finally may reach normal in from six to eight weeks. More com-

30. Downs and Eddy: Am. J. Physiol. 58:298, 1921.

31. Whipple, Robsciieit and Hooper: Am. J. Physiol. 53:236. 1920.

724 ARCHIVES OF INTERS AL MEDICI XE

monly the return toward normal will not exceed 90 per cent, of the
initial level before the anemia is produced by bleeding. Casein and
skim milk may be ranked with these food grains as regards their
influence on hemoglobin regeneration and we may say that, as a rule,
these foods do not return anemia animals to a high red cell and
hemoglobin level.

FASTING

A comparison of anemia fasting experiments with sugar feeding
experiments shows that anemic dogs actually produce more red cells
and hemoglobin during fasting periods than during periods of sugar
feeding. The question of the "sparing action" of carbohydrates comes
into this reaction, but we must refer the reader to reviews ^* of this
subject in recent papers.^^ There is evidence from these experiments
that the body conserves with much care the various pigment con-
struction units, which are then recast into red cells and hemoglobin.
It is well to keep in mind the fact that normal dogs during periods of
zero nitrogen intake (fasting or sugar feeding) are able to form
hemoglobin and red cells over and above the considerable amount
needed for daily wastage and repair — also are excreting considerable
amounts of bilirubin, stercobilin and urochrome. These must all come
directly or indirectly from the host's protein. This careful conserva-
tion of red cell and hemoglobin construction factors must be of con-
siderable importance in the body economy.

IRON

Iron furnishes a never failing topic for discussion by internist and
physiologist alike. Clinical opinion favors the use of iron in simple
anemias, but we question whether this treatment is based on sound
evidence. At least, the treatment does the patient no harm and may
soothe the doctor's conscience, but we can find no convincing evidence
that it gives patients with secondary anemia any real benefit. This
discussion does not concern chlorosis, which appears to be a distinct
disease entity. All the experimental evidence indicates that iron is inert
in secondary anemias and has no influence on blood regeneration. Our
experiments '^ indicate that iron in the form of Blaud's pills is inert.'
Ferric citrate and ovoferrin have little or no influence on blood
regeneration — nothing to compare with favorable food factors (meat).
Hemoglobin, given by mouth, intraperitoneally or intravenously, does
influence the curve of red cell and hemoglobin regeneration in a posi-
tive fashion but not to the extent noted with potent diet factors (meat).
The iron may be concerned in this reaction but there is even more

32. Davis and Whipple: Arch. Int. Med. 23:689 (May) 1919.

33. Whipple and Robscheit: Arch. Int. Med. 27:591 (May) 1921.

WHIPPLE— PIGMEST METABOLISM 723

evidence in favor of the pyrrol complex. Likewise, arsenic in sodium
cacodylate or Fowler's solution is inert in these anemia periods. We
believe it will repay the clinical workers to record careful observations
on anemic patients, paying particular attention to various diet factors
proven to be potent in controlled experiments.

The relation of hemoglobin and chlorophyll has given rise to much
speculation in the past and there are many interesting possibilities
which call for more work. It has been pointed out that under usual
conditions iron is necessary in plant metabolism for chlorophyll pro-
duction but does not form a part of the chlorophyll nucleus.^^ In the
absence of iron plants can form chlorophyll if there is a supply of
pyrrol material.^= These observations are of particular interest when
we consider the influence of certain plant leaves rich in chlorophyll on
the blood regeneration of anemia dogs (see below). But the evidence
that iron is directly concerned in this reaction is not convincing.

It has been reported by Cloetta ^° that the iron present in hemoglobin
is not absorbed from the dog's intestinal tract, and he is able to
recover the iron quantitatively after blood feeding. Weber ^' has
reported experiments to indicate that iron lactate is inert in anemia
periods in experimental animals. However, he did not control his diet
and reports on hemoglobin values only. The anemia was produced
by pyrodin. His experiments are not convincing but give no evidence
that this iron salt is potent. Musser ^* has furnished more evidence
that iron is inert under experimental conditions. Iron was given as
ferrous sulphate in capsules and the dogs rendered anemic by repeated
bleedings. He gives complete data on hemoglobin, red cell count,
resistance of red cells to hypotonic solutions and blood volume. It is,
perhaps, unfortunate that a mixed diet of hospital food scraps was used
in all these experiments. As diet factors are important it is very desir-
able that the investigator knows the amount and type of food intake.
A mixed diet gives an opportunity for much variety and also choice
by the animal. As a rule, the blood regeneration is very rapid on a
mixed diet and gives less opportunity to demonstrate the influence of
any given factor on blood regeneration. Some experiments should be
included to show the influence of the given drug when administered
during a long period of feeding on a diet unfavorable to rapid blood
regeneration — for example, a 75 calory per kilo diet of bread and
skim milk.

34. Moore: Proc. Roy. Soc. Lond. 87:556, 1914.

35. Oddo and Pollacci : Gazz. Chim. Ital. 50:54, 1920.

36. Cloetta: Arch. f. exper. Path. 37:69, 1895.

37. Weber: Ztschr. f. Biol. 70:168, 1920.

38. Musser: Arch. Int. Med. 28:638, 1921.

726 ARCHU'ES OF IXTERXAL MEDICIXE

FATS
Unpublished experiments of Robbins and Whipple indicate that
lard is inert in various diets, and influences in no degree the regenera-
tion of red cells and hemoglobin in anemic dogs. Numerous experi-
ments with cod liver oil give no evidence that this oil influences blood
regeneration. In striking contrast stand the experiments with butter
fat, which indicate that under certain conditions some substance in
butter fat is able to influence in a striking way the curve of hemoglobin
regeneration and hasten the production of hemoglobin and red cells.
When we admit that butter feeding is potent in anemia regeneration of
hemoglobin, we can scarcely admit that it actually is built into the
hemoglobin molecule. But it may act in some way to facilitate the
linkage of the various complexes which go into the large hemoglobin
molecule. We have not as yet sufficient data on the influence of
butter feeding upon the output of other body pigments during anemia
periods.

FISH

Through the friendly cooperation of the California Packing Cor-
poration and the Alaska Packing Corporation we were able to test a
variety of food fish. These experiments will be published in the near
future. The highly pigmented cooked salmon muscle was used and
in a variety of diet tests was shown to be inert in anemia experiments.
It is of particular importance to note that this muscle pigment of
the salmon is not concerned in the upbuilding of hemoglobin in dogs.
We have some evidence that the myohematin of beef muscle is actually
concerned in hemoglobin regeneration, which is so rapid after beef
feeding. Whale meat was tested and found to act exactly like beef
muscle or other striated mammalian muscle. Clams, like fish, are inert
in anemia experiments.

VEGETABLES

The data concerning vegetables will be published shortly by Robbins
and Whipple. A number of vegetables have no effect on the curve of
red cell and hemoglobin regeneration after anemia. Carrots are inert
and deserve particular attention because of their high content of
carotin, a food pigment which does enter into the body fluids and
tissues. Dehydrated celery, parsley and Brussels sprouts are likewise
inert, and it is improbable that the dehydration was responsible for
the negative reaction as dehydrated spinach is positive and exerts its
usual influence on the hemoglobin reaction. Fresh beet tops are
negative, as compared with fresh spinach, which is positive and even
more potent than the dried spinach meal. Canned spinach is some-
what less potent than is the freshly cooked material. These observa-
tions show that by this physiologic test there is a distinct diflference

WHIPPLE— PIGMEXT METABOLISM 727

between the chlorophyll of sprouts, celery, parsley, and beet leaves
as contrasted with the chlorophyll of spinach.

PERNICIOUS ANEMIA

This is a diseased condition, little enough understood, which is
of the greatest interest to any person investigating general body pig-
ment metabohsm. Let us examine the facts as known, paying par-
ticular attention to this question of pigment metabolism. We find an
excess of pigment substances everywhere in the body — increase of
pigments in the liver cells, blood stream, bone marrow, feces and, at
times, in the urine. There is a high hemoglobin index or we may say
the red cells are saturated with hemoglobin as contrasted with the
50 per cent, hemoglobin content of a simple anemia. This suggests
that the body has an excess of pigment material available and is pro-
ducing various pigment substances at an abnormal rate of speed —
compare the experiment (the splenectomy bile fistula dog, No. 16-41 '')
which showed a high color index and enormous overproduction of
bile pigment.

We are told that the stercobilin analysis in a case of pernicious
anemia is an index of blood destruction. Let us examine some of these
figures and, further, let us assume the normal red cell count as 5,000,000
and the pernicious anemia count as 1,000,000 for the sake of simplicity
of comparison. If the anemia red cells disintegrate at the same rate
of speed as the normal control, and if these products result in bilirubin
and then stercobilin we must say the stercobilin figures should be one-
fifth of normal. But the stercobilin figures during periods of remis-
sion in pernicious anemia often exceed twice or three times normal
stercobilin excretion, making no allowance for similar pigments in the
urine. This can only mean that the pernicious anemia patient with
one-fifth the number of red cells and two or three times the amount
of stercobilin output must regenerate its total red cell mass every three
days instead of the assumed normal of every thirty days. The normal
replacement factor for red cells and hemoglobin is believed to be 3 per
cent, per day.^" We must postulate from 30 to 40 per cent, replace-
ment of red cells per day in a pernicious anemia case if we persist in
explaining the stercobilin content as being due to blood destruction.
Those who wish to accept this explanation are welcome to do so, but
it would be a fleeting and troublous life period endured by the red cell
in pernicious anemia !

Ashby *" has published recent observations to show that the life
cycle of the human red cell is variable and may be thirty days or at

39. Ashby: Jour Exper. M. 29:267. 1919.

40. Ashby: J. Exper. M. 84:127, 1921.

728 ARCHIVES OF IXTERKAL MEDICIXE

times 100 days. Further than this she submits observations which
indicate that in pernicious anemia the red cells exist in the circulation
at least as lopg as in the normal human case and perhaps for a longer
time. She questions the importance of an increased blood destruction
in producing the anemia of this disease.

We prefer to explain the observed facts as done in Figure 3,
assuming that there is a great stimulus within the body for pigment
production. It may well be that there is an overproduction of pig-
ments, including hemoglobin and bile pigments and other abnormal
pigments. For example, in certain cases of pernicious anemia Hooper
and Whipple have observed an unusual pigment in gallbladder bile
obtained at necropsy — a pigment which must be treated with active
acetaldehyd before it will give the usual bile pigment tests. It is
highly probable that this pigment, after a stay in the intestine, would
be reduced to stercobilin and take part in the familiar high reading
of stercobilin in such cases.

Our conception of pernicious anemia is that there is a scarcity of
stroma building material or a disease of the stroma forming cells of
the marrow which limits the output of red cell framework. There is
plenty of pigment material (an excess in fact) as evidenced by the
high color index or the saturation of the red cell with hemoglobin.
Wherever we meet with a high color index we should suspect some
deficiency in stroma construction or some overproduction of body pig-
ments including hemoglobin. Conditions of malignancy, for example,
with the hematology of pernicious anemia should yield information of
value when examined with these points in mind. Nothing in this paper
should be construed as minimizing the importance of stercobilin
analyses for we are confident that such information is of great
value. High stercobilin figures may be a very valuable diagnostic
aid in obscure cases of pernicious anemia, as claimed by Hausmann
and Howard.*^ That these figures indicate a corresponding destruc-
tion of red cells may be doubted and an overproduction of pigment
may be a safer assumption.

Under usual secondary anemia conditions we know that this physio-
logic anemia stimulus causes a rapid production of both stroma and
hemoglobin, the production of the stroma even outstripping that of
the hemoglobin, as indicated by the low color index. It is possible
that stroma production in the body may be fatigued more easily, as
suggested by certain anemia conditions of long standing, which may
show periods of high color index and a hematology suggesting per-
nicious anemia. Such cases are deserving of particular attention and
a most careful study of the complete pigment metabolism.

41. Hausmann and Howard: J. A. M. A. 78:1262 (Oct. 25) 1919.

WHIPPLE— PIGMEXT METABOLISM 729

The experiments of McMaster and Haessler *^ may have an inter-
esting bearing on some of these questions. They show that in rabbits
the injection of hemoglobin intravenously is a stimulus to the exten-
sion of bone marrow tissue in the long bones of the experimental
animals. Simple bleeding does not stimulate to the same degree this
"spread" of the bone marrow. They believe that there is no evidence
for direct utilization of the injected hemoglobin, which is probably
broken down before use. So it is possible that the pigment construc-
tion factors or "pigment complex" may be radicles which determine
the bone marrow spread.

HEMOCHROMATOSIS

This disease is characterized by abnormal deposits of various pig-
ments in many body cells and tissues. By most writers it is thought
that this increase in body pigments is to be explained by a great increase
in blood destruction. The mere fact that there is no evidence to sup-
port his hypothetical increased blood destruction has not deterred writer
after writer from this assumption. Sprunt *^ is probably the first writer
to object to this interpretation, and he points out convincingly the entire
lack of evidence for any increase in red cell destruction — further that
the evidence favors some profound disturbance in body metabolism
which permits the deposit or production of various pigments in the
protoplasm of various body cells. MacCallum ** has shown the pres-
ence of iron in the cystoplasm of ferment forming gland cells of all
descriptions. It is interesting to recall the presence of abnormal iron
containing pigments in these same cells in cases of hemochromatosis.
The liver and pancreas are especially involved in this reaction, and the
liver may contain 100 times the normal content of iron. Various
observers record normal stercobilin elimination in cases of hemo-
chromatosis. Wilbur and Addis ^ report a long series of observations on
a case of hemochromatosis to show normal stercobilin figures and fluctu-
ating amounts of urobilin in the urine. The necropsy disclosed cirrhosis
of the liver and the familiar picture of this disease. It would be interest-
ing to speculate about the pigment abnormalities in hemochromatosis,
but we must recall that actual observed facts are but few in number. It
can be said, however, that there is no evidence of an increase in red cell
and hemoglobin destruction. Stercobilin elimination may be normal
but we do not know anything as to iron intake and elimination. It
will be of great interest to study the various pigment factors in the
blood, urine and tissues. It is, at least, possible that the usual paths
of pigment elimination and disposal are, in part, blocked at the source

42. McMaster and Haessler: J. Exper. M. 34:579. 1921.

43. Sprunt: Arch. Int. Med. 8:75 (July) 1911.

44. MacCallum: Ergcbn. d. Physiol. 7:552, 1898.

730 ARCHIVES OF IXTERXAL MEDICINE

so that these pigments which may be present in various cells in traces
are unable to escape and heap up within the cell protoplasm. This
explanation suggests that various cell protoplasm may at times be con-
cerned with the building up of certain pigment substances rather than
with a simple passive storage. We may choose to look on this disease
as resembling diabetes in certain respects — one disease with an inability
to handle the carbohydrates, the other disease associated with abnor-
mal metabolism of pigment factors. Certain end products are toxic
in the first instance (diabetes) and again other products accumulate
in the body cells , (hemochromatosis) to their detriment and final
destruction.

In this connection we wish to refer to experiments of Rous and
Oliver,*^ who induced plethora in rabbits by intravenous injection of
red cells. After months these animals developed siderosis of the liver,
kidney and other tissues. Their experimental findings in animals have
many points of similarity to those observed in human cases of hemo-
chromatosis. Their conception that the essential disease factor in
hemochromatosis lies in the liver does not wholly satisfy the writer,
but their experiments bring some support to the explanation advanced
above in favor of a disturbance of the whole intracellular pigment
mechanism.

CONCLUSION

In conclusion, we may repeat that the various body piginents have
interesting relationships which, perhaps, are not as simple as com-
monly believed. There may be at times a parallel increase or decrease
of related pigments. We wish to make clear our conception of an
underlying "pigment complex" which may be an intermediary stage
in the development of mature pigments. Perhaps all pigment con-
struction units pass through this stage and, depending upon supply and
demand, are used to construct hemoglobin or bile piginent or are dis-
charged as urobilin or urochrome or related pigments, as illustrated
in Figure 3.

What may be the fundamental stinuilus for pigment manufacture
in the body is for future work to determine. The most important
single point which we wish to emphasize is that pigment production
depends upon constructive body cell activity — a dynamic function which
concerns the formation of hemoglobin and bile pigments as well as
urobilin and urochrome.

We recall that the output of bile pigment in the dog may be
influenced by a variety of food factors but not by the feeding of
bile pigment or hemoglobin. Bile pigment output depends on liver

45. Rous and Oliver: J. E.xpcr. M. 28:629, 1918.

WHIPPLE— PIGMEST METABOLISM 731

functional activity, and we believe it is a product of liver activity
not solely a passive elimination product coming from defunct hemo-
globin. Splenectomy and bile fistula combined may give a maximum
pigment output and a clinical picture resembling in some respects
pernicious anemia. There is convincing evidence that stercobilin is
not absorbed from the intestine and further that urobilin as observed
in the urine is formed in the liver or body tissues, not absorbed from
the intestine. Urochrome is a pigment too little studied. It may
represent a shunt for pigment building material not utilized in the
body. Lipochrome seems to be an inert pigment taken in with the
food and slowly eliminated.

Hemoglobin regeneration following anemia may be influenced by
a great variety of diet factors. Among the potent factors exerting a
positive influence on hemoglobin stand, first, red meat and cooked liver,
hemoglobin and butter fat. Then come spinach and full diets of com-
mon food grains and milk. Practically inert are other chlorophyll
containing vegetable leaves — celery, parsley, beet tops and sprouts. In
the same negative group are fish and clams, onions and beets and
animal fats, including lard and cod liver oil. Iron and arsenic in the
common drug preparations are likewise inert under these conditions.

Pernicious anemia and hemochromatosis show abnormalities of pig-
ment formation and disposal. We believe these two diseases have
this in common — a definite overproduction of pigment in the body
but not an increased destruction of hemoglobin and red cells. These
are pigment abnormalities which deserve the most careful study from
this point of view. Such studies should be sufficiently comprehensive
to include simultaneous observations of these pigment elements in the
blood, feces and urine, also in the bile and body tissues whenever
possible.

It may be that this paper is too speculative, but hypotheses are
of some value for further work even if there be little or no truth in
them. I hope to furnish more data in the near future bearing on
these questions.

OCHRONOSIS

WITH A STUDY OF AN ADDITIONAL CASE *
B. S. OPPENHEIMER, M.D., and B. S. KLINE, M.D.

NEW YORK CITY

Ochronosis is the name given by Virchow in 1866 to a condition
characterized by the pigmentation of the cartilages, Hganients, tendons
and of the intima of the large blood vessels of the body. In this first
reported case the pigment deposits appeared light gray, brown and,
in places, black. On thin section, however, the pigment was everywhere
found to be yellow or yellow-brown and for this reason the condition
was called ochronosis.

Since then, forty additional cases have been studied. From the
observations made it may be stated that ochronosis is a condition
dependent on a disordered metabolism of phenol or some of its deriva-
tives ; characterized by a pigmentation of the cartilages, fibrocartilages,
fibrous tissues and epidermis, as well as of areas of degeneration,
notably atherosclerotic plaques, albuminous masses and concretions.
A further characteristic is the presence of a dark urine due to alkapton,
derivates of phenol or to melanin.

The cases of ochronosis may be divided into: (a) those due to the
circulation in the blood of certain aromatic compounds with the excre-
tion in the urine of homogentisic acid; (b) those due to the circulation
in the blood of certain aromatic compounds with the excretion in the
urine of melanin; (c) those due to the circulation in the blood of
certain aromatic compounds following the external use of phenol.

The metabolic disorder responsible for the ochronosis in the first
group is a congenital one and characterized by an alkaptonuria. More
than one-half of all the ochronosis cases observed are in this group.
The metabolic disorder responsible for the ochronosis in the second
group results in an excretion in the urine of melanin. Only a few
of the ochronosis cases are in this group. The metabolic disorder
responsible for the ochronosis in the third group is an acquired one
dependent on the prolonged external use of phenol. Eleven of the
ochronosis cases observed are in this group.

Twenty-two of the cases in the literature are females; nineteen
are males. The average age of the patients at the time of diagnosis
was about 51 years. The youngest patient was 23 years of age. There

* From the Montefiore Hospital for Clironic Diseases.

* Part of the expenses of this publication were defrayed from
by the late Dr. H. S. Oppenheimcr.

OPPEXHEIMER-KLIXE— OCHRONOSIS 72i

is a tendency for this condition to occur in families where there has been
inbreeding.

The diagnosis offers no difficulty. The cartilages of the ears and
nose have a bluish tint. The fibrous tissue, especially about small
joints, has a bluish gray appearance. There may be dark pigment
deposits in the sclerae and patches of pigmentation of the skin. There
is an excretion of dark urine or urine which turns dark on standing,
due to the presence of alkapton body or derivatives of phenol, rarely of
melanin. The pigment may be excreted to some extent by sudoriferous
and ceruminous glands.

The most frequent complications in ochronosis are: (a) deforming
arthritis of the spine or larger joints, and (b) cardiovascular lesions.

HISTORICAL RESUME

1. Clinical. — The early cases of ochronosis were recognized clinically
by the pigmentation of the external cartilages. In 1892 V. Hanseman
reported a case in which dark urine was passed. Examination of
this urine was negative for alkapton body and melanin. In 1902
Albrecht and Zdareck, reporting the seventh case in the literature,
called attention to the association of ochronosis with alkaptonuria. In
1904, Osier likewise reported two cases of ochronosis with associated
alkaptonuria. No further observations on the nature of the process
were recorded until Pick reported a case of ochronosis undoubtedly
due to the prolonged external use of phenol. From his chemical study
in cases associated with alkaptonuria and in one following chronic
phenol poisoning. Pick concluded that in the ochronosis of endogenous
origin (congenital, associated with alkaptonuria) a melanin-is formed
by the action of an enzyme on circulating homogentisic acid and tyrosin,
and that in the exogenous form (due to phenol poisoning), a melanin
is formed by the action of an enzyme on circulating hydroquinone and
pyrokatechin. This explanation of ochronosis, advanced by Pick in
1906, has received no appreciable modification since. In 1908 Gross
and Allard reported a case of ochronosis with alkaptonuria in which
there was a deforming arthritis of the larger joints. Contrary to
Virchow's belief that the pigment was deposited in the inflamed cartilage
of the affected joints, they maintained that these arthritic changes were
specifically due to the irritation of the deposited pig:nent. More
recently, Soderbergh ^ called attention to a deforming arthritis of the
spine in four cases of ochronosis with alkaptonuria. Attention has also
been called ^ to the frequent association of cardiovascular lesions with
ochronosis and it has been suggested that these changes, like the
arthritic ones, are primarily dependent on the metabolic disorder.

1. Soderbergh: Xord. med. Arch. 48: Xos. 3 and 4, 1915.

2. Beddard: Quart. J. M. 3:329, 1909.

734

ARCHIVES OF IXTERXAL MEDICIXE

The accompaning table based on Kolaczek's ^ tables shows the fre-
quency of cardiovascular and arthritic changes in the various groups.

Classification- of Ochronosis Cases with Associated Lesions

Group

Reported by

¥rf:

Sex

Patho
logic
ExamI
nation

Arthritis

Cardio-
vascular
Lesions

4T?

59

55
44
63
bO

69
73
63

F

f
S

r

F

M

P

Yes
Yes
Xo
Xo
No
No
Yes
No

III
No

meager des
?
I

Pope 1906

Poulsen 1910

Poulsen, 1910,.

Beddard 1910

-1-

Andrews and Branson, ibi'o
(Keats), 1910 . .

Beddard and Plumtree. 1911. .

-

T tnl

52.45

8 F
3 M

5

2 t,

-

57
49
45

If
66
19
35
68
61
44
35
30
23

51

m

42

M
M
M

F
P
P
F
M
M

^
P

F

M

No

No
No

Tes
No
No

No

No

+
+

+

+
+

.

Osier, 1904

Ogden, lS9o and 1904

Allard and Gross, 1907 u 1908,
Landois 190S

9

Poulsen. 1910

Kolaczek, 1910 ;....

Kolaczek 1910

-

Kolaczek, 1910

-

.Tantke 1913

^

Soderbergh 1915

^

9 F
9M

2

9 +

P. Probably

alcaptoniiria

.Ubrecht, 1902

Clemens and Wagner. 1907-8. .

47
31

M

Yes
Yes

z

+

Total 2

49

0 F
2M

2

0

1 ?

d. Melanuria and
probably
alcaptonuria
Total, 1

63

F

Yes

+

e. Melanuriano
alcaptoniiria

Hecker and Wolf, 1899

Oppenheimer, Janney, Kline,

73
40

M
M

Yes
Tes

+

-f

4veri

M.O

0 F
2 M

2

2

f. Noaleapton-

41

M

Yes

-

mclanurla
Total. 1

B. Urinonotob-

67
44

3«
52
67

65

M

I
P

Tes

lei
Tes

IS

+
+

,

Helle, 1900

Hclle 1900

-f

53.5

IS

6

2

Average age

51

22 F
19 M

19

1^,

(

2 ?

3. Kolaczek:

Beitr. z. Klin. Cliir. 71:

254. IS

10.

^/v

Moderate ocliroiiotic pigTiicntatioii of ears, eyes and axillae; tra
nnchial cartilages.

OPI'EXHEIMER-KLIXE—OCHROSOSIS 735

In the lony-one cases of ochronosis a chronic arthritis of the larger
joints or spine has been noted in sixteen. The associated arthritis
has been more frequent in the ochronosis with alkaptonuria.

Cardiovascular changes have been noted in nineteen of the forty-one
reported cases. These occurred in about equal frequency in cases with
alkaptonuria and in cases following phenol poisoning. Not only was
there extensive pigmentation of the intima and endocardium in these
cases, but also not infrequently a serious chronic valvular disease.

2. Pathologic. — In 1866 Virchow reported on a necropsy in a male,
aged 67. with an aneurysm of the ascending arch of the aorta, head

Fig. 1. — Intense ochronotic pigmenlation of costal cartilages.

injury and terminal anasarca. The striking lesion, however, was the
intense pigmentation of all cartilages and tibrocartilages, with piginenta-
tion to a less extent of ligaments, tendons, perichondrium and
periosteum. In this first case there was also some pigmentation of
the intima of the larger vessels, especially the aorta, with intense
pigmentation of the sclerotic patches in this vessel. The intensely
pigmented areas were black or bluish black. The pigmentation of the
tracheal cartilages was ochre colored. Histologically, the pigment
everywhere was brown or ochre colored, hence the name, ochronosis.
Examination of the pigment in this first case by Kiihnc showed an
organic pigment having a definite similarity to liein;itin derivatives.

736 ARCHIVES OF IXTERXAL MEDICI. \'E

\'ircho\v suggested that in ochronoj-is there is an imbihition from the
blood of hematin derivatives occurring in areas poor in vessels and
nerves but exposed to irritation. He thought that the process was
analogous to the physiologic pigmentation of the rete malpighii, the hair
and the choroid and depended on a similar relationship. Furthermore,
he believed that there were certain conditions of the cartilages and
ligaments which might be considered lower grades of ochronosis. He
had occasionally observed that the semilunar plates of the knee joints
in old people had a dark yellow or brown appearance and the costal
and bronchial cartilages a dark yellowish brown color. In these

Fig. 2. — Diffuse ochrniiotic pigmentation of cartilaginou-s matrix of costal
rtilage with granular ochronotic pigment in perichondrium.

instances also, the pigment was deposited in the intercellular substance
and was quite homogeneous and diffuse.

\'ircho\v, in this first case, observed changes in the larger joints,
particularly the knees, similar to those in arthritis deformans. The
de]30sition of the pigmetit in these irritated areas gave additional proof
to him of his theory, mentioned above. In this first report, no mention
of the appearance of the kidneys is made and no mention is made of
granular ochronotic pigment.

In concluding his article, ^'irchow states: "I believe, therefore, that
the case here presented, because nf the intensity nf the ]iigmentation.
was only an excellent example f)f the more fre(|uem iKbronosi-."

OPPEXHEIMER-KLIXE—OCHROXOSIS

737

Hanseman * observed diffuse and granular ochronotic pigment in
the tissues. In regard to the pigment he states that it is produced in
soluble form in the body and in this form absorbed and fixed by certain
tissues having but little metabolic activity and in other places changed
by cells to granular pigment.

The classical paper on ochronosis is that by Poulsen,^ who studied
ten cases clinically and two after necropsy. He described the pathologic
changes as follows :

In all cases one finds a yellowish or brown melanin-like pigment which at
times is granular, at times stains the tissues diffuseh. This pigment is deposited

principally in the cartilages ; costal, those of the air passages and larger
joints. Those of the smaller joints are usually unpigmented. The pigment
is also present in all the fibrocartilages, such as the intervertebral discs and
in the pelvic and intersternal cartilages. The pigment deposition is less intense
in the perichondrium, periosteum, tendons, fascias and joint capsules. The
bones, although usually unpigmented have shown pigment in a few cases.
Outside of the skeleton, the pigment is deposited as a rule only in the endo-
cardium, intima of the larger blood vessels and kidneys: rarely in other places.
such as bits of cartilage in the tonsils, in connective tissue of the lung, and

Hanseman : 1892.
Poulson : Ziegler's Beitr

Path. .Anat. 47: 1910.

738

ARCHH-ES OF IXTERXAL MEDICIXE

thyroid gland, in the fatty tissue about the perichondrium and in the dura mater.
The pigment is frequently found in the sclerae, epidermis and in a few cases in
the nails. Pigment masses have been observed in the prostate by a few
observers, although the authors questioned their specific character. In the
cartilage it is deposited in the matrix : the cartilage capsule, and the cells are
faintly or not at all colored. Degenerated cells, however, are deeply pigmented.
In the other tissues, this pigment is at times in the cells, at times between them.
The pigment is excreted in the urine.

Fig. 4.

KKl'DKT OF C.\SK

Hislorv.—Mak, aged 40. pressor, admitted to Monlctiorc Hospital, Nov. 18,
1915.

Chief Complaints. — Lancinating pains along spinal column radiating along the
lower intercostal spaces to both sides of the abdomen; slight productive cough;
occasional hemoptysis ; chronic constipation ; general weakness ; occasional
spells of vomiting.

Family History. — Nega*ive for consanguineous marriage.

Past History. — Occasional attacks of influenza. Frost bite of er.rs three
years ago. Habits: Ten cigarets daily. Fight years before patient's admission
to the hospital he was supposed to have had sugar in his urine. Seven years
before admission he first noticed peculiar bluish discoloration of the cartilage
of each ear. ,

OPPEXHEIMER-KLIXE—OCHROXOSIS

73)

Present Illness. — Eighteen months before admission, while bathing, he expe-
rienced sharp stabbing pains along the spinal column, extending forward along
the costal spaces to both hypochondriac regions. He left the water at once and
went home. The pains, however, continued to grow more and more severe
until the following morning when he was unable to resume his occupation. In
addition to this sharp pain he noticed stiffness of all the back muscles. He
remained at home for the next six months where he was treated with no
apparent relief. He then visited Mt. Clemens. Mich. On his return from Mt.
Clemens he began to complain of a persistent cough accompanied by profuse
greenish-yellow expectoration, blood tinged only for a period of two days.

Fig. 5. — Ochrunutic pigmentation of atherosclerotic plaques of aortic and
mitral valves and in neighborhood of attachment of aortic cusps.

three weeks before admission. In addition he suffered with night sweats and
general weakness. He lost twenty-five pounds in weight during the first year
of his illness. .*\t the time of admission, cough and loss of weight were
almost negligible symptoms.

During the first twelve months of the present illness the patient was treated
at various clinics. For the past three months the urine has been reddish-black ;
the underclothes were often stained black. Associated with this there has been
marked polyuria and dysuria. The patient became frightened because of this
condition and discontinued taking some white medicine which he was then
receiving at the St. Paul's Tuberculosis Clinic and which he felt caused the
disorder. He claims that the urinary symptoms mentioned subsided when the
drug was discontinued and recurred when the drug was again taken. On being

ARCHIJ-ES OF IXTERXAL MEDICIXE

the drug he took

given various drugs to smell he stated that he was positiv
had the same odor as creosote.

Physical Examination. — The patient, an adult male, poorly nourished,
appears to be suffering from some chronic illness. Weight. 104 pounds. Gait
is very slow and careful. The sclerae of both eyes present a faint bluish
tint. In addition there is a wedge shaped bluish-black area of pigmentation of
the sclerae to the right of each cornea. Both ears show a peculiar leaden blue
discoloration of the cartilage. The same discoloration appears to be present
in the nasal cartilage on the right side. Both axillae are diffusely bluish green
in color. Some of this discoloration is removable by soap and water and is
apparently due to pigment from the sweat and sebaceous glands. There is
a pale, brownish diffuse pigmentation of the skin of the neck and temporal
regions. The fingers and toes are clubbed ; nails pale, not pigmented. Chest :
Supraclavicular fossae deep ; clavicles exceptionally prominent. Examination
of limgs shows few signs at right apex posteriorly suggestive of pulmonary

Fig. 6. — Fine ochronc
tubules of kidnev.

tuberculosis. Heart: Not appreciably enlarged. The first and greater part
of the second sound at the apex is replaced by a loud harsh murmur trans-
mitted to the axilla. There is some thickening of radial arterial walls;
pulse regular; good tension. Liver; Palpable 4 cm. below costal margin in
right mammillary line ; tender. Extremities : Reflexes increased. Vertebral
column; Absolutely rigid, presenting a general bow deformity. I^uinbar curve
obliterated. There is a great deal of tenderness on any manipulation of either
thoracic or lumbar regions of spine.

Laboratory Findings. — Sputum ; negative on first five examinations. On
sixth examination a few tubercle bacilli were found. Blood: hemoglobin, 80
per cent.; leukocytes. 15,000; 85 per cent, neurotrophils. Wassermann reaction
of blood, negative. Urine: First specimen reddish black when voided; second
specimen, when voided, light amber color, turning to yellowish black. Next two
specimens were voided black. The following three specimens were smoky
but on standing became black. The quantity excreted in twenty-four hours was
usually 500 c.c. ; specific gravity, 1.010. .Mbumin, marked trace. Sugar, slight
reduction with Fehling's. Examination for bile and blood negative. Occasional

OPPEXHEIMHR-KLIXE—OCIIROXOSIS 741

hyalin casts. Chemical analysis of the urine by Dr. Janney showed no homo-
gentisic acid. On the other hand, a pigment was isolated exhibiting character-
istics similar to the nielanins previously obtained from the urine and tumors in
cases of melanosarcoma.

Rocntgcn-Ray Examinalioit of the Bones' — Spine : .Almost complete calcifica-
tion of the intervertebral discs from the first dorsal down. The cervical spine
appears practically normal. The lumbar spine shows marked lipping of the lower
and upper borders of the bodies of the vertebrae (Spondylitis deformanr).

Pelvic Bones : Complete calcification of the interpubic disc. Moderate
amount of irregular outgrowtli along the outer portions of the crests and the
ossa ischii.

' *. o ^

.«: v-w- ' *.

%'-'% -^

*. -* -^, ^ ©

' -* ■ .,'-'?»«■''

^. ^^-^k-^i-:^ . ^

-> ..-»

1^-^ ^ .--

: ^^.^^.^-vrx" »^|^

j|fc*v.?. a^~

*---V'3^

.i'J? % . 1

Fig. 7. — Diffuse ochronotic pigmentation of renal casts; granular ochronotic
pigment granules in cells of ascending loops of Henle and collecting tubules.

Hips: Marked lipping of the upper portion of the acetabulum. Great
amount of calcification around the trochanter major, with some bony excrcscenses
at its base.

Legs: Some calcification along tlic insertion of the ujiper portion of the
mcmbrana interossea.

Skull : Marked thinning out of both clinoid processes. Complete obliteration
of the frontal sinuses.

Shoulders : The joints are free. The upper portions of the humeri show a
condition such as wc usually see in osteitis fibrosa ; rarefaction, lack of clear
demarcation between compacta and spongiosa and beginning cystic degeneration.

Clinical Course. — The patient vomited persistently, ran an irregular sub-
febrile temperature until the day of death when temperature rose to 103 F.
A half hour after death it was noticed tliat the entire eyeball, exclusive of the

6. We are indebted to Dr. Th. Scholz for the rocntgcn-ray report in this
case.

742 ARCHIVES OF IXTERXAL MEDICIXE

cornea, had become brownish black in color and the following day at the
necropsy it was noticed that the pigment in the axillae had become much darker
than during life.

Diagnosis. — The diagnosis of ochronosis in the case was readily made
because of the bluish discoloration of the cartilages of the ears and the skin
of the axillae; the pigmentation of the sclerae and the excretion of a dark urine
becoming black on standing.

Chemical Report.' — Examination of the urine in this case was repeatedly
negative for homogentisic acid (alkapton body). On the other hand, the
pigment obtained from the urine, from a costal cartilage and from the prostate
gave reactions for melanin. The pigment from these sources had similar
characteristics.

Pathologic Report. — Anatomic diagnosis: ochronosis. Pigmentation of costal,
tracheal, bronchial, auricular and xyphoid cartilages, intervertebral discs, aorta,
endocardium, prostate, skin, sclerae, kidneys; deforming arthritis of larger joints
and spine; subacute bacterial endocarditis, mitral valve; subacute glomerulitis ;
infarct, spleen ; healed pulmonary tuberculosis ; artherosclerosis of aorta, pul-
monary arteries, mitral valve ; pulmonary edema.

Xecropsy Record (abridged). — Necropsy performed 33% hours after death.
The body is that of a considerably emaciated adult, 153 cm. in length. The
skin in general is thin, sallow in appearance. There are tattoo marks on the
left forearm. The nails show moderate double curvature. The skin of both
axills, under the arms, has a purplish color. The cartilages of the ear are
deep blue in the inner portion, less intensely colored peripherally. The
sclera have a faint blue tinge, except just below the cornea of each side
where there is a much greater deposition of the pigment and, in addition, a
wedge shaped, brownish, green-blue area about 2x1 cm. (These masses were
not present during life, but noticed a few minutes after death.) Eyes : The
right pupil is slightly larger than the left which is of about average size ; both
eyeballs sunken. The ears are small. The auricular cartilages through the
skin appear leaden gray. There is a small nodule on the upper margin of the
left ear, grape seed in size, shows grayish-pink pigmentation in the deeper
portion. The external genitalia show no abnormalities, except a faint bluish
discoloration on the upper surface of the glans.

(The organs removed through abdominal incision).

He.\rt: Somewhat enlarged, weighs 360 gm. There is considerable diminu-
tion of fat below the epicardium. The right side of the heart shows no
abnormalities, except at the base of one pulmonary valve cusp and at its attach-
ment to the artery in two places there is bluish discoloration of the intima.
The left auricle is moderately dilated, the walls not thickened ; the endocardium
has the usual appearance e.xccpt at one place above the auriculovcntricular
ring where there are numerous, small friable vegetations. Mitral valve — the
aortic leaflet shows on its upper surface, near the auriculovcntricular ring, a
number of small friable, grayish vegetations. The other cusp is strikingly
altered. There is a large irregular, friable mass along the line of closure
and free edge, yellow in color, in part calcified ; the vegetation continues down
the associated chordae. The left ventricle is moderately dilated, not appreciably
tliirkened. The papillary muscles are stretched, somewhat flaltencd ; the
endocardium thin and glistening. Aortic valve ; cusps thin and delicate. At
the attachment of the cusps to the ventricle and aorta there is extensive bluish
black pigmentation of the endothelium over a considerable area. This pigmenta-
tion is visible also on the posterior aspect of the aortic leaflet of the mitral
valve. The base of the aorta shows numerous slight elevations, due to small,

7. Dr. \. W. lanncv has aln-adv published a report of the chemi.strv of
this ca.sc. Am. 1. \L Sc. 156:. =59. 191S.

OPPEXHEIMER-KLIXE—OCHKOXOSIS ' 743

soft, yellow, opaque patches in the intima. The coronary vessels are not
tortuous. The walls are somewhat thickened and show scattered soft yellow
opaque patches in the intima; just beyond the left coronary orifice there are a
few patches of bluish pigmentation of the intima. Left myocardium on section
pale and flabby. Here and there are gray flecks replacing muscle. There are
also gray streaks associated with the vessels. No abnormal pigmentation of
myocardium.

Lungs : The right lung weighs 590 gm. It is voluminous. The upper lobe
is strikingly cushiony, especially anteriorly. The lobe also feels soggy. The
pleura in general is thin and glistening, e-xcept at the apex where there are
numerous puckered scars to which are attached dense fibrous tags. Below these
pleural scars there is an irregular, indurated pigmented mass about the size of a
robin's egg. In portions of this scarred area there are small, dry, cheesy and
calcified masses. The remainder of the lobe has a watery, dull, pinkish red
color, mottled with black ; although crepitation is made out the air spaces
contain a considerable quantity of thin fluid. Dissection of the branches of the
pulmonary artery show a number of soft yellow patches in the intima. The
bronchi show nothing abnormal. The hilic lymph glands not appreciably
enlarged, show' intense black pigmentation.

The left lung w-eighs 650 gm. It is voluminous. The upper lobe is inelastic,
cushiony. The lower lobe feels soggy. Dissection of the vessels shows
atheromatous patches in the arteries, similar to those on opposite side. The
cartilaginous rings of the larger bronchi appear bluish through the mucosa.
On cross section,' however, they appear ochre colored. The pleura is thin and
glistening everywhere. On section the upper lobe crackles. A mottled pink
and black surface presents. The air spaces contain a small amount of thin
fluid, especially in the lower portion of the lobe. The lower lobe on section
shows a pinkish-red moist surface. Thin fluid exudes in considerable quantity
from the air spaces.

Liver: Weighs 1,750 gm. ; shows no macroscopic abnormalities.

Spleen: Weighs 350 gm. ; measures 16x9x4 cm. About twice average size.
It has the average consistency. The capsule is thin. Toward the upper pole
there is a triangular area with sides 2% cm. and base 1% cm., yellow in color,
opaque, depressed a few millimeters below the general level. On section of
the spleen a striking picture presents. The surface is soft and pasty, red in
color. Scattered throughout the pulp are numerous small gray areas about
pinhead in size. The pulp scrapes oflf readily on the knife. The trabeculae are
increased in number, but not in size. The depressed area noted on the surface
is found to be a part of a typical wedge shaped infarct, homogeneous through-
out, dry, yellow and opaque, except at the ape.x where for a considerable
distance the tissue has a decidedly bluish color.

P.ANCRF.AS .\Nn SupR.\REN.M.s : No appreciable abnormalities.

Kidneys : The kidneys together weigh 500 gm. Both are apparently alike.
Each measures 12^x8x6 cm. Each moderately enlarged. The capsule strips
readily, showing a smooth surface in which the veins are prominent. In adaition,
innumerable pinpoint and larger bluish black spots are seen. On section, a
striking picture presents. The cortex is quite uniform in width, averages
from 8 to 9 mm., has a watery gray-reddish appearance, streaked and dotted
with brownish and bluish pigmentation. The striations are not very distinct
but are fairly regular. The glomeruli are inconspicuous. Brownish and bluish
pigment streaks and dots are quite extensive in the medulla and most striking
in the papillae.

Bladder : Tlie bladder of average size, the walls of average thickness.
contains turbid urine." The mucosa is pale except for a few scattered areas of
injection, especially marked in the trigone. The prostatic urethra presents
a striking picture ; there are stony, bluish pigment masses varying in size from

744 ARCHU'ES OF IXTERXAL MEDICJXE

pinpoint to grape seed; in some places entirely covered by mucosa, elsewhere
only partially covered. There is no injection about these masses.

Prostate : The prostate is of average size and consistency. On section
it contains a number of bluish black pigment niasses varymg in size from
pinpoint granules, to several as large as peas. The nodules are stony m
consistency.

Seminal Vesicles : The seminal vesicles are thin walled, not pigmented.

Vessels : The aorta is elastic, the walls of average thickness, the circum-
ference in upper thoracic portion 5 cm. There are numerous rather broad
longitudinal yellow opaque masses in the intima throughout the length of the
aorta. Just at the commencement of the thoracic portion there is an athero-
sclerotic plaque which shows considerable bluish black pigmentation over a
surface of about a square centimeter. In addition there is a slight diffuse
bluish pigmentation of the intima for a distance of 4 cm. in the neighborhood
of the intercostal vessels.

Neck Organs: Only the lower part of the trachea was removed. This shows
a pale, thin mucosa through which the cartilaginous rings have a decidedly
bluish color. This is true also of the bronchi. On cross section, the pigmen-
tation of the cartilaginous rings is found to be central ; in some, it is most
marked on the convex portion. The outer rim of pigmentation has a bluish
cast ; the deeper portions are brown.

Intestines: There is some apparent hyperplasia of the lymphoid tissue of
the small and large intestines. In the colon there are also a number of irregular
areas from 1 to 2 cm. square, having a smooth, pearly scarred appearance with
thin brown pigmented periphery, suggesting healed ulcers.

Spine : The bodies of the lumbar vertebrae are considerably flattened, the
intervertebral discs are narrower than normal and almost bony in consistency ;
the striking change of the di.scs is the diffuse, intense bluish black pigmentation.
The anterior ligament of the spine macroscopically shows no pigmentation.

Thorax: The lower portion of the sternum and adjoining costal cartilages
and ribs were removed through the abdominal incision. The costal cartilages
present a striking picture; they are hard and everywhere show an intense
bluish black pigmentation. The removed ribs and portions of sternum, however,
show no apparent pigmentation.

Owing to the fact that permission was granted for a partial necropsy only,
the larger joints of the body could not be investigated.

Roentgenograms of all the joints_were made, however, and showed changes
characteristic of arthritis deformans of the spine with well marked changes of
the larger joint (hip and knee) especially about the attachments of the capsules.
The smaller joints showed very little change. (Dr. Th. Scholz).

Histologic Report. — Tracheal CARTtLAGES : Sections show diffuse pigmenta-
tion of the matrix about the cartilage cells and clumps of fine brown granules
in the perichondrium. Most of the pigment is deposited in the matrix and
immediately surrounding the cartilage cells.

Costal Cartilage: Section shows diffuse brown pigmentation of the matrix.
In addition, a number of degenerated cartilage cells contain diffuse and granular
brown pigment. The perichondrium is pigmented in places ; the pigment present
in the form of small brown granules.

Intervertebral Disc: Intervertebral disc considerably narrower than average,
in part composed of fibro-cartilage. in part there are large cartilaginous like
plaques. In the matrix of these latter there is diffuse brown pigmentation.
In the fibrous portion near the anterior ligament there is considerable granular
brown pigmentation.

Aorta : Section shows a few atherosclerotic patches in the intima, associated
with which there is a considerable amount of extracellular brown pigment in
diffuse and small granular form.

OPPEXHEI'MER-KLIXE—OCHROXOSIS' 745

Prostate : The architecture in general is normal. There are rather numerous
corpora amylacea in the glands. These vary in appearance. A few show
a large amount of diffuse brown pigment in the central portions, the peripheral
portions unpigmented, stained pink (cosin). Various stages of pigmentation
are seen, including large and small corpora amylacea, diffusely and homogeneously
brown stained. .About a few of the glands containing the pigmented corpora,
there are accumulations of round cells, principally mononuclears. A number
of the glands containing these masses show partial or complete absence of the
epithelium.

Endocardium : This is considerably thickened. To it is attached a large
thrombus mass, composed of strands of fibrin, red cells and fragmented
leukocytes. Another section shows an area of the thrombus in which there is
beginning calcification. In the deeper layers of the endocardium there are
small masses of extracellular brown pigment in the form of fine granules.
Section stained by Gram-Weigert stain shows in the outer portions of the
thrombus numerous small round diplococci, many in small chains.

Kidney : There is some distortion of the striations. In areas there is an
increase in the interstitial connective tissue; in some of these areas there is an
accumulation of round cells in considerable number. In a number of these areas
and also elsewhere the glomeruli have an altered appearance. The glomerular
tuft is adherent to the capsule in one or more places. In places the glomerular
sac contains amorphous, pink stained material and a few large mononuclear
cells. In a very few glomeruli there are a large number of mononuclear cells
within the sac, all filled with fine, brown pigment granules. The neighboring
convoluted tubules also show a deposition of a large amount of granular brown
pigment in the epithelium. In sections stained with silver nitrate, the ochronotic
pigment is found in the form of very fine granules in cells of the pro.ximal
convoluted tubules and in the form of larger granules in the intact and
desquamated cells of many of the ascending limbs of the loops of Henle and
the various collecting tubules. The cells of the distal convoluted tubules contain
the granular pigment in moderate amount. In general the pigment in the lumina
of the tubules is diffuse. In places, however, desquamated cells containing
granular pigment are present. In addition to these changes, a number of the
tubules contain nucleated cells mostly polymorphonuclear leukocytes and in
places the interstitial tissue shows accumulations of similar cells. In addition, in
the interstitial tissue of cortex and medulla there are scattered large mononuclear
cells containing brown pigment.

Spleen : Section shows a very large, triangular area of homogeneously pink
stained amorphous material in which phantoms of former splenic structures
are seen. In addition to the pink stained material there is, in places, some
nuclear dust and more strikingly there are clumps of intra and extra cellular
brown pigment. The greatest deposits of the pigment are found immediately
surrounding this infarct, in the new formed connective tissue, which is present
as a fairly wide band; the remainder of the section shows normal looking
trabeculae and vessels. The malpighian bodies are very small, and lessened
in number. There is, however, an increase in the nucleated cells of the pulp.
There are numerous plugs of cocci in the splenic capillaries. Another section
of the spleen stows the presence of a number of clumps of yellowish brown
pigment scattered throughout the infarct. In the center of this infarct there is
the remains oP a large blood vessel plugged with homogeneous, pink stained
material.

Liver: This shows considerable engorgement of the blood vessels in the
central portions of the lobules. There is a striking increase in the number of
nucleated cells in the capillaries. In one capillary a very large clump of cocci
is seen.

Michrochemkal Report. — In sections stained by Nishimura's method, the
ochronotic pigment, diffuse and granular, does not show the reaction for iron.

746 ARCHirES OF IXTERXAL MEDIC IS E

In formalin-fixed material, the diffuse ochronotic pigment is stained orange
red by neutral red (1 per cent, aqueous solution, three hours at 56 C.) ; the
granular ochronotic pigment, however, is not stained b\- this method.

The granular ochronotic pigment behaves microchemically very much like
the pigment of brown atrophy. Both are decolorized by (1) potassium perman-
ganate, sodium sulphite, oxalic acid. (Potassium permanganate, 1/4 per cent,
solution, one half hour; equal parts of oxalic acid and sodium sulphite, 1
per cent, solution, 10 minutes.) ; (2) surgical solution of chlorinated soda (from
15 to 30 minutes), and (3) bichromate sulphuric acid solution (potassium
bichromate, 10 gm. ; sulphuric acid concentrated, 12 c.c. ; water, 100 c.c.) one
half hour. Both the diffuse and the granular ochronotic pigment are stained
brownish black by silver nitrate ( fresh 2 per cent, solution silver nitrate,
twenty-four hours at 56 C). The form and distribution of the pigment are
best demonstrated by this method.

SUMMARY OF FCRTY-FIRST CASE

Clinical. — The diagnosis of ochronosis was made in this case because
of the bluish discoloration of the cartilages of the ears and skin of
the axillae; the pigmentation of the sclerae and the excretion of a
dark urine becoming black on standing. In addition the patient had
a deforming arthritis of the spine and larger joints and a mitral endo-
carditis : complications frequently present in ochronosis.

Chemical. — Examination of the urine was repeatedly negative for
alkapton body. The pigment obtained from the urine, from a costal
cartilage and that from the prostate of the case gave the reactions for
melanin. The pigment from these three sources had similar character-
istics. The chemical findings are in accord with the belief that
ochronosis is dependent on a disordered metabolism of phenol
derivatives.

Pathologic. — As in the cases previously reported, the cartilages
(costal, tracheal, bronchial, auricular and xyphoid), and fibrocartilages
(intervertebral discs) are deeply pigmented (bluish black). Large
stony masses of bluish pigmentation are found in the prostate and
prostatic urethra. The kidneys likewise show extensive pigmentation.
The endocardium, intima of the aorta and coronary arteries, skin and
sclerae are less intensely pigmented. The pigment is not deposited in
any quantity in intact intima and endocardium but in areas of degen-
eration in these structures, however, macroscopic deposits occur.
Diffuse ochronotic pigment is present in albuminous masses (renal
casts) and concretions (corpora amylacea of prostate). Fine pigment
granules are present in the epithelium of proximal convoluted tubules,
and coarser granules are present in the cells of the ascending loops of
TIenle, distal convoluted tubules and the collecting tubules.

The pigment is predominatingly diffuse in the matrix of the
cartilage and fibrocartilage and when associated with albuminous
masses and concretions. It is predominatingly granular in perichon-
drium, periosteum, tendons, fascias, connective tissue and in certain

OPPEXHEIMEK-KLIXE—OCHROXOSIS 7A7

renal cells. It is present in diffuse and granular forms in injured and
degenerated areas.

The histologic picture in the kidney sections suggests excretion of
the pigment by the cells of the proximal convoluted tubules. The
picture likewise suggests a partial reabsorption of the fine pigment
by the cells of the loops of Henle, distal convoluted tubules and collect-
ing tubules, and a transformation of the pigment into a more granular
form. The form and distribution of the pigment is demonstrated best
in histologic sections stained with silver nitrate."

We are indebted to Mrs. H. G. Friedman for her kind assistance in pre-
paring this paper.

8. Other references bearing on this subject are: Poulsen : Literature to 1910,
Beitr. z. path. Anat. u. z. allg. Path. 48:346, 1910; Literature to 1912, IMiinch. med.
Wchnschr. 59:364. 1912; Beddard and Plumtree : Quart. J. M. 12:505, 1911;
Umber and Burger: Deutsche, med. Wchnschr. 48:2337, 1913; Jantke: Mitt. a. d.
Grenzgeb. d. Med. u. Chir. 26:617, 1913; Heymann-Giessen : 1913; Vogelius :
Hospital tidende. 1164, 1914; Sprunt : Ochronosis, Nelson's Living Medicine
3:211, 1920; Howard: Ochronosis, Oxford Medicine 4:223, 1921.

AIDS TO BASAL METABOLIC RATE
DETERMINATIONS *

H. S. NEWCOMER, M.D.

PHILADELPHIA

This paper contains charts and tables so arranged as to facilitate
considerably the otherwise laborious calculation of basal metabolism
\alues. They are in such form as to be useful in calculating basal
metabolic rate from data obtained with any form of apparatus. They
are intended primarily for use in connection with determinations
made by some method involving the use of a Haldane gas analysis
apparatus. Such methods in common use are either that described
by Boothby and Sandiford ^ in which the expired air is collected in
a large spirometer, or the method of Douglas = in which a rubber bag
is used for this purpose and the volume of its contents measured with
a meter.^ In either of these methods the gasometric data include
figures for volume, temperature, pressure, and carbon dioxid and
oxygen content of the expired air. From these are to be computed the
respiratory quotient and oxygen consumption at standard conditions
of temperature and pressure. The figure for oxygen consumption,
when multiplied by the calorific value of oxygen, gives a figure for
calory production.

Certain instruments, as for instance the Benedict portable appa-
ratus,-* read the uncorrected oxygen consumption directly and a
respiratory quotient of 0.82 corresponding to a calory value per liter
for oxygen of 4.825 is assumed.

In calculating basal metabolic rate, the patient's calory production
is compared with the normal. It is generally customary to use as
standards the figures of Aub and DuBois, giving the normal calory
production per square meter of body surface. The body surface of
the patient is calculated from the height weight chart of DuBois.
There are also available the standard multiple prediction tables of

* From the Laboratory of the Henry Phipps Institute of the University of
Pennsylvania.

1. Boothby, W. M., and Sandiford, I.: Laboratory Manual of Basal Metabolic
Rate Determination, Philadelphia, 1920.

2. Carpenter, T. M.: Carnegie Inst. Wash. Pub. No. 216, 1915, p. 67.

3. Newcomer, H. S. : J. Biol. Chcm. 47:489, 1921.

4. Benedict, F. G., and Collins, W. K. ; Boston M. S: S. J. 183:449, 1920.

XEirCOMER— BASAL METABOLISM DETERMLXATIOX 749

Benedict, as well as certain other data, applying to children.=^ In its
simplest form the calculation is laborious and when it is desirable to
obtain the additional information to be had by the use of a Haldane
apparatus, the computation involves the expenditure of an appreciable
amount of time and gives opportunity for the appearance of numerical
errors.

The use of the charts and tables at the end of this paper restrict
this labor to the performance of a simple algebraic sum. In several
instances they condense hitherto available tables of many pages to a
single sheet. There follows a description of their derivation and
then directions for their use with examples.

Derivation of Formulae and Charts. — Charts 1 and 2 are intended
for use with the Haldane gas analysis apparatus and are used to
obtain hourly calory production. Chart 2 corrects the gas volume to
standard conditions of temperature and pressure and includes a
correction for water vapor.

Apparatus such as that of Benedict which reads oxygen consump-
tion directly, that is, which reads the diminution in oxygen volume of
a closed system including the patient's respiratory tract, does not involve
a correction for water vapor. The observed diminution in oxygen
volume is, therefore, to be reduced to standard conditions by a factor
which does not include a correction for saturation with water vapor.
Chart 3 supplies the logarithm of this correcting factor. The hourly
calory production may be obtained by multiplying the oxygen consump-
tion in c.c. per minute by 0.2895 (log. 0.2895=1.4617). The other
charts and tables have to do with basal metabolism standards. In
the computation of the charts and tables five figure logarithms were
used except that in the case of Chart 1 it was necessary to use seven
place tables.

Chart 1 has as abscissae the reading of the Haldane buret after absorption
of carbon dioxid, it being assumed that the .Haldane buret was filled in the first
place to the 10 c.c. mark at atmospheric pressure. The ordinates are the
Haldane readings after carbon dioxid and oxygen absorption. It is further
assumed that these readings are corrected for any calibration error which the
Haldane buret may have, and that the patient breathes outside air. This may
be either through a tube out the window, or, if the room can be well aired,
it suffices to leave the window open for some time before starting. The chart
has two families of lines; the horizontal ones give, by interpolation between them,
the value of the respiratory quotient, volume of carbon dio.xid production divided
by volume of oxygen absorption. The more vertical lines are the logarithms of
factors, portion of oxygen absorbed times the calorific value of one liter of
o.xygen at the equivalent respiratory quotient. The value of the logarithm
of this factor corresponding to any pair of Haldane readings is obtained by
interpolation between this family of lines.

5. Carpenter, T. M. : Carnegie Inst. Wash. Pub. No. 303, 1921.

ARCHIVES OF IXTERXAL MEDICIXE

TABLE 1. — Barometer Correction for Change in the Value of Gravity with

Latitude to Be Subtracted from or Added to the Observed Height of

THE Barometer Before Going to Charts 2 or 3

0°

5°

lO"

15

20°

25°

30°

35°

40°

45°

Latitude

90°

85"

so-

75-

70°

65°

60°

_i!L

50°

45°

no Dim.

Subtract

1.84

l.Sl

1.73

1.60

1.41

1.18

092

0.63

0.32

o.co

Add

760 mm.

Add

The eorrection varies with the height of the barometer. To correct for altitude subtract
0.14 mm. oJ mercury lor everj' one thousand meters above sea level. To correct for the
capillary depression of the mercury column see the text. Combine all of these
into one and use only if signiflcant.

T.-\BLE 2.— Logarithms of Standard Reciprocal Normal Basal Metabolism

Factors for Children 17 Years of Age and Under Referred to

Weight in Pounds. (Benedict)

wt.

Boys

Girls

Wt.

Boys

Girls

Wt.

Boys

Girls

Wt.

Boys Girls

6

T264

T274

36

2?498

"27523

66

2:333

T36O

96

2r270' 27288

6

.106

.093

38

.485

.510

68

.326

.351

98

.287

10

T996

"2:974-

40

.472

.498

70

.318

.344

100

.288

12

.908

.885

42

.455

.487

72

.311

.337

102

.285

14

.835

.813

44

.444

.475

74

.304

104

.284

16

.774

.758

46

.433

.403

-6

.297

.320°

106

.283

18

.724

.712

48

.422

.4.-.1

78

.291

108

.282

20

.682

.679

50

.411

.440

80

.285

.305*

110

.281

22

.645

.649

62

.400

.430

82

.280

112

.280

24

.614

.621

54

.390

.419

84

.274

.295

114

.279

26

.589

.595

56

.379

.4015

66

.294

116

.278

28

.567

.572

58

.368

.399

88

.292

118

.277

30

.547

.564

60

.358

.389

90

.291

120

.276

32

.529

.549

62

.350

.379

92

.200

124

.274

34

.513

.536

64

.342

.371

94

.289

128

.260* -.272

The figures are the logarithms of reciprocal total calories per hour of children neighing
from 6 to 128 pounds. The figures are calculated from data by Benedict and Talbot, and
Benedict and Hendry. The numbers with an asterisk are not duo to these authors but are
supplied as probable values in order to make the table complete

XEH'COMER— BASAL METABOLISM DETERMLXATIOX
TABLE 3. — Three Figure Logarithms

0

I . a;.

5

6

7 8 9

1

2

>

4

5

6

24
21
18
16

14
12
11

8
7

6

5

4

3

7

8

32
27
24

19

16
14

12

11
9

8

6

S
5

4
4
8

9

000

301

477
602

041

079

322 342

491 605
613 623

114

362
518
633

146

380
531
643

176

398

344
653

740

204

415

556
663

748

230

255

431 447

568 580
672 681

756 763

1

279
462

690

771

4

3
3
3
2

2
2

2

1

1
1

8
7
6
6
5

4
4
3

3

2

2

12

'I

8
7

6
5

5

4

2

2
2
2
1
3

16
13

11
10

8

6

6

5

3

3

2
2

20
17
15
i3
12

10
9

8

6
6

4

3
S
3
2

24
20
18
16
14

12
11

9

S
5

3
3

36
30
26
23
21

18
16
14

12
10

9

5

699

708 716 724

732

806
669
924
973

7

6

778

S45

954

7SS 792 799
851 E37 663
908 914 919
959 964 , 968

813

875

978

820
881
931
982

826 833
886 ; 892
S«9|944
987 991

839
898
949
996

1

1
1
1

6

5
5

4

«

1 2 j 3 j 4

5

6

7 J8

9

1

'

5

6

9

The figures of the table are the mantissae of two figure numbers. The columns at the
right are the proportional parts to b(- added to the mantissae to make them the mantissae of
three figure numbers. The characteristic of the lograrithm is the number less one of Integral
digits In the number.

TABLE 4. — Three Figure .^xtilog.^rith.ms

0

1

2

3

4

5

6 , 7

8 9

1

'

3

4-

5

6

—r

8

9

0

100

102

105

107

110

112

115 117

120 123

0

1

1

2

2

2

.1

126

129

132

135

138

141

145

148

151

155

0

1

2

2

3

3

.2

158

162

166

170

174

178

182

186

191

195

0

2

2

3

3

4

.3

199

201

209

214

219

224

228

234

240

245

1

^

3

4

4

5

.4

251

257

263

269

275

282

288

295

302

309

1

'

^

3

=

5

6

-1

316

324

331

339

847

355

S63

371

380 I 389

1
1

-

—

4

4
4

-

5
8

7
8

6

8
9

7
8

.6j

398

407

417

427

437

447

457 468

479 . 460

5
5

9

10

•■}

501

613

525

537

550

582

575 589

803 617

'

6

7

10

10

11

11
12

■■•1

631

646

661

676

692

708

724 741

769 776

2

*

7
8
9

10
11
12

12
13
14

13
14
16

.o'

794

813

832

881

871

891

912

933

955 977

1

2
2

9
10
11
11

13
14
15
16

15
16
17
18

]l
19
20

0

1

2

3

*

5

6

7

8 9

1

'

8

*

6

7

8

0

The figures of the table are the numbers corresponding to two figure mantissae. The
columns at the right arc proportional part? to be added to the numbers to make them
numbers corresponding to three figure mantissae. The characteristic of the logarithm Is
omitted in going to the table. The number is to be pointed ofl so as to have a number of
integral digits equal to the characteristic plus one. If the characteristic is negative the
niiiiiber is to he iinfacid by a number of decimal zeros equal to the characteristic plus one
(with regard to sign).

K

^^jI-irT-"^3--T

1---

^^

h

*-4

i '

^n =M

g a

I 5 2 & i £

Us «.■;! .13. i£i

I ! ! ? J I ° ° f

fe^ZS_Jafiii^§

-^

^ -^ ^-^

,^rN;-

lip

754 ARCHIVES OF ISTERXAL MEDICIXE

The derivation of these two families of lines is as follows:

Let x=reading of Haldane after CO= absorption

=100 — CO2 percentage in expired air
y=reading of Haldane after CO2 + O: absorption
(1) =100— (C02 + 0=) percentage in expired air

=nitrogen percentage in expired air
If 20.93 be tlie value taken for the oxygen percentage in the inspired air,
outside air, and 0.03 the carbon dioxid percentage, then 79.04 is the nitrogen and
other inert gas percentage in the inspired air. The fraction — =— represent: the
change in volume of the air due to alveolar gas exchange, and the ox}gen
absorption in per cent, is given by
20.93 y

(2) 0.= K

79.(1-1
where O2 is the oxygen percentage of the expired air. From (1) and (2)
20.93 y

+ v — 100+COi=K

79.04
99.97

79.04 ' ^~

79.04

These are a familv of lines of slope = .790637 giving constant oxygen

99.97
absorption K. The values of the calorific equivalent, C, of one liter of oxygen
for the respiratory quotients 0.707 to 1.00 as tabulated by Lusk" are given with
an error of less than one part in one thousand by the formula
R. Q. = 0.813 C — 3.103
* 100— x— 0.03

But bv definition R. Q. =

K
100— X— 0.03

K
100— X— 0.03 3.103

C= +

0.813K 0.813

100— X— 0.03 3.103

(4) CK = + K

0.813 0.813

substituting (3) in (4)

100— X— 0.03 3.103 /99.97 \

CK = + ( y - X )

0.813 0.813 \79.04 /

CK = 4.8274y — 5.04674X + 122.9643

= per cent, oxygen absorbed times its calorific value per liter.
This family of lines is plotte.d for values of CK whose logarithms are 0.60,
0.65, 0.70. etc., to 1.55. K is the number representing per cent, oxygen absorbed

bv the patient. The fraction 7^ is, therefore, a number which multiplied by

100 CK

the expired volume is the oxygen absorbed by the patient, j^ is its calorific

value and the logarithms of

assigned as signatures to the family of lines, namely, 2.60. 2.65, 2.70 to 1.55.

6. Lusk, G.: The Science of Nutrition, Ed. 3. Philadelphia, 1919, p. 61.

v^ \

I \: : - \:

•^

:-:XLv-\:iBl

•^---;-^^

o

^

^

'igT \M-'. ^V^:\-^

I^-I

U g E
So I

Et5 c.

756 ARCHIVES OF IXTERXAL MEDICI. \E

When the respiratory quotient is greater than unity the heat production is
due to the oxygen consumption plus the heat produced in the transformation of
the carbohydrate into fat during which the extra carbon dioxid was produced.
The heat value of this transformation, according to Lusk, is 0.8 calories per
liter carbon dioxid. For readings below the line RQ =: 1 carbohydrate is
being burned at a rate producing an oxygen percentage absorption correspond-
ing to a CK line of the reading where C is the 5.047 of RQ = 1. In addition,
carbohydrate is being converted into fat with a carbon dioxid percentage pro-
duction equal to the difference betv^'een x where the CK line crosses RQ = 1,
and the observed x. To this additional CO2 there corresponds a calory
production of 0.8 per liter. In order to calculate the total calory factor, we
may draw a line of the family K:= constant through the point given by the
intersection of the CK line and RQ=:1. If this be considered as a new CK line
it is a line along which neither C nor K changes. The new calory factor line is
then one passing through the CK, RQ point and having a slope sufficiently
steeper than the constant K line to make an observed point on the constant
K line and one x unit smaller than the x of the CK. RQ point lie 0.8 of the
distance between the new line and a next parallel one having a value one
unit larger. Therefore, below the line RQ = 1, the constant calory factor lines
bend slightly, as shown in the chart, and for readings below the line RQ = 1,
the patient is putting on fat at the moment in question.

Chart 2 is computed from data by Kaye and Laby.' It is a family of curves
giving by interpolation the logarithms of the factor reducing the observed air
volume to 0 C, 760 mm. of mercury at latitude 45 degrees and sea level, dry,'
together with a correction to O'C. of the mercury columns and brass scale. The
corrections for brass and glass scales are nearly the same. At 10 C. the cor-
rection for glass to be subtracted is 0.07 mm., and at 34 C, 0.25 mm. greater
than that correction for brass which has been subtracted from the barometric
height in computing the curves. The table does not include barometer cor-
rections due to the change in 'the value of gravity with latitude and height
above sea level. If it is desired to correct for latitude, add or subtract from
the observed barometer the figures of Table 1 before reading from the chart.
The change in gravity due to altitude is such that for every one thousand
iTieters above sea level one should subtract from the observed barometer
approximately 0.14 mm. of mercury. It may be proper in addition to correct
for the capillary depression of the mercury coluinn. This correction is zero
if both mercury levels have the same cross section, as is the case with the
usual all glass barometer. For barometers having a mercury reservoir the
figures to be added to the barometer reading for bores of 4, 5, 6, 7, and 8 mm.
are, respectively, 1.2, 0.7, 0.4, 0.25. and 0.2 mm., provided the height of the
meniscus itself is 0.6 mm. For heights of the meniscus greater than this and
up to 1.6 mm. these figures are to be increased uniformly per 0.2 mm. of
height respectively by 0.4, 0.25, 0.2, 0.15, and 0.1 inni. These corrections for
gravity and capillary depression may of course be determined at any place
once for all and combined into one single correction for the barometer.

Table 2 is derived from data by Bendict and Talbot' and Benedict and
Hendry." It gives the logarithms of the reciprocal of the total number of
calories produced per hour by children seventeen years of age and under. The
numbers are computed from the tables of these authors. The numbers with an
asteiisk are supplied to make the table complete, the two for girls being easily
inserted by interpolation. The additional numbers for boys cannot be supplied
with the same assurance. There should be a continued gradual decrease in the

7. Kaye, G. W. C. and Lahy.T. H. : Talde of Physical and Che-nical Con-
stants, Ed. 2, London. 1916.

8. Benedict, F. G.. and Talbot, F. B. : Carnegie lust. Wash Pub. No. .302,
1921, p. 206.

9. Benedict, F. G.. and Ilendrv. M. F. : Boston M. & S. J. 184:329. 1921.

758 ARCHIVES OF IXTERXAL MEDIC IS E

logarithms with increasing weight. The least decrease that could be expected
is to the figure 2.260 at 128 pounds. A smaller figure could not well be chosen
because of the fact that in adults of the age of twenty-one years, having the
extreme range of weights and normal stature (Prudential 1912 statistics;, the
normal calory production for males (Benedict multiple prediction formula) is
not more than 3 per cent, higher than that of females. The antilogarithms of
2.260 and 2^272 differ by about this 3 per cent., and if the first were smaller
the percentage difference would be larger.

Chart 4 is calculated from the Harris and Benedict '° standard multiple
prediction tables for normal basal metabolism of adults. It is impossible to
put these tables in the form of logarithnre. They predict the hourly calories
as the sum of two numbers. They are to be used as follows. From the height
in inches and the age in years read by interpolation between the family of
lines a number corresponding to the lines. From the weight in pounds read
on the single line the corresponding abscissa (numbers 20 to 70). The sum
of these two numbers is the expected hourly calory production.

Ex.\MPLE. — Male, age 35, 62 inches tall. 135 pounds.
(35.62) 23

135 38

61 predicted hourly calories.

The normal metabolism data of Benedict leaves a gap from the seventeenth
to the twenty-first year. It is believed that the data of Benedict more nearly
predict the normal than do those of DuBois. The data of DuBois are. how-
ever, widely used. They are included here as Chart 5 and Chart 6. Chart 5
plots the logarithms of reciprocal square meters of body surface. It is the
same family of curves as that of DuBois." Chart 6, due to the data of Aub
and DuBois,'- gives the logarithm of the reciprocal normal calories per square
meter per hour.

Chart 3 is similar to Chart 2. It gives the logarithm of the factor reducing
the observed air volume to 0 C, 760 mm. of mercury at latitude 45 degrees and
sea level together with a correction to 0 C. of the mercury column and brass
scale. It does not correct for aqueous vapor tension as does Chart 2 and is to
be used with portable respiration apparatus. The above remarks on corrections
for latitude and altitude apply equally to this chart. .

Table 3 is a three place logarithm table and Table 4 is a table of anti-
logarithms. The logarithms of the charts consist of a mantissa, the decimal,
which is always positive, and a characteristic, the integer in front of the decimal
point which is negative. The signs of the characteristic are to be considered
in taking the sum of the various logarithms; the characteristic of the sum is
an integer which is the sum of the separate characteristics plus any digit
carried to this place in adding the adjoining columns in the mantissae. The
mantissa only is used in going to the table of antilogarithms and the charac-
teristic plus one represents the number of integral digits to be pointed off in
the number of the table. It fixes the position of the decimal point. In reading
from Table 3 the logarithm of a number such as the hourly volume in liters
assign to the mantissa of the table a characteristic one less than the integral
digits in the number.

Calculation of Calory Production and Basal Metabolic Rate. — The
charts are so arranged that the only calculation required is the per-
formance of a sum.

10. Harris and Benedict: Carnegie Inst. Wash. Pub. No. 279, 1919, p. 253.

11. DuBois, D., and DuBois, E. F.: Arch. Int. Med. 17:865 (lunc) 1916.

12. Aub and DuBois: Arch. Int. Med. 19:831 (lime) 1917.

Chart 4. — The chart is based on the ilaiidard multiple prediction tables of
Harris and Benedict for normal basal metabolism. From the height in inches
and the age in years there is read by interpolation between the family of
parallel lines a first number. The intersection of the horizontal giving weight
in pounds with the single diagonal line gives an abscissa, 20 to 70. The sum
of this and the first number is the normal calory production per hour. The
tables of Harris and Benedict do not extend to heights below si.xty inches.

XEIVCOMER— BASAL METABOLISM DETERMIXATION

761

To determine hourly calory production :

(a) using the Haldane apparatus.

Add the figures obtained by interpolation from Charts 1 and 2
and the logarithm, from Table 3, of the hourly volume in liters, as
read on a meter or spirometer. The antilogarithm, from Table 4,
of this sum is the calory production. The respiratory quotient is read
directly by interpolation between the lines of Chart 1.

(b) using a direct reading oxygen consumption apparatus
(Benedict, etc.).

Add to 1.4617 the logarithm, from Table 3. of the observed
oxygen consumption per minute in c.c. and the figure obtained by
interpolation from Chart 3. Before going to Chart 3 for the volume
correction add to the observed temperature of the bell 1 F. as sug-
gested by Benedict (i/. C). The antilogarithm of the sum is the
calory production. The respiratory quotient is assumed to be 0.82.

^ears 20

Chart 6. — The curves are the logarithms of the reciprocal of the normal
hourly calory production of males and females as given by Aub and DuBois.
DuBois data for males under 15. years is omitted, the data of Table 2 being
preferred. In the IS to 20 year interval DuBois data are higher than the data
of Benedict. Above 20 years comparison cannot be made simply.

The hourly calory production as determined under (a) or (b)
may be compared directly with the normal obtained from Chart 4.
The basal metabolic rate is the quotient of the actual by the predicted.
In the case of children, the basal metabolic rale determination may be
made by simply adding the figure of Table 2 to the sum as obtained
above under (a) or (b). The antilogarithm of this sum is then the
basal metabolic rate. If it is desired to refer to the DuBois standard,
add as a part of the sum the figures from Charts 5 and 6. The
antilogarithm of the sum is the basal metabolic rate.

The results are given with no greater error than that inherent in the data.
Each chart or table is calculated with a slightly greater degree of accuracy
than the data warrant. The charts and tables are such that the basal metabolic
rate is given correctly to within one unit in the second decimal place, that is

762 ARCHWES OF JXTERXAL MEDJCIXE

to within one per cent., and the total calory production correctly to within
one-half unit in the second figure. This amount of precision necessitates the
reading of Chart 1 to within two units in the third decimal place. This latter
may be accomplished by applying a scale between the two CK lines in which
interval the point lies and turning the scale diagonally until the ends of a ten
unit interval on it fall on the two CK lines. The proportionate distance from
the one line to the next of the point to be interpolated is then read off directly
from the scale. The error is not over two per cent, when Chart 1 is not so
taxed to its capacity. Chart 1 can easily be read more closely than the data
can be calculated with a four place logarithm table and more closely than the
Haldane gas buret can be read. Chart 4 can be read correctly to one-half
calory, Chart 5 can be read to one-half unit in the second decimal place, and
the other charts and tables are readily read to three decimals. The logarithm
tables are calculated so as to be correct to one unit in the third decimal place.
Ex.xMPLE 1.— Haldane readings 97.32 and 79.48. Temperature 20 C; barom-
eter 751 mm. Hourly volume 452 liters. Male, age 38, height 6Sy2 inches, weight
ISO pounds.

Chart 1. ri9 R.Q. .83

Chart 2, 1.9525
Log. vol. 2.655

1.7975 antilog. 62.7 calories per hour (actual)
_ Chart 4. 24

Chart 5. h755 41.5

Chart 6. 2.406 777 , •

65.5 calories per hour (expected)

r9585 antilogarithm 0.909 B. M. R. (Du Bois)
E.\.\MPLE 2. — Haldane readings 95.6 and 79.74. Temperature 18 C. ; barom-
eter 740 mm. Hourly volume 234 liters. Girl, aged 11. weight 80 pounds.
Chart 1. T405 R.Q. .83

Chart 2. 1.950
Log. vol. £.369
Table 2. 2.305

0.029 antilogarithm 1.07 7 per cent, above the normal.
Ex.\MPLE 3. — Average reading for contraction of Benedict spirometer bell,
per minute 240 c.c. Average temperature of bell plus 0.5 F. (26 C.) ; barometer
746 mm. Female, aged 32, height 62 inches, weight 125 pounds. ^

Cal. factor T.4617
Log. 240 £.380
Chart 3. 1.9505

1.7922 antilog. 62.0 calorics per hour (actual I
_ Chart 4. 6

Chart 5. L*^"^ ^"

Chart 6. 2.435 ~r~ , • , , . ,.
56 calories per hour (expected*

0.0322 antilogarithm l.OS B.M.R. (DuBois)
SUMMARY

Tables and charts have been prepared by means of wliich the
basal metabolic rate may be calculated by simply adding five numbers.

The charts condense the usual tables into a very small space. They
include one reducing saturated gas volumes to 0 C. dry, 760 mm.
mercury, sea level, 45 degrees latitude with a temperature pressure
correction for the brass 'or glass scale of the liaromcter.

THE NATURE OF THE SO-CALLED
"CAPILLARY PULSE"

ERXST P. BOAS, M.D.

NEW YORK

The "capillary pulse" has been of interest to clinicians ever since
Quincke ^ first called attention to it in 1868. Nothing, however, has
been added to his original description of the condition, nor to his
discussion of the probable mechanism by which it is brought about.
Two observers before Quincke noted the phenomenon in isolated cases,
but failed to appreciate the frequency of its occurrence or its signifi-
cance. Lebert - cites the case of a patient with an aortic aneurysm who
exhibited systolic flushing and diastolic paling of the cheeks. Ascher-
son ^ observed a child 7 years of age with varicella following scarlatina,
in whom the papules and the bases of the vesicles reddened in diastole
and paled in systole. This was very evident for four days, but after
that it was only demonstrable when the skin about the lesion was
stretched. Both authors attributed the phenomenon to a pulsation of
the capillaries. It was Quincke, however, who observed the flushing and
paling of the tissues under the finger nail not only in aortic insufficiency,
but in a variety of other conditions. In patients with incompetent
aortic valves it is most manifest, but it can be observed in many normal
individuals. Quincke calls to mind that ordinarily, because of the
elasticity of the arteries, the blood flows through the capillaries in a
continuous stream ; but that with venous obstruction, or with a
marked lowering of the blood pressure associated with a slow pulse
rate, the capillary flow may become pulsatile. A marked relaxation
of the arterial wall may have the same eflfect. Thus Claude Bernard
explained the pulsation of the capillaries and veins of the submaxillary
gland which he observed on stimulation of the chorda tympani. In a
later paper Quincke '' emphasizes that a great difiference between the
systolic and diastolic pressures is essential for the visualization of the
capillary pulse. He observed the phenomenon in anemic individuals,
and in those with low blood pressure, an overactive heart, and a pulse

* From the Medical Division of the Montefiore Hospital for Chronic Diseases.

1. Quincke, H.: Beobachtungen ueber Capillar und Venenpuls, Berl. kiln.
Wchnschr. 5:357, 1868.

2. Lebert: Handb. der praktischen Mcdicin, Ed. 3, 1:746, 1863.

3. Ascherson : Variola Versicolor. Medizin. Ztg. d. Ver. f. Heilk. in Preussen,
18,-?4.

4. Quincke. H.: Ueber Capillarpuls und Ccntripetalcn Vcncnpuls, Bcrl. klin.
Wchnschr. 27:263. 1890.

764 ARCHIVES OF IXTERXAL MEDICIXE

of the collapsing type, as well as in those with a leaking aortic valve.
In his first paper Quincke described the pulsation under the finger nail,
both with and without the application of gentle pressure. Subsequently,
however, he noted it on the mucous membrane of the lip, when com-
pressed with a glass slide, and on the skin of the forehead after it
had been rubbed with a blunt object. When a patient exhibiting this
phenomenon presents a skin lesion, such as erisypelas or urticaria, the
pulsation becomes very evident.

Quincke's observations and conclusions have been generally con-
firmed and accepted, and we find in most textbooks of medicine, as
well as of physiology, a presentation of the views first set forth by
him fifty-three years ago. Lombard ^ in 1912 described a method by
which the human capillaries can be studied directly under the micro-
scope, but in the United States little use has been made of his observa-
tions, save by Danzer and Hooker," who devised a method of measuring
the capillary pressure, based on this principle. If a drop of glycerin
or castor oil be applied to the skin at the base of the finger nail, and
this area be then studied through the microscope under direct illumi-
nation, with a magnification of from 40 to 80 diameters, the capillary
loops are beautifully visualized. Not only can the architecture of the
capillary bed be studied, but the blood flow can be observed almost as
well as in the classical demonstration of the capillary circulation in
the web of the frog's foot. For details of technic, Danzer and Hooker's
article, as well as of Weiss' ' studies should be consulted.

While engaged in a study of the capillary morphology and blood
pressure in a series of many dift'erent types of cases, I had the oppor-
tunity to observe the capillaries in eleven patients who exhibited a well
marked clinical capillary pulsation under the finger nail. The capillaries
in these cases were studied most assiduously under all kinds of con-
ditions, and in no instance was a pulsatory stream of the blood in the
capillaries seen.

MclJiod. — .-X. description of the method employed will he in i)lace
A drop of clear castor oil is placed on the dorsum of a finger just
below the nail bed, and the finger is then placed on the finger rest of
Danzer and Hooker's microcapillary tonometer, which stands on the
stage of the microscope at heart level. Light is thrown on the area

5. Lombard, W. P.: The Blood Pressure in the .-Vrterioles, Capillaries and
Small Veins of the Human Skin, Am. J. Physiol. 29:335, 1912.

6. Danzer, C. S., and Hooker, D. R. : Determination of the Capillary Blood
Pressure in Man with the Micro-capillary Tonometer, Am. J. Physiol. 52:
136, 1920.

7. Weiss, E., and Dieter, W. : Die Stronunig in den Kapillaren und ilirc
Beziehung zur Gefassfunktion, Zentralbl. f. Hcrz. u. Gefasskrankh. 12:295, 1920.

BOAS— CAPILLARY PULSE 765

to be observed by an electric bulb whose rays are focused by a con-
denser. A magnification of eighty diameters is employed. The capil-
laries are thus clearly brought into view and can be studied at leisure.
Patients with a clinical capillary pulse are, however, difficult to study,
because the finger moves with each pulse beat and the capillaries are
thus thrown out of focus. This results in a very deceptive microscopic
picture, for the finger movements rhythmically alter the focus of the
microscopic field under observation. Thus with each pulse beat the
capillaries become indistinct and may even disappear from view, to
reappear immediately thereafter with their original clearness. This
movement in and out of the focal field can readily be mistaken for a
pulsation of the capillaries themselves. However, a close study, which,
to be sure, is somewhat trying on the eyes, will convince the observer
that the blood stream through the capillaries is at all times continuous
and never pulsatile.

When the capillary blood stream is studied in this manner, it will
be found that in most instances the flow is so rapid and steady that it
can hardly be visualized. Each capillary is of a constant and uniform
caliber. However, in individual capillaries the flow may be slow and
almost halted at times but soon resumes its rapid streaming. Some
individuals have a slower flow than others, and some a more rapid flow.
With hypertension the velocity of the capillary blood appears to be
increased; in arteriosclerosis it is decreased. But in spite of these
individual variations, a capillary pulse was never observed. There was
no intermittency of the circulation in the capillaries, nor was there any
systolic lateral displacement of the capillaries.

Because of the fact that clinically the "capillary pulse" is usually
observed best under slight pressure of the tissues, such pressure was
exerted on the area studied by means of the Danzer-Hooker instru-
ment. The pressure was elevated until the flow through the capillaries
ceased, and was then gradually released to zero. At no pressure, not
even with the first reappearance of the flow, was there the slightest
departure from the normal continuous stream of blood. Both Weiss '
and Jurgensen * have described pulsation of the capillaries observed
through the microscope, but I am compelled to disagree with their
findings. It is probable that the pulsation of the finger, discussed
above, led them astray. I have seen this simulated capillary pulsation
when the capillary flow was at a standstill because of the high pressure
in the air chamber of the tonometer. Such an observation allows of
no two interpretations and shows clearly with what care the studies
must be made to avoid error.

766

ARCHIVES OF IXTERXAL .VEDICIXE

The accompanying table gives a brief summary of the cases studied,
together with the capillary pressure found in each instance. For
reasons which will be detailed in a subsequent article, I have not aver-
aged the pressures read in the different capillaries as advocated by
Danzer and Hooker, but prefer to record them as individual readings.
It seems quite certain that the variations of pressure observed in the
same individual are of significance, and are not due to inaccuracy of
measurement.

Patients Exhibiting Clinical Capillary Pulse but no Microscopic
Pulsatory Capillary Flow

case

Age

Sex

Diagnosis

Blood
Pres-
sure

Capillary
Pressure,
Mm.Hg

Temp.,

Appearance ol
Capillaries

1

21

Male

Aortic insufflcitncy,
rheumatic

140/ 50

10
10

■■ \

.Arrangement regular,
some looped; sub-
papillary plexus visi-

2

23

Male

Aortic insufficiency,
rheumatic

115/ 60

20 20
18 25
22 23
26 15

78

Capillarics long, few
loops

S

28

Male

Aortic insufficiency,
rheumatic

140/ 50

20 18.5

10 13.5
17 10

11 8

65

Capillaries numerous,
many rows, lew tor-
tuous; subpapillary
plexus visible

*

18

Female

-Aortic insufficiency,
rheumatic

190/ 0

17 17.5
17

67.5

Capillaries numerous,
not tortuous; sub-
papiUary plexus visl-

5

20

Male

Aortic insufficiency,

mitral stenosis,

rheumatic

150/ 20

8 5
7 12
12 7
10 16.5
15

70

Capillaries very long

«

12

Male

Aortic insufficiency,
rheumatic

110/ 40

10 16
9 10.5

»

Capillaries very con-
voluted and • irreg-
ular

7

8

Male

Aortic insufficiency,

mitral stenosis,

rheumatic

100/ 40

5 4
5 3

70

Capillaries normal

8

76

Male

Hypertension right
hemiplegia

180/ 60

25 24
21 25
27 21
20

68

Capillaries long, con-
voluted: extensive
subpapillary plexus

9

55

Male

Hypertension

205/110

17 31.5
12 19.5
23 22,5

18 18
17 26.5
34 38
29 28

65

CapiUaries long,
many convoluted;
some giant capil-
laries

10

57

Female

Hypertension

145/ 75

25 20
27 33
18 H
80 81

75

Capillaries very con-
voluted; several
rows; flow rapid

11

«3

Female

Hypertension,
diabetes

210/ 85

26 13

27 15
22 21
85 23
10 11.5
17 14.6

88

Longer and more tor-
tuous than normal;
subpapillary plexus
prominent

BOAS— CAPILLARY PULSE 767

Since, according to these observations the current A'iews on the
nature of the "capillary pulse" are erroneous, we must seek a new
explanation of the phenomenon. At first thought it would seem that if
the capillaries are not concerned in the production of this pulsation,
it must be the minute arterioles in the subpapillary plexus of the skin
which bring it about. This may be so in many instances, but Quincke's
observations on the centripetal venous pulse, published in the same
articles as those on the capillary pulse, suggest another possible inter-
pretation. In some patients who exhibited clinical capillary pulsation
he noted a post systolic centripetal pulsation of the veins on the dorsum
of the hand. He believed it to be due to the projiagation of the arterial
pulse wave through the capillaries to the veins. In his second paper,
however, he states that the capillary pulse is uncommonly found asso-
ciated with the venous pulse. It is not quite clear how this comes
about. Jiirgensen * ofTers a possible explanation. He reviews the
work of Hoyer,^ Grosser^" and Schumacher," who demonstrated
direct anastomoses between the arterioles and venules in the sub-
papillary plexuses, to explain some of the phenomena which he has
observed in his studies of the capillaries. It is possible that the venous
pulse, when it occurs, is caused by the transmission of the pulse wave
through these subpapillary anastomoses, and that the clinical capillary
pulse may be due to a pulsation of the subpapillary venules as well as
arterioles. ' Additional evidence that the capillaries are not necessarily
associated with the color of the skin is found in Weiss' '- observation
that on the skin of the cheek the capillaries are relatively scanty while-
the subpapillary venous plexuses are especially well developed, thus
playing a dominant part in the production of the color of the skin of
the cheek.

CONCLUSIONS

The so-called "capillary pulse" is not a manifestation of a pulsation
of the capillaries, but is due to an exaggerated pulsation of the
arterioles and possibly of the venules of the subpapillary plexus of
the skin. In view of this fact it would be well to discard the term
"capillary pulsation" and to speak of the systolic flushing of the skin.

8. Jiirgen.sen; E. : Microkapillarbeobachtungen, Deutsch. Arch. f. klin. Med.
132:204, 1920.

9. Hdver, H. : Ueber unmittelbarc Einmiindung kleinster Arterien in Gefasse
venosen Charakters, Arch f. mikroskop. Anat. 13:603, 1877.

10. Grosser, O. : Ueber arteriovenose .'\nastomosen an den Extremitatenden
beim Menschen und den krallentragenden Saugethieren, Arch f. mikroskop. Anat.
60:191, 1502.

11. V. Schumacher, S. : Ueber das glomus coccygeum des Menschen. etc.,
-Xrch. f. mikroskop. .'Vnat. 71:58, 1908.

12. Weiss, E., and Holland. M. : Zur Morphologic und Topographie der-
Hamkapillaren, Ztschr. f. exper. Path. u. Therap. 22:108, 1921.

THE ETIOLOGY AND DEVELOPMENT OF
GLOMERULONEPHRITIS *

E. T. BELL, M.D.

AND

T. B. HARTZELL, M.D.

MIXXF..\POLIS

There are four well established types of renal disease that must be
considered in a discussion of nephritis. A brief explanation of each of
these will be given in order to establish the limitations of the group
under discussion.

1. Pyelonephritis. — This is an acute or chronic exudative inflamma-
tion distributed in patches throughout the kidneys, and extending from
the cortex throughout the pyramids into the pelvis. It is caused by
bacteria, usually staphylococci or colon bacilli. In most instances the
bacteria are carried to the kidney by the blood; but in cases of obstruc-
tion of the lower urinary tract they may enter from the urine. In
the earlier stages of a hematogenous infection, before there has been
extension to the pelvis, the lesions are spoken of as abscesses. Exten-
sion of the infection to the capsule may produce perinephritis or
perirenal abscess. Infection of a dilated pelvis causes pyonephrosis.
The disease is frequently unilateral and acute cases are sometimes mis-
taken for appendicitis or other acute abdominal conditions. Cases
vary in intensity from mild to severe. The treatment of severe uni-
lateral cases is usually surgical. Bilateral pyelonephritis may result in
renal insufficiency but there is seldom any confusion clinically with
Bright's disease.

Acute interstitial nephritis is related to this group in that it is an
exudative inflammation of the interstitial tissues. The kidneys show
areas of cortex infiltrated with lymphocytes of intermediate size and
plasma cells. Apparently there is never enough of the kidney involved
to produce renal insufficiency. There is evidence in one of our cases
(Case 51) that the exudate remains in the kidney indefinitely and gives
rise to areas of cortical atrophy; but no chronic nephritis of this type
is known. Acute interstitial nephritis occurs frequently during scarlet
fever, and rarely in other infections, such as diphtheria and congenital
syphilis. It is seldom of much importance clinically, being over-
shadowed by the associated disease.

The spontaneous chronic nephritis of laboratory animals is more
closely related to pyelonephritis than to any other form of human renal

' From the Department of Pathology, University of Minnesota.
'■ Aided by grant from the National Dental Association.

BELL-HARTZELL—GLOMERULOXEPHRITIS 769

disease. It is characterized by lymphocytic exudates in the renal paren-
chyma, which ultimately cause cortical atrophy to such an extent that
the kidneys are shrunken and their surfaces covered with small
depressions. The pelvis is not extensively involved, as a rule. Death
from renal insufficiency is rare, and the majority of cases result in
healing. This is the form of nephritis usually obtained in animals by
experimental procedures. There is a gross resemblance between the
kidneys of chronic Bright's disease and those of this animal infection;
but clinically, histologically and in manner of development there is no
similarity.^ There is no known animal disease corresponding to chronic
Bright's disease.

2. Nephrosis. — This term is applied to renal lesions of a purely
degenerative character in contrast to nephritis in which the phenomena
of reaction (exudation, proliferation) have appeared. This group is
not sharply separable from glomerulonephritis since cases of degenera-
tion occur in which it is very difficult to determine whether there are
any reactive changes in the glomeruli. Again, there are degener-
ative changes of greater or less degree in practically all cases
of glomerulonephritis, and in a large percentage of arteriosclerotic
kidneys. It is, however, very desirable to limit the term nephrosis to
degenerative lesions in which there is no pronounced involvement of
the glomeruli or blood vessels. ,

Nephrosis is by far the commonest form of renal disease seen at
necropsy and it is the most frequent cause of albuminuria. It is found
at necropsy in practically all severe infectious diseases and infections,
in obscure toxemias such as the toxemia of pregnancy, in chemical
poisoning (mercury, arsenic, phosphorus, and many other substances),
in severe jaundice, and in severe cardiac decompensation. It is pre-
sumably present in life in the above named conditions when albumin
or casts are found in the urine. When found at necropsy it is proof
of some form of toxemia.

The amyloid kidney, although usually classified as a nephrosis, is
better understood as a special form of glomerulonephritis.

Clinically, nephrosis is usually definitely secondary to some associ-
ated disease, and the differentiation from Bright's disease under these
circumstances presents no difficulties ; but in cases of obscure etiology
the distinction is not made so readily. There are no cases in our
series of chronic nephrosis such as is described by Volhard and Fahr -
as "genuine" nephrosis. It seems that a case of primary chronic renal
disease should rarely, if ever, be regarded as a nephrosis.

1. Bell and Hartzell : Spontaneous Nephritis in the Rabbit, J. Infect. Dis. 24:
628, 1919.

2. Volhard and Fahr: Die Brightsche Xierenkrankheit, Berlin,- 1914.

770 ARCHIl'ES OF IXTERXAL MEDICIXE

The clinical findings attributable to the kidney in nephrosis are
usuall}' mild ; but renal insufficiency may develop in severe cases. There
is never hypertension or cardiac hypertrophy. Severe nephroses are
usually rapidly fatal.

In mild nephroses at necropsy the kidneys show cloudy swelling.
In severe cases there is cloudy swelling, fatty degeneration and some-
times necrosis of tubules and glomeruli.

3. Arteriosclerosis of the Kidneys. — Two forms of this disease
may be distinguished on clinical and anatomic grounds.

[a) The Senile Type (Arteriosclerotic atrophy) : In advanced
age and especially in association with generalized senile arteriosclerosis,
one often finds small kidneys with adherent capsules and rough gran-
ular or pitted surfaces. This appearance is due to an irregular atrophy
of the more superficial parts of the cortex which is caused by narrow-
ing of some of the arteries. Usually the larger branches are the ones
chiefly aft'ected and the disease is never restricted to the arterioles.
The senile kidney is seldom of any clinical importance since the amount
of cortex destroyed is relatively small; but in occasional instances
the atrophy is so extensive that renal insufficiency develops. It has
not been determined how often arteriosclerotic" atrophy gives rise to
the clinical picture of chronic Bright's disease.

(b) The Hypertension Type: In many cases of chronic hyper-
tension normal Tvidneys are found at necropsy and in some a glomerulo-
nephritis is found ; but a large percentage of cases are associated with
disease of the renal arteries. When no disturbance of renal function
is demonstrable we speak of essential hypertension, but when there
is evidence of serious renal injury w-e consider the condition chronic
Bright's disease. The kidneys from cases of essential hypertension
usually show hyaline degeneration of some of the afferent glomerular
arteries and often there is also disease of medium sized and small
arteries, and there are gradual transitions between the slight involve-
ment of the arteries in these cases and the extensive involvement in
chronic Bright's disease of the vascular type. The majority of patients
with essential hypertension die of cardiac or cerebral complications
without developing serious renal involvement, but some cases extensive
destruction of renal tissue occurs and they are then regarded as being
cases of chronic Bright's disease of the arteriosclerotic type. There
are about twenty-six examples of this form of Bright's disease in our
series of 3,300 necropsies. We expect to discuss this subject fully
in a subsequent report.

4. Glomerulonephritis. — This group includes renal disease in which
the structural changes are due almost entirely to primary inflammatory
and degenerative changes in the glomeruli. It includes all acute and
subacute and a majority of chronic cases of Bright's disease. ■

BELL-HARTZELL—GLOMERULOSEPHRITIS 77\

The older terms "chronic parenchymatous" and "chronic inter-
stitial" correspond to stages or degrees of severity of glomerulo-
nephritis. They do not designate any important feature of the disease
and a large number of cases are intermediate in type, i. e.. neither typical
parenchymatous nor typical interstitial. This terminology has another
element of confusion in that some observers consider the arteriosclerotic
kidney as chronic interstitial nephritis. The literature of nephritis
would be clearer if these older terms were discarded.

Chronic glomerulonephritis is not sharply separable from the
arteriosclerotic kidney since a few cases of the former show some
disease of the renal arteries; but these borderline cases are not very
numerous and we see no justification for the view that the two diseases
are indistinguishable (Moschcowitz ^j. Certainly the great majority of
cases are anatomically distinct, although they may be indistinguishable
clinically. It may be, as Ophiils * believes, that the same toxin attacks
the arteries in one case and the glomeruli in another and that the
differences are really only anatomic ; but it seems better to adhere to ,
anatomic distinctions until we know more about the etiology of
arteriosclerosis.

M.\TERI.\L

Microscopic sections from the kidneys of about 3,300 consecutive
necropsies have been examined. Small pieces of kidney from nearly
all of these had been preserved so that it was possible to make serial
sections when desirable. The clinical history and gross necropsy find-
ings were always considered, but the final diagnosis was usually made
on the microscopic appearances. All the subacute and chronic cases
had been recognized clinically or at necropsy, but many of the acute
cases had been overlooked. Sixty-nine cases of glomerulonephritis were
identified. In a number of the acute cases nephritis was not the main
cause- of death, and these kidneys furnish abundant illustrations of the
early stages of the disease.

We have arranged our cases somewhat arbitrarily into acute, sub-
acute and chronic groups. There are striking clinical and pathologic
differences between typical examples of each group, but there are many
intermediate forms and one may easily become convinced from the
study of a large series that there is a fundamental relationship between
the different forms of glomerulonephritis. Senator = called attention to
this point many years ago. As will be brought out later, one of the

3. Moschcowitz. E. : Clinical and Anatomic Relations in Chronic Xephritis,
Arch. Int. Med. 26:2S9 (Atfg.) 1920.

4. Ophiils, Wm. : Arteriosclerosis and Cardiovascular Disease, Stanford
Univ. Pub.. Med. Sc. 1:1. 1921.

5. Senator, H. : Xothnagel's Encyclopedia of Practical Medicine, 1903,
Am. Ed., p. 180.

in ARCHIVES OF IXTERXAL MEDICINE

arguments for the infectious origin of chronic glomerulonephritis is
that it is linked to the acute form by numerous intermediary cases.

ACUTE GLOMERULONEPHRITIS

There are thirty-two acute cases. Brief protocols of each will be
given. Cases 1 to 9 are mild cases in early stages in which death was
clue to extrarenal causes. In Cases 1 and 2 onl}' a minority of the
glomeruli are involved.

C.\SE 1 (A-16-182). — Female, aged 33 years. Case of advanced chronic
aortic and mitral endocarditis with fresh thrombi on the thicl<ened leaflets.
Kidneys weighed 260 gm. ; no gross changes. Some of the glomeruli showed
changes in a few of their lobules. These changes consisted in swelling of the
endothelium, partial closure of capillaries, and accumulation of polymorpho-
nuclears in the capillaries. The majority of the glomeruli are entirely normal.
This represents an early glomerulitis of very limited e.xtent.

C.^SE 2 (A-19-276). — Male, aged 16 years. Acute endocarditis. During the
last week of life, blood culture on two occasions gave hemolytic streptococci.
Hematuria. Leukocytes, from 20,000 to 42,000. Necropsy: Heart weighed 620
gm. ; many large thrombi on mitral and aortic leaflets. Infarcts in kidneys and
spleen. Kidneys not enlarged. Many petechial hemorrhages. Some glomeruli
showed swelling of endothelium with partial closure of capillaries. Majority
were normal.

C.\SE 3 (A-19-S8). — Male, aged 39 years. Chronic mitral endocarditis with
fresh thrombi on the stiffened leaflets. Leukocytes, 11,200. Heart weighed 585
gm. Kidneys not enlarged ; all glomeruli much enlarged. There arc a number
of polymorphonuclears in the capillaries. The endothelial swelling is not
sufficient to close any of the capillaries. There is moderate injury of the
tubules but no atrophy.

C.\SE 4 (.\-20-94). — Male, aged 61 years. Had a suppurative infection of
the right carpus for two months before death. Erysipelas of face and neck the
last five days of life. No examination of urine. Necropsy: Suppuration of
carpus with partial destruction of the os magnum. Erosion of the cartilaginous
surfaces. Spleen weighed 550 gm. Hemolytic streptococcus in pure culture from
the spleen. Large numbers of gram-positive cocci demonstrable in sections
of the spleen. Kidneys weighed 370 gm. Cloudy cortices. Glomeruli are not
enlarged, but their capillaries contain many polymorphonuclear leukocytes.

Case 5 (A-20-1 18) .—Female, aged IS years. Normal labor February 11.
Fever began four days later. Continuous fever and leukocytosis. Streptococcus
from blood culture, February 28. Death, March 13, 1920. Necropsy: 1,500 c.c.
pus in right pleural cavity. Large thrombus on tricuspid valve. Abscess of
right lung. Spleen weighed 260 gm. Kidneys weighed 295 gm. Cloudy cortices.
Glomeruli not enlarged. No swelling of endothelium. A number of polymorpho-
nuclears in some of the glomerular capillaries.

C.\SE 6 (.'^-20-220).— Male, aged 55 years. Death from chronic myocardial
degeneration and chronic alcoholism ; no clinical history ; no edema. A small
amount of fluid found in the pleural cavities. Heart weighed 520 gm.; spleen,
335 gm.; kidneys, 397 gm. Glomeruli are all enlarged. There is moderate
swelling of the endothelial cells, with partial closure of some capillaries. Many
polymorphonuclears in the capillaries.

Case 7 (A-21-144).— Male, aged 20 years. Measles in Noveml)er, 1919.
complicated by bilateral suppurative otitis media. Discharge from the ears for
several weeks. Present illness began about four weeks before death with septic
sore throat, complicated by acute bilateral suppurative otitis media. Both

BELL-HARTZELL—GLOMERULOXEPHRITIS Hi

tympani were incised. Discharge from ears continued until March 20. Symptoms
of meningitis appeared March 20. Lumbar puncture, March 21, gave a purulent
fluid. The urine showed a faint trace of albumin at times. Death March 22.
Necropsy : purulent meningitis ; streptococci in smears. Bilateral suppurative
otitis media with mastoiditis. Kidneys were not enlarged but showed cloudy
cortices. Glomeruli not enlarged. There is no notable swelling of the endothe-
lium but the capillaries contain large numbers of polymorphonuclears.

Case 8 (A-17-114). — Female, aged 44 years. Death from lobar pneumonia.
No edema. No urinalysis. May 19, 10,900 leukocytes. Blood pressure 115/88.
Death ^lay 23. Necropsy : lobar pneumonia. Heart weighed 310 gm. ; kidneys,
310 gm. Pale cloudy cortices. All glomeruli slightly enlarged. Large numbers
of polymorhponuclears in the capillaries. No swelling of the endothelium.

C.^SF. 9 (A-13-24). — Male, aged 55 years. Had nosebleed at frequent intervals
for two weeks before death. No edema. Urine not examined. Necropsy:
pneumococcus bacteremia; acute bronchopneumonia (small areas); localized
suppuration in the pharynx which extended deep into the pharyngeal tissues.
Kidneys weighed 400 gm. and contained a large number of small peripheral
infarcts which were due to thromboses of the small arteries. Most of the
glomeruli are enlarged and the capillaries are partly occluded by the swollen
endothelial cells. \\\ occasional polymorphonuclear leukocyte is seen in the
glomerular capillaries.

Cases 10 to 21 are examples of fairly severe glomerulonephritis in
which the renal condition was obscured by the associated disease.

C.\SE 10 (A-13-190). — Female, aged 35 years. History of many sore throats.
Inflammatory rheumatism at 15. Has deformities of joints and swelling of
feet. November 7. blood pressure 126/110. Urine: faint trace of albumin,
November 6 and November 28. Leukocytes, December 6, 5,500 ; December 26,
26,600; 95 per cent, polymorphonuclears. Streptococcus from blood culture.
Clinical diagnosis : cardiac hypertrophy, endocarditis, mitral insufficiency,
terminal bacteremia. Death, December 27. Necropsy: edema of ankles. Heart
weighed 525 gm. ; mitral stenosis ; fresh thrombi on mitral valve and mural
thrombosis of left ventricle. Kidneys not enlarged, but cloudy. Glomeruli are
all greatly enlarged. A large percentage of the capillaries are partly or
completely occluded by the swollen endothelium. Many capillaries contain
numerous polymorphonuclear leukocytes.

Case 11 (A-14-49). — Male, aged 25 years. Has had arthritis at intervals
for many years. Present illness began about Dec. 1, 1913, with dyspnea and
swelling of the legs and face. .Admitted to Hospital, Feb. 13, 1914. Marked
dyspnea. Temperature, 102 F. on admission ; later not over 100 F. Urine :
traces of albumin during February, severe albuminuria in March. Feb. 14,
13.600 leukocytes. Clinical diagnosis: cardiac hypertrophy and endocarditis.
Death. .April 4. 1914. Necropsy: general anasarca, ascites, hydrothorax. Heart
weighed 575 gm. ; chronic mitral endocarditis; many large thrombi on aortic
leaflets. Spleen, greatly enlarged. Kidneys enlarged and cloudy. Glomeruli
are all enlarged and show partial closure of the capillary lumina, due to
swelling of the endothelium. There are very few polymorphonuclears in the
glomeruli. No tubular atrophy.

C.\SE 12 {A-14-255).— Male, aged 39 years. Patient was taken ill witli a
cough and fever after working three days in a wet trench. Diagnosis of
pneumonia, made by a physician four days later (Nov. 23, 1914). Admitted to
hospital, December 1. Leukocytes, 20,000. Urine: moderate albuminuria,
numerous hyalin and granular casts. No edema. Clinical diagnosis : lobar
pneumonia. Death, December 12. Necropsy: Lobar pneumonia; large thrombi
on aortic leaflets. Kidneys weighed 405 gm. The glomeruli are greatly enlarged

774 ARCHH'ES OF IXTERXAL MEDICIXE

and the capillary lumina are partially occluded by swollen endothelial cells.
There are many polymorphonuclear and some large mononuclear cells in the
capillaries. No atrophy of the tubules.

Case 13 (A-13-16S). — Male, aged 33 years. Had acute articular rheumatism
a few years ago. Four months ago he first noticed shortness of breath and
swelling of feet. Symptoms became more intense. Purpuric rash appeared on
lower extremities on October 1. Admitted to hospital, October 4. Leukocytes,
15,800. Urine: albumin, numerous hyalin and granular casts. Death, October
21. Necropsy: moderate anasarca, ascites. Heart weighed 640 gm. ; chronic
mitral endocarditis with many fresh thrombi on mitral and aortic leaflets.
Spleen weighed 595 gm. ; kidneys, 460 gm. Glomeruli moderately enlarged.
Partial closure of capillaries. A number of polymorphonuclears in the capil-
laries. No tubular atrophy.

C.\SE 14 (A-lS-97).— Female, aged 29 years. Illness of nine weeks' duration.
Continuous high fever. Pus in urine. No edema. Leukocytes, 13,500. Necropsy :
acute vegetative mitral and aortic endocarditis. Infarction of spleen. Kidneys
cloudy but not enlarged. Glomeruli are all moderately enlarged. The capil-
laries are partially closed and contain large numbers of polymorphonuclears.

Case 15 (A-15-230).— Male, aged 34 years. Acute symptoms began about
April 1, 1915. Cough, night sweats, pulmonary hemorrhage. Temperature 97
to 102. Tubercle bacilli in sputum. No edema. Death, July 16, 1915. Necropsy :
ascites (700 c.c). Pulmonary tuberculosis with cavities. Intestinal ulcers.
Heart weighed 320 gm. ; kidneys, 490 gm. Glomeruli all moderately enlarged.
Increased number of nuclei. Partial closure of many capillaries by swollen
endothelium ; complete closure of some. A number of polymorphonuclears in
some glomeruli. No tubular atrophy (Fig. 1).

C.^sE 16 (A-15-323).— Male, -aged 49 years. First hospital admission in
June. 1914. Discharged. Readmitted, .^lug. 3. 1915. Recurrent attacks of
arthritis. Loss of weight and strength. Dyspnea. Edema. Enlarged heart.
Dilated aortic arch. Right pleural cavity aspirated several times. Blood
Wassermann, positive. Urine, August 5 : albumin and casts. Blood pressure,
.\ugust 11, 160/110. Death, Oct. 18, 1915. Necropsy: ascites (200 c.c.) hydro-
thora.x. Enlarged heart : advanced mitral stenosis with ulceration ; aneurysm
of arch of aorta. Kidneys weighed 320 gm. A number of glomeruli appear as
abscesses because of the enormous number of polymorphonuclears present
( Fig. 2) ; in other glomeruli there are only a few polymorphonuclears. A fairly
large proportion of the glomeruli show epithelial crescents (the extracapillary
type of glomerulrtis). There is no appreciable swelling of the endothelium.
This case represents a blending of the pyelonephritis group with the extra-
capillary form of glomerulonephritis. There are large numbers of polymorpho-
nuclears in the tubules. No tubular atrophy.

Case 17 (A-16-48). — Male, aged 58 years. Diabetic gangrene of foot. No
edema. Blood pressure. February 8, 140/110. Urine, February 9: trace of
albumin and a small amount of pus. February 5, 13,500 leukocytes; 91 per cent,
polymorphonuclears. Death, February 10. Necropsy: ascites (200 c.c); right
hydrothorax (400 c.c). Heart weighed 370 gm. Streptococcus pyrogenes
secured from heart's blood. Kidneys weighed 573 gm. Cloudy cortices. Large
numbers of polymorphonuclears in the glomeruli, tubules and interstitial tissue
(Fig. 3). Glomeruli are all enlar,ged. Marked increase of nuclei and partial
closure of capillaries by swollen endothelium. No tubular atrophy.

C.\SE 18 (.^-16-165). — Male, aged 35 years. Became intoxicated and was
injured in an automobile accident April 20. Superficial injuries. No broken
bones. Admitted to hospital, April 22. Temperature around 101 F. Excessive
hematuria, vomiting, hiccoughing. Leukocytes, 15,000; 82 per cent, poly-
morphonuclears. No edema. Death, April 26. Necropsy: heart weighed 345
gm. ; kidneys, 415 gm. ; cloudy cortices. No focus of infection found. Blood

BELL-HAKTZELL—CLOMEKCLOXEPHRITIS

775

culture sterile. Glomeruli are all moderately enlarged and show great increase
of nuclei. Capillaries are partially closed by swollen endothelium. Many
polymorphonuclears in the capillaries. No atrophy of tubules.

C.\SE 19 (A-17-202).— JIale. aged 25 years. Acute arthritis in 1908 and
again in 1909. Smallpox in 1910. History of cardiac disease for the past
six years. Frequent appearance of petechial hemorrhages in the skin since
February, 1917. These would disappear, then reappear after a variable interval.
Admitted to hospital Aug. 20, 1917. On admission there was marked cardiac
hypertrophy, enlargement of spleen and edema. Urine : abundant albumin,
casts found at one examination. Leukocytes, August 21, 4,700; 70 per cent,
polymorphonuclears. Septic temperature. Phenolsulphonephthalein, August 23,
65 per cent. ; September 14, 45 per cent. One blood culture sterile. Blood
pressure, October 3. 140/?0. Death, Oct. 4, 1917. Xecropsy : anasarca, ascites,
hydrothorax. .Adlurent pL^ricardium. Hesrt weighed 6(X) gm. : old ulcerative

Fig. l.-Cas
magnification.

type. Low

mitral lesion with many large fresh thrombi. Spleen weighed 1400 gm. ;
kidneys, 515 gm. .\11 glomeruli are greatly enlarged with a notable increase
of nuclei. The capillaries -are largely closed by the swollen endothelium.
There- are only a few polymorphonuclears in the glomeruli. Some of the
tubules show an early stage of atrophy. There are no embolic lesions in the
glomeruli although one would expect this lesion with the type of involvement
of the heart and spleen.

C.xsE 20 (A- 18-9) .—Female, aged 3 mnmh>. .Marked jaundice developed
October 2. .'\dmitted to hospital, Oct. 4, 1917. .Marked jaundice and exoph-
thalmos. Enlargement of liver and spleen. iMnaciation. Xo edema. Wasser-
mann positive. Hemoglobin, 35 per cent. Erythrocytes, 2,500,000. Bile and
pus in the urine. Treated for congenital syphilis. .-Attack of lobar pneumonia,
Jan. 3, 1918. Partial recovery from the pneumonia, but fever and listlessness
persisted. Leukocytes, Jan. 16, 1918, 14,800. Death, January 17. Necropsy:
100 c.c. of thick pus in peritoneal cavity, and fibrinopurulent exudate over

776

ARCHirES OF IXTERXAL MEDICIXE

both lungs. Kidneys weighed 45 gni. ; pale, cloudy cortices mottled with
hemorrhages. Pneumococci in smears from pus from serous cavities and in
pure culture from the blood. A large percentage of the glomeruli show
infarction due to thrombosis of the afferent arteriole (Fig. 4). The majority
of the glomeruli show no special abnormality. There is no swelling of the
endothelium, and there are very few polymorphonuclears in the capillaries.
There is severe tubular degeneration.

C.\SE 21 CA-18-122).— Male, aged 30 years. Dated his present troulde from
March 1, 1918. Abundant albumin in urine. Xo other clinical data. Death,
June 11, 1918. Necropsy: general anasarca, hydrothorax. hydropericardium.
Heart weighed 565 gm. Numerous large vegetations on mitral and aortic
\alves. Spleen weighed 680 gm. ; contained several infarcts. Kidneys weighed
530 gm. Cortices thickened, clouded and yellowish. Chronic passive con-
gestion of viscera. The glomeruli are nearly all moderately enlarged and

show an increased number of nuclei. There is swelling of the endothelium and
narrowing of the capillaries, liut only a small per cent, of the capillaries are
completely closed. There are many mononuclear leukocytes in the capillary
lumina. Some of these may 1)e endothelial cells that have separated off from
the walls. There is moderate injury of the tubules but no atrophy.

Cases 22 tn ,i2 may be ccmsidered fairly typical eliiiical examples of
acute glomerulonephritis.

22. (.^-13-140).— Female, aged 12 years, .\fter a severe colil, about Tune 7,
1913. patient developed fever with swelling of the face and feet. She was
confined to her bed for two weeks at this time. On the third day of the
illness a diagnosis of pneumonia was made. About June 28, a severe general-
ized edema appeared and persisted until death. Admitted to hospital, Aug. 9,
1913. Empyema was recognized, and was drained by rib resection .August IS.
Urine was scanty and contained many granular, waxy and hyalin casts, and
large quantities of albumin. Phenolsulphonephthalein, ^9 per cent, .'\ugust 25 ;

BELL-HARTZELL—GLOMERULO\EPHRITIS

777

36 per cent., September 1. Death. September 24. Duration of nephritis about
three months. Xecropsy : Purulent peritonitis. Pure growth of pneumococci
from peritoneal exudate. Empyema. Kidneys cloudy and swollen. Severe
glomerulitis ; capillaries largely closed by the swollen endothelial cells (Fig. 5).
Polymorphonuclears rare in most glomeruli ; numerous in a few. A few
glomeruli show beginning hyaline degeneration with early tubular atrophy.

C.\SE 23 (A-13-150).— Male, aged 35 years. Illness began three and one-
half months before death with frequent chills, fever and vomiting. Became
better and went back to work. About one month later he de\eloped a cough
with weakness, loss of appetite and pain in the right chest. Admitted to hos-
pital Sept. 18. 1913. Temperature of septic type. Slight edema of both legs.
Abundant albumin and many casts in the urine. Leukocytes. 18,000. Death,
Oct. 4, 1913. Necropsy : double empyema. Kidneys swollen and cloudy.
Glomeruli are not enlarged but the capillaries contain large numbers of poly-
morphonuclear leukocytes. These cells are also found occasionally in the
tubules and in the interstitial tissue. There is very little swelling of the
glomerular endothelium. This may be considered an e.xudative glomerulitis.

Fig. 3. — Case 17. Acute exudative glomerulitis. Note the polymorplionuclcars.
There is some enlargement of the endothelial cells.

C.\si; 24 (A-13-153). — Female, aged 13 years. Developed a sore throat,
Sept. 22. 1913. This continued with some improvement until September 28
when there was frequent emesis, anorexia and malaise. September 30 there
were definite symptoms of peritonitis, and two days later a laparotomy revealed
seropurulent peritonitis. Many streptococci were shown in smears of the exudate.
Leukocytes, October 2, 24,000. During the next live days there was a very small
amount of urine. Each specimen contained abundant albumin, blood and many
casts. There was no edema. Death. Oct. 7. 1913. Necropsy: Purulent peri-
tonitis; great enlargement of spleen and kidneys. Glomeruli are all enlarged
and the capillaries are for the most part closed by the swollen endothelial
cells (Figs. 6 and 7). There are many polymorphonuclears in some of the
glomerular capillaries, and there are large numbers of them in the lumina of
the tubules. There is severe tubular injury and many tubules are filled with
blood. Many mononuclear leukocytes are seen in the interstitial tissue. There
is no tubular atrophy. The total duration of the illness was only fifteen days.

Case 25 (A-15-144).— Female, aged 6 years. April 28. 1915, child was dull
and tired and had lost her appetite. Severe diarrhea and vomiting for the

778

ARCHIVES OF 1XTERX.4L MEDICIXE

next six days. Xo edema. Convulsions, May 7. Urine: marked albuminuria;
hyalin, granular, waxy, pus and blood casts. Leukocytes, 10,600; erythrocytes,
2,400,000. Death, May 7. Necropsy: No anasarca. Edema of lungs. Kidneys
weighed 175 gm. ; very cloudy cortices. Severe tubular degeneration and
necrosis of many tubules. Large numbers of casts. Glomeruli are not
enlarged but many of them show extensive hyaline, granular, hydropic and
fatty degeneration and occasionally some necrosis of the endothelium. These
changes are due to thromboses of the afiferen't glomerular arteries. Rarely
a mitotic figure is seen in the endothelium. This case may be considered a
severe nephrosis, since the changes are almost entirely degenerative in character.
C.ASF. 26 (A-15-16S). — Male, aged 55 years. Illness of about six weeks"
duration. Began with severe pain in the chest, dyspnea, edema of legs and
anorexia. L'rine : abundant albumin, hyalin and granular casts, leukocytes and
erythrocytes. Temperature about 100 F. Necropsy: Marked anasarca, ascites
and hydrothorax. Edema of lungs. Heart weighed 410 gm. No focus of

Fig. 4. — Case 20. Degenerative g

after

infection found. Kidneys not enlarged, but cortices are grayish yellow. Glomeruli
are all enlarged. Enormous increase of nuclei. Almost complete obliteration
of all glomerular capillaries. Many polymorphonuclears mark the position of
the capillaries. No tubular atrophy.

C.^SE 27 (A-17-176).— Female, aged 66 years. Severe bronchitis first week
of August, 1917. This was followed by dyspnea, precordial pain, vertigo and
general weakness. Gradual increase in severity of symptoms. On admission
to hospital, Aug. 25, 1917, there was edema of the feet. Urine, .\ugust 28
and September 6, contained a large amount of albumin and many casts of
all types. August 28, leukocytes, 17,000; 94 per cent, polymorphonuclears.
Blood pressure, August 30, 220/180; September 6, 200/80. August 29: urea
nitrogen 72; creatinin 1.7. August, 31, urea nitrogen 46.5; creatinin 2.5. Phcnol-
sulphonephthalein, August 31, 27 per cent.; September 6, 12 per cent. Death,
September 8, 1917. Necropsy: a large amount of seropurulent fluid in each
pleural cavity. Small areas of bronchopneumonia. Heart, normal. Kidneys
weighed 300 gm. ; cloudy surfaces. Glomeruli are all moderately enlarged.

BELL-HARTZELL—CLQMERVLOXEPHRITIS 779

Thert is a notable increase in the number of nuclei. Nearly all the capillaries
are completely closed by swollen endothelium. :Many disintegrating polymorpho-
nuclears are seen in the positions of the closed capillary lumina. The tubules
connected with a few of the glomeruli show a definite early stage of atrophy.
The glomeruli are not all injured to the same degree ; a few have a number
of permeable capillaries. Duration of illness about one month.

C.\SE 28 (.•\-18-62').— Male, aged 27 years. Right kidney removed February 21.
This kidney showed tuberculosis but no glomerulonephritis. Extensive sup-
puration of the surgical wound developed. Septic temperature. Urine: March
10 and March 14. large amount of albumin, pus and casts. Erysipelas appeared
April 1. Death, April 2. Necropsy: edema of left leg. Enormous dissecting
abscess of abdominal wall extending from the surgical wound. .\ small amount
of excess fluid in the serous cavities. Left kidney weighed 290 gm. Cortex
opaque with yellowish and reddish mottling. A majority of the malpighian
bodies show epithelial crescents which compress the glomeruli (Fig. 8). There
is very little swelling of the endothelium. There are large numbers of poly-
morphonuclears in the glomeruli, capsular spaces, and surrounding the
malpighian bodies. There is severe tulnilar degeneration. This is the extra-
capillary ty])e of glomerulonephritis.

Fig. 5.— Case 22. .\cutc pruliiL-r.ai\c yluiriLrulni.,- ,-iiiall portion of a
glomerulus. Note complete obliteration uf the capillaries.

Cask 29 ( A-18-251 ).— Female, aged 18 years. Admitted to hospital Xov. 27.
1918. Septic temperature. Severe prostration. Peritonsillar abscess, .\bscess
was opened November 27 and again November 28. There was erysipelas of the
face on admission and the entire face was involved before death. Blood
pressure, 96/54. November 27, 2 ounces of urine was removed by catheter.
November 28, 1 ounce removed by catheter. Urine contained a large amount of
albumin with enormous numbers of casts. Death, December 1. Necropsy:
Slight edema of ankles. No fluid in the serous cavities. Heart not enlarged.
Extensive bronchopneumonia. Hemolytic streptococcus from the blood.
Kidneys weighed 540 gm. Cortices were swollen and very cloudy. All the
glomeruli are moderately enlarged. There is a little swelling of the endo-
thelium. There are large numbers of polymorphonuclears in the glomeruli capil-
laries, in the capsular spaces, and in the tubules. There is severe tubular
injury.

Case 30 (.\-19-152).— Male, aged 12 years. Attack of smallpox about the
middle of June, 1919. Has had edema of the face and extremities ever since.
Vomiting and convulsions July 18. Admitted to hospital July 19. Dyspnea.
Frontal headache. Rales throughout the chest. Dulness over right lower lobe.
Leukocytes, 19,800; 94 per cent, polymorphonuclears. Blood pressure, 132/94.

780

AKCHirES OF IXTERXAL MEDICI XE

Urine: abundant albumin, many granular and epithelial casts. Death. July 21,
1919. Necropsy: Ascites (500 c.c.) ; hydrothorax (each cavity about 1,000 ex.),
edema of face. Heart weighed 200 gm. Small areas of bronchopneumonia.
Spleen weighed 220 gm. ; kidneys, 245 gm. Swollen, cloudy cortices. .-Ml
glomeruli enlarged. Nearly all the capillaries closed by swelling of the endo-
thelium. Very few polymorphonuclears. Severe tubular injury. No tubular
atrophy.

C.\SF. 31 (.^-21-62 1. — Female, aged 7 weeks. .\ few days after birth the
mother noticed puffiness of the eyelids which gradually increased. On the
eighth day there was a profuse purulent discharge from the eyes with swelling
of the lids. This discharge continued. When the child was 5 weeks old it was
brought to the university dispensary for treatment. Xo gonococci were found.
Some improvement under treatment. February 4, the mother first noticed
swelling of the face, hands and neck. .Admitted to hospital, Fe1i. 7, 1521.
General anasarca. Svstolic murmur at liase of heart. No convulsions. Xo

Fig. 6. — Case 24. I'ruliferative glomcruliti;., intracapillai
cation. Note solid lobules, also polymorphonuclears in 1

)\v magnih-
tulndes.

spasticity of muscles. Phenolsulphonephthalcin, a trace. Creatinin, 4.23 ing. ;
urea nitrogen, 29.16 mg. Very little urine excreted. A few drops removed
by catheter showed hyalin, granular, epithelial and erythrocyte casts. Hemo-
globin, 6.3 per cent. Leukocytes, 24,800; 30 per cent, polymorphonuclears: 65
per cent, lymphocytes. Temperature from 97 to 98.6 F. Death Feb. 14, 1921.
.•\cutc tibrinopurulent pleuritis of right side. Infected infarct of right lung
(streptococci in smears). Streptococcus in pure culture from heart's blood.
Kidneys weighed SO gm. Cloudy yellowish cortices. Smooth external surfaces.
F.xtensive involvement of nearly all the glomeruli. The capillaries are closed,
and there is beginning hyaline degeneration in many glomeruli. There arc very
few polymorphonuclears. There is beginning tubular atrophy.

C.\SF, 32 (.\-20-215).— Male, aged 60 years. .A.dmitted to hospital May 25,
1920, in coma. Temperature as high as 105 F. Blond pressure, 160/80. Urine:
specific gravity 1020; moderate amount of albumin: many erythrocytes; many
granular casts. Leukocytes. 21.000; 70 per cent, polymorphonuclears. Blood

BELL-HARTZELL—CLOMEHLLOXEPHRITIS

781

chemistry : creatinin, 5.8 mg. : urea nitrogen, 86.2 mg. ; sugar 0.19 per cent.
Death. May 26. 1920. Necropsy: ascites (300 c.c). Heart weighed 350 gm.
Edema of lungs. E.xtensive ulcerative colitis. Kidneys weighed 305 gm. ;
cloudy cortices. Severe degenerative changes in both tubules and glomeruli.
Very slight evidence of reaction in glomeruli. This may be considered a severe
nephrosis.

The Clinical Pliciioniciia in Acute (iloincnilnnephritis. — In twenty-
one of the thirty-two cases, ihe renal symptoms were so masked by the
associated disease lliat a diagnosis of nephritis would have been very
difficnlt ; and in Cases 1 to 'i the in\oh-ement nf the kidnevs was

Fig. 7. — Case 24. .\cute glomerulitis, proliferative and exudative. Note swell-
ing and increase of endothelial cells with clo.sure of capillarie.s. There are a
number of polymorphonuclears.

probably too slight to produce any prominent tindings. Cases of this
kind are, of course, not ordinarily considered as Bright 's disease by the
clinician ; but they are very valuable in the study of the early stages
of the disease, since the changes in the kidneys are of the same lyjie
as those of typical clinical acute Bright 's disease.

Considering only the eleven typical cases (Cases 22 to 32), it will
be noted that eight showed edema of some part of the body. In Cases
24, 25 and Z2. edema was absent. All showed heavy albuminuria and
numerous casts, usually granular. Renal function was tested in four
patients. Cases 27 and .^1 showed a slight, and Ca.se 32 a marked

782 ARCHIVES OF IXTERXAL MEDICI XE

retention of metabolites in the blood. The phthalein elimination was
found decreased in Cases 22, 27 and 31. The phthalein measurement
in Case 31 is not consistent with the blood chemistry and was probably
incorrectly determined because of the very small amount of urine
excreted. The duration is difficult to determine in many instances but
seems to vary from a few days to three months. In only three patients
(Cases 2^. 26 and 30) did death ^eeni to be due mainly to renal involve-
ment; in the other eight cases there were very sexere complicating
infections. The data as to age and sex have little \-a!ue because of the
small number of cases.

f •'« -T^^'JC^^^lt*!

%'..
.*^!

(w-o.w Chaiu/cs in the Kidneys. — With the cxceinion of the very
mild cases the C()rtice.■^ are invariably cloudy, and in the more se\ere
cases the cloudiness and opacity are very pronounced and sometimes
a yellowish tinge can be detected. The kidneys are not always enlarged.
In general, the enlargement is pro])ortional to the severity, but occasion-
ally severe injury is found in kidneys of normal weight (Cases 26. 27
and i2). Among seventeen adults whose weights are recorded there
are eight cases in which the kidneys weighed more than 400 gm.. and
four weighed more than 500 gm. The single kidney in Case 28 weighed
290 gm. It is apparently not possible to distinguish a nejibrosis from
acute glomerulonephritis by the gross appearance.

The Normal Cloiiieniliw;. — Before studying the ylmuerular changes
in .-R-utc nejihritis attenticm -biuild be direcled lo liu' siruclure of the

BELL-HARTZELL—0L0MERVL0\EPHR1TIS

783

normal glomerulus. The microscopic appearance of a glomerulus varies
with the amount of blood it contains. When distended, the lumina
of the capillaries are large, their endothelial walls are thin and the
individual capillaries are fairly distinct. When empty, the endothelial
walls of the capillaries are thicker, their lumina are very small or even
invisible, and it may not be possible to see the individual capillaries.
The glomerulus is, of course, much larger when distended than when
empty. In necropsy material the great majority of the glomeruli
ordinarily contain only a little blood ; and it is often difficult in disease
to distinguish an empty capillary from one in which the lumen has been
closed by endothelial swelling. Thin sections of well fixed material are
necessarv for the studv of glomerular structure.

Fig. 9.— Case 55. Chronic case. Section through entrance of artery. Note
lobules. A few polymorphonuclears are visible at the site of closed capillaries.

The glomerulus is composed almost entirely of capillary loops which
anastomose freely. The arrangement of the capillaries is shown well in
Johnston's ° reconstruction. Corresponding with the capillary loops, the
glomerulus is subdivided into a number of small lobules which have
their narrow apex in common near the point of entrance of the artery
and their wider bases free at the periphery. This lobulation is best seen
in diseased glomeruli where there has been some ?hrirl<age but is some-
times visible in the normal. The lobulation is well shown in Figure 9
in a section through the entrance of llic artery. In ordinary histologic

6. Johnston. W. B. :
Kidney, .Xnat. .Xnzcigcr

.\ Reconstructi
16:260. 1859.

)f a Gloiiieruhis of the Human

784 ARCHII'ES OF IXTERXAL MEDICINE

preparations no connective tissue is visible between the capillaries, but
Johnston was able to demonstrate fine reticulum fibrils. These fibrils
apparently do not take any part in inflammations of the tuft.

In Figure 10 a small part of a normal congested glomerulus is
shown under high magnification. Most of the capillaries are distended
with erythrocytes, and the endothelial cells appear as thin plates where
the plane of the section is about perpendicular to the course of the
vessel.

The capsule of Bowman lines the outer wall of the capsular space
and is reflected over the glomerulus. The inner layer is prominent in
infancy but apparently does not form a continuous layer in the adult
kidney. Some of the cells of this layer are found in the clefts between
the glomerular lobules (Fig. 10). It is often prominent in shrunken

Fig. 10. — Xormal congested glomerulus. Note the thinness of the endothelium
when the capillaries are distended.

glomeruli. The outer layer takes a prominent part in liie extracapillary
type of lesion which will be described below.

The Glomerulus in Acute Inflamm.vtions. — The tliree funda-
mental phenomena of inflammation, viz., degeneration, exudation and
proliferation — occur in the gknuerulus as they do in other tissues and
their relative prominence determines the microscopic appearance. The
peculiar structure of the glomerulus, however, influences markedly the
course and outcome of the inflammatory process. In mo.st instances the
only fixed tissue concerned is capillary endothelium and the occlusion of
capillaries produces permanent damage that seems out of proportion
to the intensity of the injury. In accordance with the prominence of
the fundamental processes three types of glomerulitis may be described.

(a) Degenerative Glomerulitis. — In this form there is disintegration
and necrosis of the glomerular endothelium with escape of blood into
the capsular space. It is usually associated with severe tubular injury

BELL-HARTZELL—GLOMERVLOXEFHRITIS 785

also. The reactive changes are very slight. Case 32 is a good e.xample
of a severe case with involvement of all the glomeruli. In Cases 25 and
20 the degenerative changes are largely due to thrombosis of afferent
arteries and many glomeruli escaped serious injury (Fig. 4).

Mild degenerative changes may be responsible for a hematuria.
Figure 11 is from a case of severe hematuria apparently due to a
bacteremia. A large percentage of the glomeruli show a similar appear-
ance. The blood escapes from ruptured caiiillaries but the glomerulus
as a whole shows no signs of permanent injury. Tl is possible that
some of the cases in children described by Mill ' as hemorrhagic
nephritis have only trivial glomerular injuries such as this. Hill states

Fig. 11. — De.tjcnerativi
case of bacteremia with

glomerulitis. .\o pennaiieiit rena
svere hematuria.

that these cases have an especially favorable prognosis. Renal hematuria
does not necessarily indicate a serious glomerular injury. .\ few
erythrocytes in the urine may be due to passive congestion or to mild
injuries such as occur in any nephrosis.

(b). Exudative Glomerulitis. — This lyjje is characterized by the
appearance of an unusual number of polymorijhonuclear leukoc^-tes
in the glomerular capillaries. The leukocytes pass through the capillary
walls into the capsular space and are carried away in the urine. There
is usuallv an associated >welling «l the ciKldthclium but in seven of

7. Hill. L. W.: StlKlie.^ in th,
17:270 CApril) 1919.

\eplir

Children

786 ARCHIVES OF IXTERXAL MEDICISE

our cases this is inconspicuous and the lesions are mainly exudative.
Some polymorphonuclears were found in the glomeruli in nearly all
our cases of glomerulitis, but in ten of them they were very rare.
Figure 3 (Case 17) is a good example of an exudative glomerulitis.
There is apparently no narrowing of the capillary lumina in this case
and it does not appear that any serious permanent damage has been
done. Possibly the plugging of the capillaries with leukoc}^es may
interfere with function to some extent. It is an attractive hypothesis
that mild cases of acute Bright's disease are of the exudative type, but
there are no supporting observations. We do know, however, that our
severe clinical cases were either mainly proliferati\e or mainly
degenerative.

Fig. 12. — Case 47. From a chronic case. Low magnilicati
morphonuclears in the soHd lohiilcs. See Figures 13 and 14.

Xote poly

When acute cases become chronic the leukocytes remain permanently
in the occluded capillaries and remnants of them are recognizable for
a long time (Figs. 12, 13. 14). This feature helps to establish the
relationship between the acute and the chronic case.

In one instance (Case 16) the exudate was so abundant that a
numlaer of glomeruli were converted into small abscesses (Fig. 2).
The other glomeruli showed the changes characteristic of typical
glomerulonephritis. There was a large amotmt of exudate in the
interstitial tissues also. This case shows the close relationship between
the pyelonci)hritis group and glnnu'ruloncphritis. and furnishes an
argument in favor of the infcclinu-- nalurc df ihe latter di>caNC.

BELL-HARTZELL—GLOMERLLOSEPURITIS 787

( c- ) Proliferative Glomenilitis. — Occurs in two forms : the extra-
capillary and the intracapillary. The extracapillary type consists in
proliferation of the cells of the outer layer of Bowman's capsule. These
newly formed cells usually become arranged in the form of a crescent
(Fig. 8), which compresses the glomerulus and ultimately causes it
to undergo atrophy and hyaline degeneration. The epithelial crescent
itself finally undergoes hyaline degeneration but remnants of nuclei
are visible for a long time. The presence of old epithelial crescents
in a chronic case (Fig. 15) suggests an acute beginning. Extracapillary
lesions are not frequently seen. They were found in only two of our
acute cases (Cases 16 and 28). According to ^'olhard and Fahr they
occur especially in infections with violent onset, and this view is sup-

Fig. 13. — Case 47. High magniticatiun uf I'lgiirc I
are visible in the solid lobules at the site of capillaries
by swelling of their endothelium.

Polymorphonuclears
lich have been closed

])(irtccl by our cases, Case 28 being tht- most severe infection of the
entire series.

The intracapillary type is the most common and the most important
form of glomerulitis. It consists in swelling of the endothelial cells and
increase in their number. The glomerulus as a whole is enlarged
and there is usually a notable increase in the number of nuclei. Rarely
a mitotic figure is seen. More important than the increase in the number
of the endothelial cells is their increase in size. The capillary lumina are
obliterated so that entire lobules may appear solid (Fig. 5). Low
power views of the actite stages in the glomeruli are shown in Figures
1 and 6. and high power views of a few lol)uIcs are seen in T'igures

788 ARCHU-ES OF IXTERXAL MEDICIXE

5 and 7. Polymorphonuclear leukocytes are frequently present, occa-
sionally in large numbers (Fig. 7); but some times they are absent
entirely from large portions of the glomerulus ( Fig. 5) . These leukoc\tes
are frequently seen also in the lumina of the tubules (Fig. 6), and in
the interstitial tissue (Fig. 16). They are often caught in the closed
capillaries where they are easily recognized in subacute stages before
hyaline degeneration of the glomerulus begins (Figs. 12, 13 and 14);
and they may be seen by careful examination in many chronic stages
(Figs. 9 and 17). This microscojjic feature furnishes an important
connecting link between acute and chronic glomerulonephritis. As a
result of intracapillary glomerulitis, the entire capillary network of the
tuft may be closed completely, in which event the glomerulus undergoes
hyaline degeneration and its tubule atrophies to the point of complete
disappearance. Frequently, however, some of the capillaries are not
occluded (Fig. 17), and the glomerulus continues to function to a
limited extent, in which event only a partial atrophy of the associated

Fig. 14. — Case 47. See Figures 12 and 13. High inagnili
remnants of polymorphonuclears in the solid lobules. Capill;
obliterated by growth of endothelial cells.

tubule occurs. This type of damaged glomerulu> is \ery cnnimun in
chronic glomerulonephritis.

The tubules in acute glomerulonephritis usually slmw some degen-
erative changes. An increased amount of fat is often demonstrable;
and hyaline granular degeneration is occasionally seen. In the degen-
erative ty])e there may be some necrosis. As soon as the glomerulus
ceases to function the tubule begins to atrophy, but this is never marked
in an acute case. In Cases 27 and 31 atrophy of ^ibules had begun.

The interstitial tis.sues frequently show an exudate of mononuclear
or polymorphonuclear leukocytes (Fig. 16). This exudate, liowever,
does not seem to destroy any tubules and probably does not affect the
course of the disease to any noteworthy extent.

ETior.onv OF .xcute gi.omf.rulonepiiritis
The prevailing opinion in the literature is that acute glomerulo-
ne])hritis is closely associated with infectious processes. Our experience

BELL-HARTZELL—GLOMERULOSEPHRITIS 789

is entirely in accord with this view. Table 1 gives the associated
infections found at necropsy or determined clinically in our thirty-two
cases. The numbers refer to the individual cases. It will be noted
that some of the patients had more than one infection.

The frequency of acute endocarditis in our series is very impressive
— twelve times in thirty-two cases, or Z7.5 per cent. In seven of
these hearts the valve leaflets showed chronic changes also, so that
these are to be regarded as acute exacerbations of a chronic endocarditis.
That the frequent association of these two conditions is not accidental

TABLE 1. — .'Vssoci.ATED Infections in C.\sf.s of Acute Glomerulonephritis

Associated Infections Cases

Vegetative endocarditis 1.2,3,5,10.11,12,13.14,16,19,21

Puerperal sepsis 5

Empyema, or purulent pleuritis 5,20,22,23,27

Peritonitis 20,22,24

Streptococcic bacteremia 2,5,10,17,29.31

Pneumococcic bacteremia 9.20

Septic sore throat 7,24,29

Erysipelas 4,28.29

Lobar pneumonia 8,12

Suppurative otitis media 7

Suppurative pharyngitis 9

Streptococcic arthritis and osteomyelitis 4

Infected surgical wound 28

Meningitis 7

Acute arthritis (not present at time of death) 10,11,13,16

Severe bronchitis 27

Variola 30

Tuberculosis with cavities 15

Ulcerative colitis 32

Diabetic gangrene 17

Purulent conjunctivitis 31

No localized infection 6,18,25.26

is shown by the fact that in our 3,300 consecutive necropsies, not
including those with acute nephritis, there were only sixty-three with
acute endocarditis, or only 1.9 per cent.

This relationship has been observed by others. Leyden ' noted
the as.sociation of rheumatism, endocarditis and nephritis. Councilman '
found ten instances of acute endocarditis in twenty-eight cases of acute
nephritis. Klotz '" also found endocarditis and nephritis in frequent
association. Ophiils " cites three cases of acute nephritis associated
with acute exacerbations of a chronic endocarditis.

S. Leyden. Cited from Mannalicrg."

9. Councilman. W. T. : .^n .iXnatomical and Bacteriological Study of .^cute
Diffuse Nephritis, Am. J. M. Sc. 114:23, 1897.

10. Klotz, O. : Chronic Interstitial Nephritis and .Arteriosclerosis, .-Xm. J.
M. Sc. 150:832, 1915.

11. Ophiils. \Vm.: A New Series of Cases with a Review of Recent
Literature. Stanford Univ. Med, Bull. No. 3. 1915.

790

ARCHIVES OF IXTERNAL MEDJCIXE

Four cases in our series gave a history of rheumatic fever prior to
the terminal iUness, and several writers seem to consider this disease
closely related to acute nephritis.

The relation of tonsillitis to acute glomerulonephritis has been dis-
cussed by a number of investigators. Kannenberg ^' in 1879 reported
three cases of tonsillar infections (two were peritonsillar abscesses)
followed by acute nephritis. In one cases the urinary changes appeared
on the seventh day. Mannaberg " mentions tonsillitis among the
infections sometimes followed by Bright's disease. Bluhm ^* reported

> , '

; ;f f J

m

¥4

m

1

m

I

j

Fig. 15.— Case 56. Old epitli.
beginning hyaline degeneration.

crescent. Coniiiress

glomerulus with

that five out of seventy-four cases of tonsillitis developed Bright's
disease. Lohlein '■' and Aschotif ""' mention angina among the causes
of glomerulonephritis. \'olhard and Fahr ' attribute to angina seven-
teen out of seventy-one acute, and seven of thirty-two subacute cases.
In one instance cited by these authors nephritis appeared fourteen days

nsl<i

12. Kannenberg: Ucbcr Xepliritis hci acutcn
f. klin. Med. 1:506, 1879.

13. Mannaberg, J.: Zur .^etiologie des Morbus Brightii acutns
f. klin. Med. 18:223. 1890.

14. Bluhm, A.: Zur. .Aetiologie des Alorbus Brightii. Deutscl
Med. 47:193, 1891.

15. Lohlein, M. : Ueber Nephritis nach dem hcutigcn Stam
anat. Forschung, F.rgeb. dcr inn. Med. n. Kindcrh. 5:411, If 10.

16. .^schonf. L.: Pathologischc .Xnatoinie. 2:483. 1<;21.

Ztscbr.

BELL-HARTZELL—GLOMERULOXEPHRITIS 791

after an aitack ot sore throat. Ophiils '" attaches great importance
to tonsillar infection as a source of glomerulonephritis, and
Herxheimer '* seems to hold a similar view. Hill '=• found tonsillitis
the most frequent cause of nephritis in children, and James' -" obser-
vations are in agreement with those of Hill. Hill states that the renal
symptoms usually appear about one week after the tonsillar infection.

Evidently there is a close relationship between tonsillitis, endo-
carditis, arthritis and acute glomerulonephritis. It seems probal)le from
all the. accumulated evidence that the pathogenic bacteria gain access

Fig. 16.— Case 43. From a chn.nic case. The
lobules is largely clo.setl. Many polymorphonuclears

[lillary network of the
the interstitial tissues.

to the blood streaiu from an infected throat and that the suhseciuent
clinical picture depends on whether they attack tlie lieari valves, joints
or glomeruli.

There were three cases of septic sore throat in our scries, and
there were ten cases with aciUe endocarditis in wliicli the throat was
a possible source of the infection.

17. Ophiils. \Vm.: The Etiology and Development of Nephritis, J. \. M. A.
69:1223 (Oct. 13) 1917.

18. Herxheimer, G. : Ueber den jetzigen Stand unserer anatom, Kenntnisse
der Xephritis u. Xcphropathien, Miinch. med. Wchnschr. 65:283, 1918.

19. Hill, L. W. : .Acute Xephritis in Childhood. J. A. M. A. 73:1747
(Dec. 6) 1919.

20. James, R. F, : Prognosis of Xephritis in Childhood, J. A. M. A. 76:505
(Feb. 19) 1921.

792 ARCHirES OF IXTERXAL MEDICIXE

Scarlet fever was one of the first diseases in which a relationship
to nephritis was observed. According to Friedlander -'^ post-scarlatinal
glomerulonephritis was first described by Klebs. Urinary disturbance
occurs in from 20 to 25 per. cent, of cases of scarlet fever (McCrae,^''
25 per cent, of 1,034 cases ; Sorensen,-^ 20 per cent, of 365 cases). But
the abnormal urine is usually due to acute interstitial nephritis or
nephrosis. Only 2 per cent, of McCrae's patients had physical signs
and symptoms of a true nephritis. Friedlander, Sorensen and others
believe that true glomerulonephritis usually develops in the convalescent
stage, during the third or fourth week after the onset of the disease.

'W0

/r

Fig. 17. — Case 43. A common form of damaged glomerulus in clironic
glomerulonephritis. Some capillaries are closed but many are pern\cable. The
tulnile associated with this glomeruhis showed moderate atrophy.

hi fatal cases of scarlet fever glomerulonephritis is frequently seen
(I'^riedlander, in 18 per cent, of 229 necropsies; Reichel,-* in 29 per
cent, of fifty-eight necropsies). Aschoff considers scarlet fever a
frequent cause of glomerulonephritis. \'olhard and Fahr found nine-
teen of seventy-one acute cases, and two of thirty-two subacute cases
due to scarlet fever. The older observers considered the glomerular

21. Friedlander. C. : Ueber Neiihritis scarlatinosa. Fortsclir. d. Med. 1:81,
1883.

22. McCrae, J.: Incidence of Nephritis Following Scarlet Fcvor. Tr. .\ssn.
Am. Phys. 28:194, 1913.

23. Sorensen. S. T.: Ueber Scharlachncphritis, Ztschr. f. klin. Med. 18:298,
1890.

24. Reichcl, H.: I'eber Nephritis bei Scharlach. Ztschr. f. llcilk. 6:72. 1905.

BELL-HARTZELL—GLOMERULOSEFIIRITIS 793

injury due to poisomius substances excreted by the kidneys, but recent
investigators attribute it to streptococcic infection.

There are no cases in our series in which a history of scarlet fever
was obtained.

One of our cases followed an attack of xariola. but apparently
this is not a common cause since Bluhm found only one case of Bright's
disease among 481 cases of variola.

Lobar pneumonia appears twice in uur series as a possible source
of infection; but Councilman found glomerulonephritis only once in
107 cases of pneumococcus pneumonia.

Fig. 18. — Case 51. Chronic case six j-cars after an attack of scarlet fever.
These areas are probably the result of an acnte interstitial nephritis during
the attack of scarlet fever.

Three of our series were asstjciated with erysipelas and Bluhm
found Bright's disease seven times in 162 cases of erysipelas. This
disease is mentioned by several observers as a cause of acute Bright's
disease.

A large number of other infections, some of which appear in our
series, are mentioned by various investigators as having a causal
relationship to acute glomerulonephritis. .Xmong these are infected
wounds, puerperal sepsis, peritonitis, empyema, impetigo, osteomyelitis,
tuberculosis with cavities, otitis media, etc. Gaskell ='' describes a case

25. Gaskell. J. F. : On the Chans-es in
Inflammatory and .\rteriosclerotic Kidney Dis>
1911.

("ilomeruli ai
I I'aih .V Bn

.Arteries in

riol. 16:287,

794 ARCHIVES OF IXTERXAL MEDICIXE

of less than one week's duration following peritonitis. McEIroy -"
cites a characteristic case developing two or three weeks after a severe
infection of one hand. Cases such as these are apparently frequent in
the practice of many clinicians, and there are probably few of wide
experience who have not seen a case of acute Bright's disease that
followed some infectious process.

The various infections which apparently cause acute glomerulo-
nephritis are generally due to streptococci, less frequently to pneumo-
cocci. In scarlet fever it is believed that a streptococcic infection of
the throat develops in the period of convalescence, which gives rise to
the renal involvement. Lcihlein. ^'olhard and Fahr, AschofY and Ophuls

Fig. 19. — Case 53. Chronic case, ^llo\^lll^ alnn■^t complete occlusion of all
tlie capillaries. Xo hyaline degeneration. Marked atrophy of associated tubule.

all agree that streptococci are chiefly responsible for acute glomerulo-
nephritis. Ophiils apparently believes that other bacteria, e. g., B.
influenzae and B. eoli. may occasionally produce this disease.

In our series there were six cases with -trejitococcic bacteremia.
In three of these the organisms were hemolytic: but there is no record
as to the type in the other three cases. Xonhemolytic streptococci have
not been found in our cases of acute glomerulonephritis, but they have
been cultured from the blood in several cases of embolic glomerulo-
nephritis. Further study is necessary to determine whether the non-
hemolytic strains are ever responsible for acute Briglit's disease.

26. McElroy. J.

Xephropathics, M. Clin. X. America 1:14.^7. lOlS.

BELL-HARTZELL—GLOMERILOXEPHRITIS

795

Two instances of pneumococcic bacteremia appear in our series —
one associated with suppurative pharj-ngitis and bronchopneumonia, the
other with suppurative pleuritis and peritonitis.

It has not been determined whether the bodies of the bacteria or
some diffusible toxin produces the glomerular injury. In favor of the
toxin theory is the absence of bacteria in the glomeruli, and the diffuse
uniform character of the lesion — practically all the glomeruli are usually
involved. A few of the earlier workers described bacteria in the
glomeruli, but Ophiils has pointed out that more recent contributors
by careful histologic examination have failed to find bacteria in the
glomerular endothelium. Sections from ten of our acute cases were

Fig, 20.— Case 56. Chronic case, showing complete closure of capillaries and
beginning hyaline degeneration. Lobules are very distinct. Note disintegrated
polymorphonuclears in the lumina of the tubules.

Stained by the dram-Weigert method, but no bacteria were found in
the endothelium. However, in sections from an acute case following
erysipelas, which were furnished us by Dr. J. P. Schneider, large
numbers of gram-positive cocci were easily seen in the swollen endo-
thelial cells of the glomerular capillaries. Ophiils has suggested that
the bacteria undergo rapid lysis in the endothelium and are therefore
seldom seen. This view is supported by Pappenheimer, Hyman and
Zeman," who injected bacteria directly into the renal artery of the

27. Pappenheimer, Hyman and Zcman :
Injections of Bacteria Into tlie Renal
16:73. 1516.

.'\cute Glomerular Lesions Following
rterv. Froc. New York Path. Soc.

796

ARCHU'ES OF IXTERXAL MEDICI XE

rabbit. They found that the bacteria were taken up by polymorpho-
nuclears and endothelial cells in the glomeruli within a few minutes
and that within four hours they had nearly all undergone complete
intracellular digestion.

War nephritis is anatomically acute glomerulonephritis, according
to all who have studied necropsy material. It differs clinically in its
epidemic character. Several observers have noted a frequent association
with infectious processes ( Brown,-* Tytler and Ryle.'-" .\meuille ^'') ; but
all seem to agree that a localized primary infection is not demonstrable
in a majority of cases. Streptococci have been demonstrated in the

Fig. 21.— Case 6U Typical lir nic e 1 w ma^nilication. The larger

tubules on the ktt ire (.oinie t 1 itl ..,1 lerul thit lias a numlier of

permeable capillaries.

urine but they may have been secondary invaders. No satisfactory
etiologic studies have been re])orted.

SUB.\CUTE .XND CIIKOXIC GLO.M ERfLCNKrilRITIS

The arrangement of our cases into subacute and chronic groups is
arbitrary, since these subdivisions are not sharply defined. Cases 33
to 41 may be considered subacute. When the knciwn dtiralinn of the
disease is over one vear it is Cdiisidcrcd chronic. When ;is many as

28. Brown, L. : Epidemic Nephritis, Brit. M. J. 2:723. 1916.

29. Tvtler and Rvle: Clinical and Pathologic Notes on Trench Nephritis,
Quart. J^ M. 11:112, 1917.

30. Ameuille, P.: Du role de rinfcctinn dans les nephrites <le guerre, .\nn.
de med. 3:298, 1916.

BELL-HARTZELL—GLOMERULOXEPHRITIS

797

ten per cent, of the glomeruli have become complete hyalinized and their
tubules show advanced atrophy, the case is considered chronic, since
the study of cases of known duration has shown that a long time is
required for this change to occur. But the rate of tubular atrophy is
not the same in all cases and it is probable that this feature is not an
accurate index of the duration of the disease.

Case 33 (A-12-131).— Male, aged 33 years. Admitted July 17, 1912. About
one month before admission he developed dyspnea which was increased by
exertion and gradually became worse. One week before admission his feet
began to swell and dyspnea became very alarming. On admission he corn-

Fig. 22. — Case 69. Chronic case, showing old epitheli
tubules and interstitial exudate.

crescents, atrophi

plained of precordial pain, cough, dyspnea, and swelling of the feet. The urine
contained albumin at all times. The specific gravity was 1.030 July 18; 1.012
September 8. Death, Oct. 7, 1912. Necropsy: Edema of ankles, moderate
ascites and hydropericardium ; edema of lungs. Heart weighed 460 gm. ;
moderate thickening and retraction of the mitral and aortic leaflets. Kidneys
weighed 415 gm. ; cloudy cortices; smooth external surfaces. The glomeruli
are all involved but in varying degrees. There are a number of epithelial
crescents which compress the glomeruli. Most of the glomeruli show swelling
of the endothelium with partial or complete closure of the capillaries but none
of them show any hyaline changes. Nearly all the tubules are moderately
atrophic. A few polymorphonuclear leukocytes are seen.

Case 34 (A-13-8).— Female, aged 26 years. Attack of pericarditis in October,
1912. The pericarditis cleared up but hematuria appeared and persisted. Read-
mitted to hospital Dec. 23, 1912. Edema was never very prominent. Blood and
casts were continuously present in the urine. Toward the end of her illness
there was severe hematuria and hemorrhages from the vagina and mouth.
Death Jan. 10, 1913. Necropsy: No edema; ascites C200 c.c). Heart not

798 ARCHIVES OF IXTERXAL MEDICIXE

enlarged. Kidneys not enlarged. E.xternal surfaces smooth. Cortices cloudy.
Glomeruli are all involved. The great majority show epithelial crescents with
compression of the tufts. Some show endothelial swelling with closure of the
capillaries. There are numerous polymorphonuclears in some glomeruli. Many
tubules and capsular spaces are filled with blood. Numerous mononuclear
leukocytes (mainly plasma cells) are seen in the interstitial tissues. Nearly
all the tubules show notable atrophy. No hyaline glomeruli.

C.»,SE 35 (A-16-223).— Infant, aged 6 months. Normal at birth. At end of
fourth month developed pallor, slight jaundice and weakness. At end of fifth
month (May 29, 1916) was very anemic — hemoglobin 20 per cent.; erythrocytes
2,000,000. Blood stained vomitus and stools. Transfused. Suppuration of the
transfusion wound. General anasarca. Urine at various times showed albumin,
blood and pus cells. Death, June 26. 1916. Suppurating wound of thigh (site
of transfusion). Seropurulent pleuritis. Bronchopneumonia. Kidneys weighed

S|S^

^'

€

%

^.r

4P^

.%

^^^^

Fig. 23. — Case S3. Chronic case showing only a few permeable capilla
The associated tubule was very small but had not disappeared.

S4 gm. ; swollen; pale and cloudy; smooth surfaces. Nearly all the glomeruli
show marked swelling of their endothelium with closure of the capillaries.
Very few capillaries are patent. There are no hyaline glomeruli. Some tubules
are filled with polymorphonuclears, but these cells are very rare in the glomeruli.
Tubular atrophy is marked.

Case 36 (A-18-118).— Male, aged 21 years. Syphilitic infection in October,
1917. Treated for three months with mercury injections. One injection of
neodiarscnol early in December, 1917. Three days after this injection the patient
developed edema with a large amount of albumin, blood and casts in the urine.
Phenolsulphonephthalein at this time. 15 per cent.; urea nitrogen, 14 mg. :
creatinin, 1.7 mg. December 30, lobar pneumonia developed in right lower lobe.
Crisis si.x days after the attack. April 15, 1918, thrombosis of right popliteal
vein developed. Blood pressure at this time was 130/80. Later blood pressure
was 140/80. .Vscites was present a long time before death. Death, June 2, 1918.
Necropsy: Edema of feet and scrotum. .Ascites (2,000 c.c.) ; hydrothorax
(1,000 c.c). Heart weighed 375 gm. Fibrinopurulent pleuritis. Kidneys weighed

BELL-HARTZELL— GLOMERULONEPHRITIS 799

425 gm. Cloudy cortices. Smooth external surfaces. Nearly all the glomeruli are
severely injured. There is a great increase of nuclei with closure of most of the
capillaries, and beginning hyaline degeneration in a few glomeruli. Numerous
adhesions between the layers of Bowman's capsule. A few polymorphonuclears
in some glomeruli. Rather marked atrophy of tubules.

Case 37 (A-13-67). — Female, aged 35 years. Appendix, ovaries and tubes
removed about seven years ago. Has had swelling of feet for several years.
Abdominal paracentesis several times recently. Urine ; abundant albumin and
casts. Necropsy: marked general anasarca, ascites (3,000 c.c), hydrothora.x
(1,500 c.c.) and hydropericardium. Heart weighed 4(X) gm. Thromboses of
pelvic veins with infarction of lungs. Kidneys normal size; surfaces slightly
uneven; cortices yellowish. The great majority of the glomeruli are enlarged
and show a great increase in number of nuclei. Their capillaries are largely
closed by swollen endothelium, and hyaline changes have appeared in many.
Many polymorphonuclears are seen in the interstitial tissues and capsular spaces,
and there are a few in the closed glomerular capillaries. There is moderate
atrophy of most of the tubules. This case appears to be of longer duration
than Nos. 3i to 36.

C.\SE 38 (A-15-253).— Male, aged 51 years. Smallpox at age of 32 years.
Well until about April 19, 1915, when he developed a cold with cough following
wetting of his feet. There was a yellowish expectoration. About April 26
his face began to swell, and shortly afterwards his legs and abdomen became
swollen. There was shortness of breath, headache and sOme decrease in the
amount of urine. Admitted to hospital. May 10, 1915. Blood pressure 172/100.
Leukocytes, 13,200. Urine: specific gravity, from 1.020 to 1.030; moderate
amount of albumin; casts. Death, Aug. 5, 1915. Necropsy: Pronounced gen-
eral edema, ascites, hydrothorax and hydropericardium. Heart weighed 375
gm. ; kidneys, 500 gm. Smooth external surfaces. A number of glomeruli are
completely sclerosed, but these are probably not concerned with his terminal
illness. The great majority are somewhat enlarged and their capillaries are
largely occluded by endothelial swelling. There is a notable increase in the
number of nuclei. No hyalin changes have appeared. There is marked tubular
atrophy. Polymorphonuclears are abundant in the interstitial tissues but rare
in the glomeruli.

Case 39 (A-18-237).— Female, aged 22 years. Admitted to hospital, Nov. 1.
1918. A diagnosis of nephritis had been made by a physician three or four
months previously, and a therapeutic abortion had been performed about two
months before. On admission the patient was in coma and was having con-
vulsions every hour. Albuminuric retinitis was noted. She recovered from
the coma but had irrational periods and was stuporous most of the time until
her death. There was no urine voided on some days and only from three to
six ounces on other days. A friction rub over the heart was heard on
November 7. Abdominal paracentesis was performed several times. Blood
pressure 220/120. Urine: specific gravity, 1.018; numerous waxy and pus casts,
many erythrocytes, large amount of albumin; creatinin, 11.4 mg. ; urea nitrogen,
33 mg. Death November 13. Necropsy: Marked anasarca, ascites and hydro-
thorax. Fibrinous pericarditis. Some pus cells and cocci in plueral and peri-
cardial exudates. Heart weighed 300 gm. ; kidneys, 260 gm. External surfaces
smooth. Marked atrophy of nearly all the tubules. Very few hyalin glomeruli.
Numerous epithelial crescents. Most of the glomeruli show closure of their
capillaries with little or no hyaline degeneration. Many disintegrated poly-
morphonuclears in some of the tubules and in the interstitial tissue.

Case 40 ( A-19-5).— Male, aged 55 years. Was in the hospital about one
year ago at which time albumin and casts were found in the urine. There was
no edema then. Readmitted, November, 1918. Edema appeared about three
weeks before admission. Blood pressure, 200/120. Albuminuric retinitis.
Ascites. Phenolsulphonephthalein on admission 20 per cent.; shortly before

800 ARCHIVES OF JXTERXAL MEDICIXE

death, 0 per cent. Death, January, 1919. Xecropsy : Marked general anasarca.
Ascites (1,000 c.c.) — fibrinopurulent e.xudate on a few intestinal coils. Hydro-
thora.x (1,S00 c.c). Edema of lungs. Heart weighed 420 gm. ; kidneys, 520 gm.
External surfaces smooth. There are very few glomeruli that are completely
sclerosed. Nearly all of them show closure of most of the capillaries with
beginning hyaline degeneration of the lobules. There are many remnants of
polymorphonuclears enclosed in the lobules. A number of epithelial crescents
are to be seen. There is moderate atrophy of all the tubules.

C.^SE 41 (A-21-108).— Male, aged 51 years. First admitted to the hospital
in October, 1919, with an attack of acute articular rheumatism. The heart
was enlarged at this time but the urine was normal. The second admission
was Feb. 26, 1921. At this time he had edema of the arms, legs and back,
dyspnea and epistaxis. Blood pressure, 228/110. Width of heart 19 cm. Urine:
specific gravity, 1.020; large amount of albumin; a few hyalin and granular
casts, a few erythrocytes. Hemoglobin, 80 per cent. Erythrocytes, 4,800.000;
16,950 leukocj'tes — 82 per cent, polymorphonuclears. Blood pressure, March 3,
190/100. Ammoniacal odor to the breath. Fever developed during the last
few days coincident with the appearance of the physical signs of broncho-
pneumonia. Death, March 2, 1921. Necropsy: Marked edema of all dependent
portions ; no ascites ; 500 c.c. of thin purulent fluid in the left pleural cavity.
Heart weighed 475 gm. ; no valvular lesions. Bronchopneumonia. Kidneys
weighed 630 gm. External surfaces smooth. No hyalin glomeruli. Large
numbers of fairly fresh extra capillary lesions. A great many polymorphonu-
clears in the swollen glomeruli. Fairly well marked atrophy of nearly all the
tubules.

Cases 42 to 45 are apparently intermediate between subacute and
well defined chronic cases. The kidneys are not shrunken and there
are not many hyalin glomeruli. The tubular atrophy and the hyalin-
ization of the glomeruli are more pronounced than in the previous
group. The inflammatory exudate (polymorphonuclear.s) appears
fresher in Cases 42 and 43 than in Cases 44 and 45.

C.-lSe 42 (A-16-368).— Male, aged 39 years. Father and one brother died
of renal disease. Severe attack of scarlet fever in childhood. Diphtheria
at 20. Acute arthritis at 25. Present illness began May 30, 1916. with gastric
disturbances and nausea. June 19 he first noticed puffiness of the face, and
shortly afterward a general edema developed. Precordial pain for two or
three weeks before death. Admitted to hospital in extremis. Oct. 26, 1916.
Had generalized edema, ascites, gastric disturbances and precoridal pain. Urea
nitrogen, 104 mg. ; creatinin, 10.2 mg. ; blood sugar, 0.074 per cent. ; hemoglobin,
35 per cent. Phenolsulphonephthalein, 0 per cent. Urine : specific gravity,
1.024; abundant albumin, many casts. Leukocytes. 12,000. Death, Oct. 27, 1916.
Necropsy: Marked edema of lower half of body; ascites (5.000 c.c), left
hydrothorax (3,000 c.c), hydropcricardium (500 c.c). Heart weighed 375 gm.;
kidneys, 280 gm. ; finely granular surfaces. Nearly all the glomeruli show
beginning hyaline degeneration. Very few glomerular capillaries are visible.
Large numbers of partially disintegrated polymorphonuclears are seen in the
closed capillaries and in some of the dilated tubules. There are large num-
bers of old epithelial crescents. There is advanced atrophy of nearly all the
tubules.

Case 43 (.^-19-220).— Male, aged 43 years; was well until one year ago
when he began to have severe headaches and weakness. He was told by his
physician that he had kidney trouble. Admitted to hospital Oct. 2. 1919. com-
plaining of weakness, dyspnea, palpitation of heart, anorexia, precordial pain

BELL-HARTZELL—GLOMERULOXEPHRITIS 801

and constipation. His heart was enlarged. Blood pressure, 140/60; hemo-
globin, 30 per cent.; erythrocyte.-;, 1,300.000; leukocytes, 13,300. Urine: specific
gravity, from 1.012 to 1.014; trace of albumin; many granular casts. Phenol-
sulphonephthalein, October 4, 2 per cent.; October 7, 0 per cent. Blood
chemistry : October 7, creatinin, 13 mg. ; urea nitrogen, 125 mg. ; blood
sugar, 0.13 per cent.; October 22, creatinin, 17 mg. ; urea nitrogen, 155 mg. ;
blood sugar, 0.19 per cent. Eye-grounds negative. Death, Oct. 23, 1919.
Necropsy: No edema; no ascites; hydrothorax (300 c.c). Heart enlarged.
Kidneys weighed 300 gm. ; finely granular surfaces. Some of the glomeruli
are completely sclerosed but the great majority show the hyaline change just
beginning. There are enormous numbers of disintegrated polymorphonuclears
in the interstitial tissues and in the damaged glomeruli (.Fig. 16). There is
advanced atrophy of most of the tubules and moderate atrophy of the others.

C.-vsE 44 (.A-IO-HS).— Male, aged 32 years. Duration unknown. Urine:
abundant albumin, hyalin and granular casts. No other clinical data. Necropsy:
Ascites, hydrothora.x, hydropericardium, edema of the lungs. Heart weighed
530 gm. ; kidneys, 314 gm. Nearly all the glomeruli are of about the same
appearance. They are enlarged, their capillaries are practically all closed by
swollen endothelium, but none of them has yet become hyaline. Occasional
remnants of polymorphonuclears are seen in glomeruli and tubules. There is
very marked tubular atrophy.

Case 45 (A-11-77).— Male, aged 70 years; admitted to hospital June 7, 1911.
Complained of edema and shortness of breath. Urine: albumin, hyalin and
granular casts. Specific gravity, from 1.018 to 1.024. Death, July 18, 1911.
Necropsy : General anasarca, ascites, hydrothora.x and hydropericardium. Heart
weighed 475 gm. ; old mural thrombus in left auricle. Kidneys weighed 335
gm. ; smooth external surfaces. There are a few hyalin glomeruli ; but the
great majority are not so far advanced. There is complete closure of all but
a very few capillaries in each glomerulus. Occasional nuclear fragments are
seen in the glomeruli which may be remnants of polymorphonuclears. There are
many partially disintegrated polymorphonuclears in the tubules. There arc
a large number of old epithelial crescents. Tubular atrophy is very marked.

There is very little clinical history available in Cases 46 and 47,
but the microscopic structure suggests cases of long duration with a
fairly recent inflammation in the persistent glomeruli.

Case 46 (.^-12-40). — Male, aged 33 years. Duration of illness unknown.
Was comatose and breath had a uremic odor. Edema of face. Systolic blood
pressure, from 200 to 210. Diarrhea. No marked changes in eye-grounds.
Moderate amount of albumin and numerous casts. Necropsy : Slight edema
of face and neck. A little fluid in the serous cavities. Heart weighed 735
gm. ; kidneys, 250 gm. ; granular surfaces. A large percentage of the glomeruli
are hyaline. Other glomeruli are enlarged and show closure of most of the
capillaries. Numerous well preserved polymorphonuclears are visible in the
open and closed capillaries, suggesting a recent acute attack.

Case 47 (0-12-83). — Female, aged 17 years. For several years the patient
had symptoms which were referred to the urinary tract. No details of the
clinical picture were recorded. A diagnosis of renal tuberculosis was made
and one kidney was removed Nov. 18, 1912. Subsequent history unknown.
The kidney shows a finely granular surface. A large number of glomeruli
are hyaline; the others are enlarged and show a great increase of nuclei with
closure of most of their capillaries. Large numbers of disintegrating poly-
morphonuclears are to be seen in the closed capillaries (Figs. 12, 13 and 14).
The tubules belonging to the hyaline glomeruli have completely disappeared ; the
others show varying degrees of atrophy.

802 ARCHIVES OF IXTER.WAL MEDICINE

Cases 48 to 51 give histories which may be interpreted as examples
of chronic nephritis following acute nephritis, and in Cases 49 and 51
there is good evidence of a terminal acute exacerbation.

Case 48 (A-15-363). — Male, aged 30 years, has had frequent urination for
past five or six years. Three years ago he suddenly developed general edema.
Was in the hospital three weeks at this time. Left hospital much improved.
Has had attacks of edema and shortness of breath since then. Edema is
worse after alcoholic excesses. Has been confined to his home since about
Sept. 10. 1915. Admitted to hospital Oct. 10, 1915. with a severe attack similar
to the one he had three years ago. Phenolsulphonephthalein, October 16, a
trace. Urine: specific gravity, from 1.010 to 1.014; abundant albumin; many
casts. Leukocytes, 7,400. Death, Nov. 14, 1915. Necropsy : Marked edema
of lower half of body, ascites, hydrothorax and hydropericardium. Heart
weighed 575 gm. ; kidneys, 155 gm. ; granular surfaces. A large majority of the
glomeruli are hyaline and their associated tubules show extreme atrophy; the
others are damaged in varying degrees, due to closure of part of their capil-
laries. Many of the tubules associated with these injured glomeruli show-
very little atrophy.

C.\SE 49 (A-17-62).— Female, aged 27 years. Mother died of Bright's dis-
ease at 49. One Ijrother has kidney trouble now. Patient had otitis media in
1905, and smallpox in 1908. In October, 1909, she developed severe headaches,
high fever, edema of ankles and ascites. Blood pressure during this attack,
160 systolic. She was in bed from October, 1909, to April, 1910. There was
gradual improvement but she had recurring attacks of headache, edema of
ankles and vomiting. Had a mastoid infection in 1915. Scarlet fever in the
fall of 1916. Polyuria the past three years. Eyesight always poor. In October,

1916, she had a sore throat with high fever and the edema became worse.
Vision has gradually grown poorer. Has been in bed since Feb. 24, 1917.
Admitted to hospital March 1, 1917. Urine: specific gravity, from l.OOS to 1.008;
large quantities of albumin, many leukocytes, a few granular casts. Hemoglobin,
40 per cent.; erythrocytes, 2,300,000; leukocytes, 11,000 (March 3); 22,400
(March 13) — ^92 per cent, polymorphonuclears. Blood pressure, March 1, 185/100.
Temperature about normal. Blood chemistry: creatinin, March 1, 26 mg. ;
March 16, 22.2 mg. Urea nitrogen, March 1, 101.2 mg.; March 16, 68 mg.
Phenolsulphonephthalein, 0 per cent. March 2 and March 10. Death, March 17,

1917. Necropsy : Marked edema of face, ascites, hydrothorax and hydroperi-
cardium. Heart weighed 455 gm. ; kidneys, 150 gm.; rough granular surfaces.
A large majority of the glomeruli are completely sclerosed and their tubules
have almost disappeared; the others are enlarged and show closure of most of
their capillaries, and partial atrophy of their associated tubules. In this latter
group of glomeruli there are many polymorphonuclear leukocytes in both the
open and closed capillaries. These histologic features suggest an acute exacer-
bation of a chronic nephritis. There are a few old extracapillary lesions.

Case 50 (A-18-117).— Male, aged 32 years. No history of scarlet fever or
rheumatism. Ten years ago he was in a hospital with a condition diagnosed
Bright's disease by his physician. He remembers that he had a considerable
amount of edema at that time. His present illness began about the end of
March, 1918, with weakness, loss of appetite, vomiting after eating and unpro-
ductive cough, .\dmitted to hospital May 21, 1918. Patient thinks he has lost
al)0Ut 20 pounds during the last two months. Physical Examination : Enlarged
heart, systolic murmur at apex, dyspnea, slight edema of lungs. Blood pressure,
190/140. Daily urine excretion from 175 to 225 c.c. Moderate albuminuria, casts;
many erythrocytes. Phenolsulphonephthalein, 0 per cent, on two occasions. Urea
nitrogen, from 15.5 to 17 mg. ; creatinin, from 2 to 3.6 mg. Blood sugar, from
0.092 to 0.11 per cent. Death, June 3, 1918. Necropsy: No edema; ascites (100

BELL-HA R TZELL—GL OMER UL OXEPHRITIS 803

c.c.) ; hydrothorax (1,300 ex.). Fibrinous pericarditis. Heart weighed 630 gin.;
spleen, 240 gm. ; kidneys, 405 gm. ; granular surfaces. About one half of the
glomeruli are completely sclerosed, and their tubules have almost disappeared.
The others are enlarged and show closure of most of their capillaries but no
definite hyaline changes. Their tubules are notably decreased in size. .^ number
of these glomeruli show hyaline degeneration of the afferent arteriole. Remnants
of polymorphonuclears are visible in most of the enlarged glomeruli and
throughout the interstitial tissues. There are many epithelial crescents. It is
probable that the sclerosed glomeruli are due to the attack ten years ago.

Case 51 (.■\-19-264).— Male, aged 29 years, had scarlet fever at 23. Was
unable to work for si.x months following this disease (no information as to
edema or albuminuria). Had an attack of rheumatism at 27 which lasted six
months. At 28 he had arthritis in the ankles and wrists, relieved by tonsillec-
tomy in the spring of 1919. In October, 1919, he contracted a severe cold from
exposure in a cold rain. The ne.xt day he was in bed with severe headache,
palpitation, dyspnea, precordial pain, edema around the eyes, epistaxis and
blurring of vision. He has grown worse gradually since that time, .'\dmitted
to hospital, Nov. 6, 1919. Examination showed enlargement of heart with a
systolic murmur, pufiiness of face and conjunctival hemorrhages. Blood
pressure, 170/80. During his stay in the hospital he had frequent attacks of
vomiting, nosebleed, hiccough, headache and dizziness. Became unconscious
December 8. Hemoglobin, from 32 to 25 per cent. Erythrocytes from 2,650,000
to 1,800,000. No fever. Urine: large amount of albumin, hyalin and granular
casts, a few erythrocytes. Phenolsulphonephthalein, November 14, a trace ;
November 21, 5 per cent. Blood chemistry: November 9, creatinin, 7.4 mg. ;
urea nitrogen, 87 mg. ; blood sugar, 0.153 per cent.; November 20, creatinin,
10.8 mg. ; urea nitrogen, 36 mg. ; blood sugar, 0.189 per cent. Alkaline reserve,
November 20, 40.6. Death, Dec. 8, 1919.

Necropsy: Slight edema of face; hydropericardium (100 c.c). Heart
weighed 435 gm. ; vegetations on wall of left auricle. Spleen weighed 365 gm. ;
kidneys, 340 gm. ; finely granular surfaces. A number of glomeruli show
complete sclerosis with disappearance of their tubules. A large number show
old extracapillary lesions without hyalinization of the tuft but with well
advanced atrophy of their tubules. More than one third show an appearance
corresponding with the subacute stage, i. e., great enlargement due to swelling
and proliferation of the endothelium, with many polymorphonuclears in the
closed capillary loops. The tubules belonging to this latter group show moderate
to well advanced atrophy. There are masses of polymorphonuclears in some
of the tubules. There are large areas of cortical tissue densely infiltrated with
mononuclear leukocytes— an appearance characteristic of the acute interstitial
nephritis of scarlet fever (Fig. 18). The appearances suggest an old nephritis
dating from the attack of scarlet fever, and an acute attack of rather recent
date.

Case 52 is a good illustration of terminal acute glomerulitis in a case
of very long duration.

Case 52 (A-17-207).— Female, aged 46 years, has had a discharge from the
left ear as long as she can remember. Frequent attacks of sore throat and
pleurisy. Operation for myoma of uterus when 27 years old. Present trouble
began in 1900 with nocturia and gastric disturbances. Some improvement on
a milk .diet. Was refused life insurance in 1906 because of albuminuria. Was
under treatment for this condition from time to time but did not improve. In
bed, on a milk diet, for eight days in .'\ugust, 1915. Admitted to hospital,
May, 1917. Poor vision. Cramps in muscles. Urine: trace of albumin; casts
very' rare. Hemoglobin, 49 per cent.; leukocytes, 11,000. Phenolsulphone-
phthalein, a trace. Blood pressure, 117/80. Urea nitrogen, from 39 to 66 mg.;

804 ARCHIVES OF IXTERXAL MEDICIXE

creatinin, from 9.7 to 16.8 mg. ; blood sugar, from 0.16 to 0.2 per cent. Dis-
charged June 23. Readmitted, Sept. 29, 1917. Headache, nausea, vomiting,
dyspnea, swelling of face and legs, loss of appetite, cramplike pains in fingers
and frequent urination. Eleven hundred c.c. of fluid was drawn from right
pleural cavity. October 3, urine: moderate amount of albumin, occasional
hyalin casts. Hemoglobin, 25 per cent.; erythrocytes, 3,000,000; leukocytes, 7,600.
Blood pressure, 160/110. Phenolsulphonephthalein, a trace. Urea nitrogen,
89 mg. ; creatinin 16.2 mg. Temperature normal. Death, Oct. 6, 1917. Necropsy ;
general anasarca: ascites (1,500 c.c); hydrothorax (800 c.c). Heart weighed
400 gm. ; spleen, 120 gm. Pus in left middle ear. Slight general arteriosclerosis.
Kidneys, weighed 120 gm. : rough, granular surfaces. .\ large percentage of
the glomeruli are completely sclerosed ; the others show an acute or subacute
glomerulitis. These latter are enlarged with greatly swollen endothelium, which
closes most of the capillaries. The numerous polymorphonuclears in the capil-
laries and tubules as well as fresh epithelial crescents indicate that there has
been a recent acute attack.

Cases 53 to 59 are clinically examples of the type with slow insidious
onset and long duration. Microscopically they show very old lesions
without any acute processes.

C.«iSE 53 (A-11-76).— Female, aged 44 years. .Admitted to hospital .\ug. 9,
1910. Complained of edema which bad been present at varying intervals since
her first pregnancy nearly fifteen years before. Had scarlet fever when a
child. Has had headaches and vomiting spells for many years. Shortness of
breath and swelling of feet for the past three years. .A.lbumin and casts con-
tinuously present in the urine. Hemoglobin, from 50 to 60 per cent. Systolic
blood pressure, 140 to 170. Extreme edema. Eighty-one gallons of fluid was
drawn from the pleural and peritoneal cavities during her stay in the hospital.
Multiple abscesses of the skin of the arms and legs appeared a few months
before death. Streptococci were found in these abscesses. Low fever during
most of her stay in the hospital. Death, July 17, 1911. Necropsy: .Anasarca,
ascites, hydrothora.x and hydropericardium. Heart weighed 492 gm. Fresh
mural thrombus in left ventricle. Kidneys weighed 260 gm. Granular surfaces.
About one-third of the glomeruli are completely hyalinized and their tubules
have largely disappeared. Most of the other glomeruli are associated with
markedly atrophic tubules and show closure of nearly all the capillaries with
beginning hyaline degeneration (Fig. 19). A few glomeruli show a number
of permeable capillaries, and these glomeruli belong to tubules of normal size.
Renal function was performed by a number of damaged glomeruli.

C.\SE 54 (A-13-178).— Female, aged 32 years. Many attacks of tonsillitis
when young; none recently. Present illness began about one year ago with
shortness of breath on slight exertion. In June, 1913, she first noticed swelling
of legs which would disappear after a night's rest. Was in bed sixteen weeks
before admission to hospital. On admission, Nov. 21, 1913. there was general
subcutaneous edema, enlargement of heart, and hypertrophied tonsils. Urine:
abundant albumin, many hyalin granular casts. Leukocytes, November 22,
13,700; 80 per cent, polymorphonuclears. Blood pressure, November 24, 160/110;
December 8, 110/100. Death, Dec. 9, 1913. Necropsy: General anasarca,
hydrothorax, hydropericardium, ascites and edema of lungs. Heart weighed
543 gm. Double pyosalpinx. Streptococcus from heart's blood. Kidneys
weighed 190 gm. Granular surfaces. Large majority of glomeruli sclerosed.
Other glomeruli show partial closure of their capillaries with a corresponding
atrophy of the associated tubules.

C.\si-: 55 (A-13-184).— Female, aged 34 years. Duration of illness about
10 months. Hemoglobin 30 per cent. No clinical history available. One

BELL-HARTZELL—GLOMERULOXEFHRITIS 805

kidney removed at necropsy ; granular surface. Many sclerosed glomeruli. No
normal glomeruli are found. The great majority show severe injury and are
connected with markedly atrophic tuhules. These injured glomeruli are enlarged
and nearly all their capillaries are closed (Fig. 9"). There are a number of
polymorphonuclear remnants in the closed capillaries. A number of old
epithelial crescents are seen.

Case 56 (A-17-108).— Male, aged 39 years. Has had scarlet fever, date
unknown. Present illness began three years ago with swelling of ankles. Later
his legs and face began to swell. Still later he developed dyspnea and palpitation
of the heart. Admitted to hospital. April 18, 1917. Had edema of ankles,
headache, poor vision. Urine: specific gravity, 1.014; albumin, trace to a large
amount. Leukocytes, 11,800. Phenolsulphonephthalein, 0 per cent. May 2.
Death, May 18, 1917. Necropsy: No edema; no ascites; hydrothorax (600 c.c.) ;
hydropericardium (200 c.c). Heart weighed 640 gm. : kidneys, 350 gm. ; surfaces
roughened and pitted. Large numbers of hyalin glomeruli are seen ; the other
glomeruli are all enlarged, most of their capillaries are closed and there are
varying degrees of hyaline degeneration in the lobules (Figs. 15 and 20).
Numerous degenerated polymorphonuclears are seen in the closed capillaries
and in atrophic tubules. There are a few rather recent epithelial crescents.
An occasional afferent glomerular artery shows hyaline degeneration.

Case 57 (A-17-118).— Male, aged 69 years. One brother died of Bright's
disease. Patient first noticed swelling of his feet, especially at night, thirteen
years ago. For the last few years he has noticed puffiness under the eyes.
He has been in a hospital on two previous occasions for the same condition
that troubles him now. Admitted to hospital, April 18, 1917. Has severe
dyspnea. Left leg is greatly swollen. Blood pressure, April 27, 185/140. Urine:
trace of albumin, many casts : specific gravity from 1.012 to 1.022. Leukocytes,
9,600. Phenolsulphonephthalein, April 27, 12 per cent. Blood chemistry, April
27: creatinin, 1.5 mg. ; urea nitrogen, 28 mg. Alkaline reserve, 46. Death,
June 8, 1917. Necropsy: Marked edema of lower extremities; left hydrothorax
(100 c.c.) ; edema of lungs. Heart weighed 495 gm. ; extensive sclerosis of aorta
and large vessels. Kidneys weighed 160 gm. ; rough granular surfaces. A large
number of glomeruli are completely sclerosed; many others are enlarged with
closure of most of their capillaries. There are a number of old extracapillary
lesions. Many of the afferent glomerular arteries show hyaline degeneration.
and the obliteration of some of the glomeruli is evidently due to this cause.

Case 58 (A-18-94).— Male, aged 50 years. Never had scarlet fever, rheu-
matism or tonsillitis. Well until his present illness began six or seven years
ago, when he first noticed swelling above the tops of his shoes, especially in
the evening. This condition grew worse and at times there was also swelling
of the hands. For the past five years he had noticed weakness and dyspnea,
most pronounced during the past few weeks. Admitted to hospital, April 7,
1918. Physical Examination: Loss of weight; enlargement of heart to the left;
systolic murmur at aortic area and at apex. Urine: specific gravity from 1.006
to 1.015; trace to a heavy precipitate of albumin; numerous granular casts.
Low fever. Hemoglobin, 75 per cent. Erythrocytes, 4,420,000; leukocytes, 7,800.
Blood pressure, April 11, 188/124. Death, May 3. Necropsy: Edema of face;
no fluid in serous cavities. Heart weighed 640 gm.; kidneys, 205 gm. ; finely
granular surfaces. A large majority of the glomeruli are completely sclerosed;
the others are enlarged and show closure of most of their capillaries.

Case 59 (A-19-2).— Male, aged 25 years. No history of scarlet fever, tonsil-
litis or rheumatism. In good health until April 1, 1918, when he began to have
attacks of headache and vomiting. The attacks would last a day, after which he
would be able to work again. Has had a few attacks of precordial pain, one
of which lasted eight hours. Shortness of breath was first noticed about the

806 ARCHIVES OF IXTERXAL MEDICI. \'E

middle of November, 1918. He has been in hospitals on four different occasions
since April 1, 1918. Admitted to University Hospital Dec. 14, 1918. No edema.
Mitral regurgitation. Blood pressure, December 15, 210/118. Urine: large
amount of albumin; many granular casts; specific gravity, 1.010. Leukoctyes,
December 19, 10,500. Low fever. Blood chemistry : December 27, urea nitrogen,
57.3 mg. ; creatinin, 12 mg. ; January 4, urea nitrogen, 72 mg. ; creatinin, 12.4
mg. Phenolsulphonephthalein, December 20, a trace. Albuminuric retinitis.
Death, Jan. 5, 1919. Necropsy: No edema; ascites (150 c.c.) ; hydrothorax
(,400 c.c.) ; early bronchopneumonia. Heart weighed 500 gm. ; kidneys, 195 gm. ;
rough granular surfaces. Over three fourths of the glomeruli are completely
sclerosed and their tubules have almost disappeared. The remainmg glomeruli
are badly damaged, only a small part of their capillaries being permeablp.
Their tubules are moderately atrophic. An occasional old extracapillary lesion
is seen. There are no polymorphonuclears in the glomeruli.

Cases 60 to 68 are cases with short chnical histories but with gross
and microscopic evidence of very long duration.

Case 60 (A-13-145).— Male, aged 35 years, has had diphtheria and scarlet
fever. Present illness began three months before admission with dyspnea and
swelling of feet. Admitted to hospital Sept. 10, 1913. Chief complaints were
gastric distress and edema. Heart was enlarged. Urine: specific gravity, 1.010;
abundant albumin ; hyalin, granular, waxy and leuckocyte casts. Systolic blood
pressure, September 27, 132. Died in coma, Sept. 29, 1913. Necropsy: Ascites
and hydrothorax. Heart weighed 475 gm. ; kidneys, 169 gm. ; surfaces very
granular. Apparently about two thirds of the glomeruli are hyalinized and their
tubules have disappeared. The remaining glomeruli are enlarged, and a large
percentage of their capillaries are occluded (Fig. 21). In some of these, there
are remnants of polymorphonuclears in the closed capillaries. The tubules
connecting with these defective glomeruli are moderately atrophic. There are
large areas of lymphocytes in the cortex, probably representing an interstitial
exudate which occurred during the attack of scarlet fever. This suggests that
the glomerular involvement may date from the attack of scarlet fever.

C.^SE 61 (A-14-192).— Male, aged 30 years. Admission, Sept. 28, 1914. No
clinical history. Urine: specific gravity from 1.008 to 1.018; abundant albumin,
many casts. Blood pressure, October 7, 192 systolic. Death, Oct. 13, 1914.
Necropsy: Slight edema of extremities; ascites. Heart weighed 490 gm.;
kidneys, 360 gm. ; granular surfaces. A small proportion of the glomeruli are
completely sclerosed ; the majority show extensive obliteration of their capil-
laries with marked atrophy of their tubules. A few glomeruli are normal.
Rarely an old extracapillary lesion is seen. .\ great many disintegrating
polymorphonuclears are found in the atrophic tubules.

C.\SE 62 (A-15-373).— Male, aged 23 years, admitted to hospital Sept. 26, 1915,
has had frequent attacks of sore throat and rheumatism (?). About one month
before admission he noticed slight swelling of his feet, and gradually increasing
weakness and shortness of breath. Kept at work until about Sept. 12, 1915. On
admission he complained of weakness, dizziness, headache, cough and nocturia.
Urine (nine examinations) : specific gravity, from 1.010 to 1.012: albumin, trace
to a heavy precipitate; many casts. Blood pressure. October 13, 180/115. Phenol-
sulphonephthalein, October 1, 11 per cent. Leukocytes, November IS. 5,800.
Death, Nov. 23, 1915. Necropsy: No edema ; no fluid in serous cavities. Heart
weighed 557 gm. Kidneys small with granular surfaces. A large percentage of
the glomeruli are completely hyalinized, and the others show partial closure of
their capillaries.

Case 63 (A-15-394).— Male, aged 54 years, admitted to hospital, Oct. 19,
1915. Seven years ago he had an attack of acute arthritis (ankles were swollen,

BELL-HARTZEL L—GLOMER ULOXEPHRITIS 807

red and painful). Xine days before admission he began to have dyspnea and
five days later his legs began to swell. Abdomen has been distended since about
October 5. On admission he had marked edema of lower extremities and
external genitals, and enlargement of the heart with mitral regurgitation. Urine
(sixteen examinations): albumin, trace to a heavy precipitate; casts. Blood
pressure, October 19, 150/SO; November 4, 168/90; December 6, 100/58. Death,
Dec. 7, 1915. Necropsy: No edema; hydropericardium. Heart weighed 650
gm. Mitral leaflets greatly thickened and retracted; aortic leaflets adherent.
Lobar pneumonia. Advanced sclerosis of aorta. Kidneys weighed 319 gm.;
roughened granular surfaces. A large percentage of the glomeruli are com-
pletely sclerosed ; the others show partial closure of their capillaries. There is
advanced sclerosis of some of the larger arteries and a few of the afferent
arterioles. The fact that most of the afferent arterioles are not diseased indicates
that the vascular disease is not the main cause of the glomerular changes.

Case 64 (A-16-132). — Male, aged 27 years. About three months before
admission he began to have attacks of nosebleed. These recurred frequently
and have gradually become more severe. About seven weeks before admission
his left wrist pained him severely. Entered hopsital, March 26, 1916, com-
plaining of frequent epistaxis, pain in stomach with occasional vomiting, sore
throat and general weakness. Small hemorrhages were noted on the left wrist
and on the palate. There was edema of the uvula and pharynx, and a mitral
systolic murmur. Temperature, from 100 to 101 F. Hemoglobin, 36 per cent. ;
erythrocytes, 2.800,000 ; leukocytes. 9,200; 87 per cent, polymorphonuclears. Urine:
large amount of albumin ; some blood. Death, March 30, 1916. Necropsy : No
edema; ascites (SO c.c.) ; hydrothorax (100 c.c). Aorta, normal. Heart weighed
425 gm. ; kidneys, 150 gm. ; roughened granular surfaces. More than three
fourths of the glomeruli are completely sclerosed, and the others show many
closed capillaries. Occasional partially atrophied tubules shows masses of
disintegrating polymorphonuclears. A few afferent arterioles show hyaline
degeneration.

Case 65 (A-16-384).— Male, aged 24 years. One brother died of Bright's
disease at 23. Patient stated that he had been perfectly well until August,
1916, when he had an attack of nosebleed. At first the attacks came about once
a week, and would last one to two hours, but later, they became more
frequent. He gradually became weaker. Did not notice swelling of feet or
puffiness of eyes until about November 5. Admitted to hosi>ital. Nov. 12,
1916. Urine: specific gravity, 1.010; large amount of albumin; many casts of all
kinds. Hemoglobin, 20 per cent.; erythrocytes, 1,500,000; leukocytes, 7,500.
Blood pressure, 170/90. Phenolsulphonephthalein, November 12, 0 per cent.
Blood chemistry, November 13: creatinin, 27.2 mg. ; urea nitrogen, 131 mg. ;
sugar, 0.26 per cent. ; alkaline reserve, 18.3. Death, November 18, 1916. Necropsy :
No edema; no ascites; hydrothorax (200 c.c). Heart weighed 415 gm. Aorta
normal. Kidneys weighed 205 gm. ; roughened granular surfaces. Practically
all the glomeruli are severely involved. There is advanced atrophy of all the
tubules. Many atrophic tubules contain masses of disintegrating polymorpho-
nuclears. There are some old extracapillary lesions.

Case 66 (A-19-160).— Male, aged 32 years. Recently discharged from the
army because of heart and kidney trouble. No details of history available.
Admitted to hospital July 26, 1919. Normal temperature. Frequent vomiting.
Became irrational July 27. Blood pressure, 120/80. Systolic murmur at apex.
Dilated heart. Urine: large amount of albumin; no casts; no erythrocytes.
Death, July 30, 1919. Necropsy: No edema; no fluid in serous cavities.
Heart weighed 350 gm.; dilated. Kidneys weighed 140 gm. ; finely granular
surfaces; thinned cortices. A large majority of the glomeruli are completely
sclerosed, and most of the others are severely damaged and their tubules are
markedly atrophic. .'Kn occasional glomerulus is practically normal. There are
a large number of old extracapillary lesions.

808 ARCHIVES OF IXTERXAL MEDICI XE

Case 67 ( A-18-180).— Male, aged 26 years, admitted to hospital, Sept. 17,
1918. Death, Sept. 18, 1918. No history was obtained. Urine : specific gravity,
I.OIS; small amount of albumin; many granular casts. Necropsy: Edema of
ankles ; marked ascites, hydrothorax and hydropericardium. Heart weighed
360 gm. ; spleen, 235 gm. ; kidneys, 290 gm. ; external surfaces smooth. About
one-third of the glomeruli are completely sclerosed and their tubules are very
small. The other glomeruli are moderately enlarged, most of their capillaries
are closed and their tubules are moderately reduced in size. There are
numerous disintegrated polymorphonuclears in the closed capillaries and some
of the tubules. Many glomeruli show a thick band of hyaline around the
afferent artery after its entrance into the glomerulus ; but there is very little
change in the artery before it reaches the glomerulus.

Case 68 (A-20-204).— Male, aged 34 years. Typhoid fever at age of 17
years ; was in bed 10 weeks. In March, 1920, he first noticed impairment of
vision. At this time he had frequent headaches with a catarrhal condition in
the upper respiratory tract. His physician told him that he had nephritis.
May 4, 1920, examination revealed enlargement of heart, albuminuric retinitis,
and a blood pressure of 200/120. Death, May 20, 1920. Necropsy: Slight
edema of face; ascites (100 c.c.) : hydrothorax (2,200 c.c.) ; edema of lungs.
Heart weighed 425 gm. ; kidneys, 195 gm. ; granular surfaces. Many sclerosed
glomeruli with disappearance of their tubules. Majority of glomeruli are
enlarged and their capillaries are largely occluded by swollen endothelium but
there is very little hyaline degeneration. There is a notable proliferation of
the capsular epithelium. The tubules show advanced atrophy.

Case 69 (A-17-230).— Male, aged 45 years. Well until the autumn of 1916
when he developed swelling of the feet and abdomen with marked weakness.
Improved under hospital treatment but was too weak to return to work. In the
summer of 1917 he had another similar attack. There was marked general edema
and a systolic murmur at the apex of the heart. Admitted to hospital Sept. 29,
1917. Blood pressure, October 1, 140/70; October 20, 130/80. Urine, Sep-
tember 29 to November 6; abundant albumin, many casts of all types. Leuko-
cytes, October 15, 8,700. Phenolsulphonephthalein, October 1, 35 per cent. ;
October 15, 31 per cent. Blood chemistry, October 2 : urea nitrogen, 24.2 mg. ;
creatinin, 2.3 mg. November 1, left side of face became swollen and six teeth
were extracted. Lobar pneumonia was recognized November 9. Leukocytes
26,000 on November 9. Death, Nov. 11, 1917. Necropsy: No edema; no
fluid in serous cavities. Extensive old pleuritic adhesions. Lobar pneumonia.
Heart weighed 400 gm. ; spleen, 120 gm.; kidneys, 360 gm.; adherent capsules;
slightly roughened external surfaces ; grayish yellow cortices. Very few
glomeruli are hyaline. The great majority of the corpuscles show old
epithelial crescents surrounding small partially collapsed glomeruli with few
or no permeable capillaries. The tubules connected with these glomeruli are
markedly atrophic but still easily visible (Fig. 22). There are a number of
glomeruli, however, that have many permeable capillaries and a're coiuiected
with fairly normal tubules. These presumably explain the fairly good renal
function.

Age. — The age of the subaciiie and chronic cases at the time of
death is shown by decades in Table 2. It will be noted that two-thirds
of the cases occurred in the third and fourth decades. Fitz ^' at the
Massachusetts General Hospital, found an avera<^e age of 32 years
in cases of chronic glomerulonrpliritis, and 52 years in the arlerio-

31. Fitz, R. : The Phenolsulphonephthalein Test and the Nonprotein
Nitrogen of the Blood in Chronic Nephritis, Boston M. & S. J. 183:247, 1920.

BELL-HARTZELL—GLOMERULOXEPHRITIS 809

sclerotic form of Bright's disease. Two-thirds of our cases of arterio-
sclerotic renal disease were over 50 years old.

T.\BLE 2.— .Alge at De.mh
•^ge Number of Cases

0-10 years 1

11-20 years 1

21-30 years 12

31-40 years 12

51-50 years 5

51-60 vears 4

61-70 years 2

Sex. — There were twenty-seven males and nine females, but the
proportion of males to females in our necropsies is approximately 2 : 1
(2,325 males and 1,063 females in 3,388 necropsies) so that there is
probably only a slight preponderance of the disease in males.

Duration. — The duration of the subacute cases was usually less
than four months; but one patient (Case 36) lived six months, and
another (Case 40) had an abnormal urine one year before death.

The actual duration of a chronic case cannot be determined entirely
by the patient's statement as to when his illness began. Seven well
defined chronic cases with contracted kidneys gave histories of an
illness of only two or three months' duration. Probably the disease
could have been recognized long before in these patients, but it evi-
dently did not produce serious discomfort until well advanced.

Twelve patients had symptoms of nephritis for one year or more
before death, and in seven of these the disease was present over three
years. Case 49 lasted eight years and Case 52 over eleven years.

Symptoms. — The disease usually develops slowly, but sometimes the
onset is rapid, even in cases in which extensive destruction of renal
tissue must have been present long before the appearance of symp-
toms. The most frequent initial complaints are dyspnea and edema ;
but in occasional cases epistaxis or gastrointestinal disturbances, such
an anorexia, vomiting and gastric distress may bring the patient to
his physician. Other symptoms which may be prominent at some time
during the course of the disease are headache, weakness, precordial
pain, frequent urination and disturbances of vision.

Edema. — This is a very common sign of nephritis. Subcutaneous
edema was present in thirty-one of thirty-six cases in which data
were available ; and even in the five dry cases there was a little fluid
found in the serous cavities at necropsy. The intensity of the edema
varies greatly in different patients and in the same patient from time tc
time. It is influenced by rest, the amount of salt in the diet and othei
unknown factors. It may disappear before death. The accumulation
of fluid in the serous cavities is sometimes a prominent and serious

810 ARCHIVES OF INTERXAL MEDICIXE

feature of the disease. Eighty-one gallons of fluid were drained from
the pleural and peritoneal cavities of one patient (Case 53) during
a period of eleven months. Hydrothorax and edema of the lungs are
partly responsible for dyspnea, which is a very frequent symptom.

Hypertension and Cardiac Hypertrophy. — These are almost con-
stantly present in chronic glomerulonephritis. Blood pressure was
recorded in twenty of the chronic cases, and in only two instances was
the systolic reading below 140. These low pressures were both taken
shortly before death and may not be exceptional since it frequently
happens that the pressure falls a short time before exitus. The systolic
pressure was from 160 to 180 in six patients, and 180 or above in ten.

The average weight of the heart in twenty-six chronic cases was
498 gm. Only three hearts weighed less than 400 gm. One of these
(Case 42) was a case of comparatively short duration, judged both
by the clinical history and the structural changes in the kidneys. In
the other two cases (Cases 66 and 67) no clinical history was available,
but the histologic picture indicated a disease of long duration, so that
these seem to be examples of chronic glomerulonephritis without car-
diac hypertrophy. There is no notable enlargement of the heart in the
subacute cases, but hypertension was found in the three cases in which
the blood pressure was taken.

Anemia. — This is apparently a frequent complication in advanced
chronic glomerulonephritis and it seems to increase in intensity toward
the end of the illness. The hemoglobin was determined in ten chronic
cases, and found to be below 50 per cent, in eight of them. One
patient (Case 58) had a hemoglobin of 75 per cent, one month before
death.

Urimiry Changes. — Urinary examinations are recorded in thirty-
four of the thirty-seven cases of this group. Albumin was found in
every sample of urine from every patient. Usually a large amount was
found but occasionally only a trace was present. When a large num-
ber of examinations are made on the urine of one patient the amount
of albumin may vary from a trace in some samples to a heavy
precipitate in others. We have no cases of glomerulonephritis with
normal urine. Casts were found in thirty of thirty-four cases, and
usually in large numbers. In the cases in which no casts were found
only one sample of urine was examined. Casts may be rare in one
sample and numerous in another sample of urine from the same patient.
Bloody urine was recorded in one chronic and three subacute cases. Of
course, only a small percentage of persons who have albumin and casts
in the urine are suffering from Bright's disease. Brown " in a study

32. Brown, P. K. : A Study of the Etiology of Chornic Nephritis, J. A.
M. A. 66:793 (March 11) 1916.

BELL-HARTZELL—GLOMERULOXEPHRITIS 811

of 7,000 hospital admissions found 594 with albumin and casts, of
which only thirty-eight (6.4 per cent.) were considered to have Bright's
disease.

Functional Tests. — Functional studies were made on sixteen patients.
The tests were applied five months before death in two patients, and
within the last two months of life in all the others. The phenolsul-
phonephthalein output was invariably greatly reduced, and in eleven
patients in which it was determined shortly before death it varied
from 0 to 20 per cent. In Case 69 the readings were higher but this
patient died of lobar pneumonia before serious renal insufficiency had
developed.

Marked retention of urea nitrogen and creatinin was usually
observed. The exceptions are Case 36, in which the test was made very
early in the attack, and Case 50, which is inconsistent with the
phthalein excretion and probably an error. In Case 69, in which death
was due to lobar pneumonia there was only slight retention. As is
well known urea nitrogen begins to increase before there is any change
in the amount of creatinin.

These tests usually run parallel and are of about equal value in
estimating renal sufficiency. It is desirable to use both tests since
each acts as a control on the other and both are entirely consistent
with the structural changes found at necropsy. We have not seen any
cases of marked inconsistency between these functional tests and the
findings at necropsy. Functional tests are of no great value in acute
glomerulonephritis since in these kidneys the damage is usually not
permanent ; but in chronic nephritis they give a fairly accurate mea-
surement of the extent of the permanent injury. The difficulty with
these tests is that they do not reveal the presence of nephritis until
two-thirds or more of the renal filter has been closed, and they are
therefore of no value in the recognition of the early stages of the
disease when there might be some chance for successful therapy.

Functional tests are frequently useful in distinguishing a case of
primary hypertension from one of chronic Bright's disease, and also in
determining the extent of the renal injury in the arteriosclerotic kidney ;
but it is not often possible to distinguish this last named renal disease
from chronic glomerulonephritis by functional tests. Occasionally
patients may live a long time with a high degree of retention of
metabolites in the blood and a very low phthalein excretion
(O'Hare^-), but, as a rule, such findings indicate that death will occur
within a few months at most. In making a prognosis it is to be remem-
bered that in acute nephritis and in acute exacerbations of a chronic

33. O'Hare. J. P.: Compatibilitv of Long Life with Low Renal Function,
J. A. M. A. 73:248 (July 26) 1919.

812 ARCHIVES OF IXTERXAL MEDICIXE

nephritis there is a temporary obstruction of many capillaries which
open up again as the acute process subsides. Tests made during these
acute stages indicate a more extensive destruction of renal tissue than
actually exists.

Our data are not complete enough to warrant a statement as to
the frequency of albuminuric retinitis.

Death in subacute and chronic glomerulonephritis seems to occur
almost invariably as a direct result of renal insufficiency. A severe
anemia is usually present in the terminal stages. Cardiac decompensa-
tion seems never to be the direct cause of death. Occasionally a
seropurulent pleuritis or a bronchopneumonia of limited extent is
found at necropsy. In only one case was death definitely due to a
complication, viz., lobar pneumonia in Case 69.

Gross Changes in the Kidneys. — In subacute glomerulonephritis
the kidneys may be of normal size but are usually enlarged. They
are never contracted. In live out of eight adults the kidneys weighed
more than 400 gm. In one instance the weight was 500 gm. and in
another 630 gm. The external surfaces are smooth. The cortices
are cloudy.

In the twenty-five chronic cases in which the weights are recorded
the average was 248 gm. In ten cases the kidneys weighed less than
200 gm., in eight cases more than 300 gm., and in three cases more
than 350 gm. The maximum weight was 405 gm. In a general way
there is a relation between the duration of the disease and the size of
the kidneys, the smaller kidneys being found in cases with histories
of a long illness: but there are many exceptions. The external sur-
faces are almost invariably roughened and granular, a condition pro-
duced by an uneven atrophy of the cortex. The cortex is cloudy,
reduced in thickness, and indistinctly marked ofif from the pyramids.
Multiple small superficial cysts and adenomas do not seem to occur
in uncomplicated glomerulonephritis but these structures are very
common in arteriosclerotic kidneys.

Microscopic Appearances. — In chronic cases there are always a
large number of hyalin glomeruli. These are smaller than normal and
arc of homogeneous structure. Their tubules may appear as small
solid cords of cells, or they may have disappeared entirely. Roughly
estimated, the number of hyalin glomeruli varies from one-third in
some cases to three-fourths in others. These tubule systems arc of
course nonfunctional. The other glomeruli arc practically all injured
in varying degrees. Very few are entirely normal. The most
important change is the permanent closure of various parts of the
capillary network. Some glomeruli show closed portions here and
there wliilc others mav show only an occasional permeable cajiillary

BELL-HARTZELL—GLOMERULOXEPHRITIS 813

(Fig. 23). The size of the associated tubule depends on the permea-
bihty of its glomerulus. The tubular atrophy is pronounced when the
capillaries are mainly closed. The work of the kidney is performed
by damaged glomeruli (Figs. 17, 19, 23).

In the subacute kidneys the changes are not so far advanced.
The glomeruli show closure of the capillaries by swollen endothelium
with beginning hyaline degeneration, but very few glomeruli have
become completely hyalinized. The subacute case differs from the
chronic in the more uniform injury of all the glomeruli. Death in
the subacute case occurs before any glomeruli have had time to reach
the hyaline stage. In the chronic case the severely injured glomeruli
become hyalin, while those less injured carry on the renal function.

The evidences of previous acute processes in chronic kidneys will
be discussed later on.

ETIOLOGY OF SUB.'^CUTE .\.\-D CHRONIC GLOMERULONEPHRITIS

The clinical and pathologic evidence that acute glomerulonephritis
is due to bacterial infection is so convincing that this view has very
few opponents at present ; but there are many who do not believe that
infection is the basis of the chronic form, and either attribute it to
obscure metabolic disturbances or consider the etiology entirely
unknown. The evidence as to the etiology of subacute and chronic
glomerulonephritis will now be discussed.

1. The Relation Between Acute and Chronic Glomerulonephritis.— h
is a common observation that acute glomerulonephritis does not often
pass into the chronic form. Volhard and Fahr state that the majority
of acute cases recover in days, weeks or months according to their
severity, and that only the very severe die. Of their seventy-one
acute cases three became chronic. One of their patients developed
acute nephritis in 1903, fourteen days after an attack of tonsillitis. In
1906 edema of the legs developed and never completely disappeared.
In 1910 he had a typical chronic nephritis. Death in 1913. Duration,
ten years.

Aufrecht'* gave a very thorough description of a case of twenty
years' duration which began acutely six months after an attack of
scarlet fever. Albumin was found in the urine almost constantly
throughout the illness. The patient had seven attacks of hematuria,
one or more years apart, usually following inflammations of the throat
During the last' six months the .symptoms were very
necropsy contracted kidneys and an enlarged heart were found

severe. At

M Aufrechf Eine zwanzig Tahrc dauernde Xephritis nacli Scharlach.
Deutsch. Arch. klin. Med. 42:517, 1888.

814 ARCHIVES OF ISTERXAL MEDICIXE

Mann ^^ described a very similar case of twenty-eight years' duration,
which was under his observation the entire period. An acute nephritis
began in 1866 at the age of 14 years, following an attack of scarlet
fever. The acute symptoms subsided in a few weeks and the patient
regained her health, except that albumin continued to be present in the
urine. For seven or eight years after the original attack the patient
had occasional subacute attacks some of which were accompanied by
edema. Attacks usually followed exposure to cold and usually lasted
two or three weeks. From 1880 on the urine became of lower specific
gravity and increased in amount. Arterial tension gradually rose.
Retinal hemorrhages and uremia preceded death in 1894. At necropsy
the heart was found hypertrophied. The kidneys were contracted,
each weighing about 2 ounces.

Lohlein had a case in which the acute attack occurred in 1894 at
the age of 16 years. The patient spent 1897-99 in the army. In 1901
he developed symptoms of contracted kidneys and died with anuria,
severe edema and uremic symptoms. Lohlein believes that a chronic
nephritis "not very rarely" develops from an acute nephritis of scar-
let fever or angina.

Eichhorst ^® reported two cases of intermittent albuminuria of
several years' duration which followed scarlet fever. The patients
finally recovered. Another of his patients developed a hemorrhagic
nephritis on the fourth day of an acute tonsillitis. He still had blood,
casts and albumin in the urine two years later.

Sorensen and Volhard and Fahr believe that scarlet fever is rarely
responsible for a chronic nephritis. Reichel says "numerous observa-
tions show that a chronic nephritis may develop from a scarlatinal
nephritis."

Ernberg ^' located forty of 106 adults who had had acute nephritis
before the age of 15 years, and found normal urine in all. In sixteen
of fifty adults who had had acute nephritis between the ages of 15
and 30 years he found normal urines in all but four.

Reports on soldiers who have had acute nephritis (war nephritis)
show that many of them have not recovered completely (Robinson,^^
Patterson^"). There are indications that some of these men will
develop chronic nephritis.

35. Mann. J. D. : On Granular Kidneys Following Scarlatinal Nephritis,
Lancet 2:670, 1895.

36. Eichhorst, H. : Leber clironische intermittierende .-Mhumcnurie als
Nachkrankheit infektioser Nephritiden, Med. Klin., 1919.

37. Ernberg (cited from Hill) : Boston M. & S. ]. 177:313, 1917.

38. Robinson, A. R. : The .\fter History of War Nephritis, J. Roy. Army
Med. Corps 30:205, 1918.

39. Patterson, D. W. : British M. J. 2:431, 1921.

BELL-HARTZELL—GLOMERULOXEPHRITIS 815

Our Cases 48 to 51 are apparently examples of chronic glomerulo-
nephritis developing from acute cases. It is possible that in Case 48
the sudden onset of symptoms was due to a severe exacerbation of a
chronic nephritis and not to a primary acute attack. In Case 51 there
is abundant microscopic evidence of an old attack of acute interstitial
nephritis such as occurs in scarlet fever, and this observation is strong
support for the assumption that his weakness after scarlet fever was
due to nephritis. Cases 50 and 51 seem to be clean cut cases.

The evidence accumulated shows that a chronic nephritis may
develop from a typical acute case but that such instances are rare.
The great majority of acute cases do not terminate in chronic di.sease,
and the great majority of chronic cases do not begin as clinical acute
nephritis. But this does not exclude the origin of chronic nephritis
from a "clinically veiled" acute case, as Lohlein has expressed it. It
is conceivable and even probable that an infection of slow development
may occur without producing intense clinical symptoms. If edema
failed to develop there would be a striking change in the clinical
picture.

The clearly established fact that an occasional case of chronic
nephritis does develop from an acute case seems to prove that at least
a few cases of chronic nephritis are due to infection, since the infec-
tious nature of the acute type can hardly be denied.

2. The Blending of the Different Clinical Types of Glomerulo-
nephritis.— It is maintained by some observers that chronic nephritis
has very little clinical resemblance to the acute type and might there-
fore have a different etiology. But such a view encounters difficulties
when a large series of cases is studied. Senator was impressed with
the fact that the nephritides form a continuous group with numerous
intermediate cases between the acute and chronic types. The recogni-
tion of a subacute type by most observers is a further support of
this view.

Nine of our cases (Cases 2iZ to 41) have been classed as subacute,
but some of these might be considered acute by other observers.
Cases 42 to 45 show structural changes somewhat more advanced than
the subacute group but much less advanced than the typical chronic
case. Case 43 was of at least one year's duration. This group (Cases
33 to 45) seems to us to illustrate gradual transitions between acute
and chronic nephritis.

3. Associated Infections. — In the subacute group there are .several
cases with evidence of a preliminary infection. Case ZZ began with
acute pericarditis, traces of which were found at necropsy. Case 38
followed a cold and bronchitis which was apparently due to wetting
of the feet. (M healed endocardial lesions were found in Case ii.
In Case 35 there was jaundice early in the disease and a sero])urulent

816 ARCHIVES OF IXTERXAL MEDICIXE

pleuritis was found at necropsy. In Case 36 symptoms of nephritis
appeared three days after an injection of neo-diarsenol. There was a
severe attack of arthritis in Case 41 four months before death, and
thrombosed pelvic veins were found in Case Z7. Ophiils found in most
of his subacute cases a comparatively recent history of tonsillitis,
rheumatism or other streptococcic infection.

In the chronic group there were sixteen cases without history of
a primary infection and without associated old infections at necropsy.
The clinical data in some of these were, however, incomplete. In the
other cases there was some evidence of a primary infection. Old
healed mural or valvular lesions were found twice (Cases 45 and 63).
One patient (Case 52) had a chronic purulent otitis media of many
years' duration, as well as frequent attacks of tonsillitis. Two other
patients (Cases 54 and 62) gave a history of frequent attacks of
tonsillitis. A number of the patients had had scarlet fever, and micro-
scopic evidence of scarlatinal interstitial nephritis was found twice
(Cases 51 and 60). The high incidence of tonsillitis and scarlet fever
among persons who do not develop nephritis of course detracts from
the value of our data. Two patients had had acute articular rheu-
matism (Cases 42 and 63). Four cases apparently followed acute
nephritis.

On the whole, it may be said that the evidence of associated strepto-
coccic infections is very convincing in the acute group, fairly good in
the subacute, but not at all satisfactory in the chronic group. It does
not seem to us that these data exclude the primary infection in the
chronic case. The clinical data are often incomplete, the patient having
forgotten things that happened several years before; or the primary
infection may have been comparatively mild and ignored by the patient.
The primary infection may heal in a short time. Even in acute cases
following tonsillitis, the throat is usually about normal when the renal
symptoms appear.

4. Microscopic Evidence. — The strongest argument for the infec-
tious origin of chronic glomerulonephritis is obtained from microscopic
study of the kidneys. Every abnormal glomerulus in a chronic kidney
is readily explainable as the outcome of one or more acute inflam-
matory processes. The various forms of acute glomerulonephritis are
recognizable in their healing stages in the chronic kidney, and sometimes
even acute glomerular lesions are seen along with the chronic changes.

The proliferative form of glomerulitis is the most important since
it causes permanent closure of the capillaries. When there is com-
plete closure of all the capillaries in the acute stage (Fig. 5) the tubule
atrophies rapidly and the glomerulus soon becomes small and hyaline.
If, however, some of the capillaries are not closed the glomerulus
continues to function to some extent and undergoes hyaline degenera-

BELL-HARTZELL—GLOMERULOXEPHRITIS 817

tion only in the clo.sed portions (Figs. 19 and 17). In chronic nephritis
practically all the glomeruli are damaged by closure of part of the
capillary network.

The epithelial crescents which characterize the extracapillary type
of glomerulitis are easily recognized in chronic nephritis. When first
formed the crescents are composed of cells with distinct nuclei and
well stained cytoplasm (Fig. 8). On long standing they gradually
become homogeneous and hyaline. All intermediate stages are recog-
nizable (Fig. 15). Extracapillary lesions were numerous in five of
the subacute cases. In the chronic group old crescents were found in
twelve of twenty-eight cases. They were numerous in four, frequent
in three, and rare in five cases. Epithelial crescents have been pro-
duced experimentally by bacterial inoculations (Pernice and Scag-
liosi*"). Their presence in chronic nephritis is a strong argument for
the infectious nature of this disease.

Exudative glomerulitis is characterized by polymorphonuclears in
the capillaries, and when this is combined with the proliferative type
the leukocytes become imprisoned in the closed capillaries (Figs. 12,
13 and 14). Frequently masses of polymorphonuclears are retained
for an indefinite time in the lumina of tubules or in the interstital
tissues. This gives another form of evidence by which we may prove
that a previous acute infection was present.

Polymorphonuclears in closed glomerular capillaries were found in
six subacute "and thirteen chronic cases, sometimes in large numbers.
Masses of pus cells were found in the tubules in six chronic cases.
There were only seven chronic cases that had no epithelial crescents
and no pus cells in glomeruli or tubules.

The histologic evidence is convincing that the glomerular lesions
in chronic kidneys are the healing stages of acute lesions, and if one
admits that acute glomerulitis is due to infection it is difficult to escape
the conclusion that chronic glomerulonephritis is due to the same cause.

Progrcsske Character of Chronic Glomerulonephritis. — It is known
that patients with chronic nephritis may live many years. A duration
of ten years is frequently seen and Mann's patient lived twenty-eight
years. During the course of the disease there may be long intervals
when the patient is in good condition with only a trace of albumin in
the urine and fairly efficient kidneys as shown by functional tests.
But at irregular intervals acute exacerbations occur, characterized by
the usual symptoms of renal insufficiency. The patient improves after
many of these acute attacks but finally develops a permanent renal
insufficiency which soon ends in death. These exacerbations are well

40. Pernice and Scagliosi ; Bcilrag zur .^etioIogie dcr Nephritis, etc., Vir-
chows Arch. f. path. .Nnat. 138:521. 1894.

818 ARCHIVES OF IXTERXAL MEDICI. XE

illustrated in the excellent case records published by Mann and
Aufrecht. They seemed to be due usually to exposure to cold in
Mann's patient and to throat infection in Aufrecht's. Emerson *'^ has
recently emphasized the importance of these acute exacerbations in
causing progressive destruction of the kidneys. He attributes these
exacerbations to infections, chilling of skin, errors in diet, too much
exercise, etc., and believes that chronic nephritis is not necessarily a
progressive disease.

Our clinical data are not complete enough to warrant an extensive
analysis of this topic, but there are records of acute exacerbations in
eight patients (Cases 48, 49, 51. 52, 53, 57, 59 and 69). In Cases 49
and 51 the exacerbations followed attacks of sore throat.

In Case 51, in addition to the chronic changes, a number of fairly
recent glomerular lesions were found, which were in all probability
due to the severe throat infection which developed two months before
death. In Case 52 a chronic purulent otitis media was present. The
kidneys of this patient showed abundant evidence of a recent infection,
viz., acute glomerulitis, fresh purulent exudate, etc. In these two
patients at least there is convincing evidence that fresh infection con-
tributed to the destruction of the renal tissue. Lohlein mentions rare
instances in which acute glomerulitis is found in chronic kidneys.

Both the clinical and the pathologic evidence, therefore, seem to
support the view that the downward progress of chronic glomerulo-
nephritis is largely due to recurring infection. These infections are
probably not often focal in character, but it seems a reasonable
inference that known foci of infection, such as diseased tonsils,
abscessed teeth, etc., should receive attention. An eflfort should be
made to protect the chronic nephritic against infection, but diseased
glomeruli are abnormally sensitive to bacterial toxins and glomerular
injury will probably occur no matter how careful the patient may be.
Throat infections and exposure to cold are to be especially avoided.
The influence of diet and excessive muscular effort in causing exacer-
bations is not well established.

In explaining the downward tendency of chronic nephritis after
renal insufficiency has developed another factor .should be considered.
Dr. Hilding C. Anderson,*^ in some experiments carried on in this
laboratory, has found that a permanent renal insufficiency may be
produced by the surgical removal of about three-fourths of the renal
tissue. In the course of a few weeks the kidney remnant undergoes
degeneration which is not explainable on the basis of obstruction or
infection and seems to be due to excessive functional strain. If these

41. Emerson, C. P.: The Acute Element in tlic Chronic Xcphropathies,
J. A. M. A. 77:745 (Sept. 3) 1921.

42. This paper is being prepared for publication.

BELL-HARTZELL—GLOMERVLOXEPHRITIS 819

conclusions be correct they may be applied to the final stages of chronic
nephritis in man.

Atrophy of the Renal Tubule. — Lohlein's interpretation of tubular
atrophy as a disuse atrophy is in accord with our observations. From
a study of serial sections it has been observed that the size of the
tubule corresponds closely to the permeability of the glomerulus and
that the tubule disappears when the glomerulus becomes completely
hyalinized. It has been claimed that the atrophy of the tubule is due
to inadequate blood supply because of the closure of the glomerular
vessels ; but one frequently sees atrophic tubules surrounded by well
filled capillaries and there are free anastomoses between the capillaries
of adjacent tubules. Besides, Dehoff " has shown that the terminal
branches of the interlobular arteries pass directly to the capillary net-
work of the tubules. Tubular atrophy occurs as readily at the surface
of the kidney as elsewhere.

The evidence that chronic glomerulonephritis is due to infection may
be summarized briefly as follows:

There is abundant clinical and pathologic evidence for the infectious
origin of the acute type and good evidence for a similar origin of many
subacute cases. When this is admitted there is no escape from the
conclusion that the chronic type has a similar cause.

In our series there is a gradual transition from the acute through
the subacute to the chronic type, and it is obvious that we are not
dealing with diff'erent diseases. On clinical grounds alone a funda-
mental relationship between the different forms must be admitted.

An occasional acute case becomes chronic and it must be conceded
that at least these chronic cases are due to infection. The .structural
changes in these kidneys are not different from those of other chronic
cases.

The glomerular lesions in chronic kidneys are obviously healing
and healed stages of acute glomerulitis. The old epithelial crescents and
the closed capillary loops filled with disintegrated polymorphonuclears
cannot be explained on any other basis.

Definite acute and subacute glomerular lesions are sometimes found
in chronic kidneys. These cannot be satisfactorily explained except
as exacerbations of an inflammatory process.

SUMMARY

Thirty-two cases of acute glomerulonephritis have been studied. In
many of these cases death was due to extrarenal causes and early
glomerular lesions are available for study.

43. Dehoff, E. : Die arteriellen Zuflusse (les Capillarsystcms in de
rinde des Menschen, Virchows Arch. f. path. .Anat. 228:134, 1920.

820 ARCHIVES OF IXTERXAL MEDICISE

Degenerative, exudative and proliferative types of inflammation
occur in the glomeruli. Proliferative changes are chiefly responsible
for permanent glomerular damage.

Acute glomerulonephritis is nearly always due to some acute infec-
tious process, usually a streptococcal infection. The bacteria gain
access to the blood and it is probable that the injury is produced by
the direct action of their bodies on the glomerular endothelium.

An occasional case of acute glomerulonephritis passes into the
chronic form; but the great majority of chronic cases do not begin
as frankly acute nephritis.

Acute glomerulonephritis is linked with the chronic form by
numerous intermediate cases.

Glomerular lesions in chronic kidneys correspond to healing or
healed stages of acute glomerulitis. Old epithelial crescents are common,
and disintegrating polymorphonuclear leukocytes are frequently found
in the closed glomerular capillaries and in the partially atrophied
tubules. In a few chronic cases acute and subacute glomerular lesions
were found, indicating acute exacerbations.

In chronic glomerulonephritis many glomeruli are obliterated com-
pletely and those persisting show permanent closure of a part of the
capillary network. Function is carried on by damaged glomeruli, and
is depressed not only because of reduction in the total number but
also because of the reduced capillary network in those that persist.

The progressive nature of chronic glomerulonephritis is apparently
due, in part, to repeated acute exacerbations.

All forms of glomerulonephritis are due directly to bacterial invasion
of the glomeruli ; and the various clinical and pathologic types depend
on the degree and extent of the permanent glomerular injury.

BIOCHEMICAL STUDIES IX A FATAL CASE OF
METHYL ALCOHOL POISONING*

I. .M. RABIXOVITCH, M.D.

MONTREAL

It is not sufficiently appreciated that methly alcohol is very toxic.
For economic reasons, methyl (wood) alcohol is employed as a
substitute for ethyl (grain) alcohol, in what may be termed compara-
tively innocent products, such as perfumes, hair tonics, skin lotions,
polishes, varnishes, etc. Government analysts, not infrequently, find
it employed in the manufacture of various extracts for the flavoring
of food products. Since prohibition has come into force, pure methyl
alcohol, being somewhat similar in odor and taste to ethyl alcohol, has
been employed in the preparation of various alcoholic beverages. This
in great part is due to ignorance arid has resulted in many deaths.

The literature, both experimental and clinical, on this subject shows
a preponderance of papers relative to the effects of this drug on the
central nervous system, especially the brain and optic nerves, and little
reference has been made to lesions elsewhere in the body. More
recently the importance of an acidosis as the cause of the more general
signs and symptoms in methyl alcohol poisoning has been emphasized.

Little consideration has. however, been given to the changes which
may occur in the kidney and other functions which may be mani-
fested by variations from the normal chemical composition of the
blood. In the following case special attention was given to this.

REPORT OF CASE

History ( Hosp. No. 4937/21).— .\ female, aged 70 years, was admitted to
the medical wards of the Montreal General Hospital, into the service of
Dr. H. A. Lafleur, with a history of having taken, with suicidal intent,
Oct. 24, 1921, one drinking-glassful of wood alcohol. There was a history
of vomiting prior to admission, but no vomiting occurred while she was in
the hospital, a period of five days. On admission, the patient was drowsy
and very much confused, so that it was not possible to obtain a detailed
history.

Physical Examination.— The patient was a white female apparently of the
stated age. There was slight cyanosis. The right pupil was larger than the
left. Both reacted to light and accommodation. The tongue protruded in
the midline with a slight tremor at the edges. There was a slight sweetish
(acetone) odor to the breath. Physical examination otherwise was negative,
with the exception of the fundi oculi. There was a slight effusion into the
retina.

Course.— From the time of admission the patient grew progressively weaker.
The respirations, at first of the Kussmaul (acidosis) type, became very

* From the Department of Metabolism of the Montreal General Hospital.

822

ARCHIVES OF JXTERXAL MEDICI XE

shallow during the last two days. October 29 the cyanosis became more
marked and there was clinical evidence of bronchopneumonia.

The patient was incontinent throughout her stay in the hospital, and it
was not possible to obtain a specimen of urine at proper periods for chemical
analysis. This was unfortunate as no examination could be made of the
excretion of organic acids. A single specimen obtained by the house physician
at the time of admission showed a clear urine, specific gravity 1.018, albumin
7.8 gm. per liter, many hyalin and granular casts, a few blood cells, and
a trace of acetone. The accompanying table shows the combined results
of the blood examination made every twelve hours during the patient's
illness.

Chemical Analysis of the Blood

Venous Oxygen

Date
Oct. 26, p.m.
Oct. 27, a.m. .

p.m. .
Oct "8, a m. .

Acid

3.1

5.3

7.6

8.4

92

Oft. 29, a.m. .
p.m. t

y.o

3:30) 9.3

DISCUSSION

It has long ago been demonstrated ' that the difference in the
character and degree of intoxication between ethyl and methyl alcohol
is due to the fate of these substances following their administration.
Ethyl alcohol is oxidized into easily excreted products, carbon dioxid
and water. Methyl alcohol is, however, only partially oxidized. The
products of this incomplete oxidation being meth or formaldehyd and
formic acid.

H H H H

H-O-C-H + 0 = C = 0 + HjO and C = 0 + 0 = C = 0

i A A o'h

methyl meth or Formic acid

alcohol formaldehyd

These partially oxidized substances are very toxic. It has been
found - that formic acid is six times as toxic, and formaldehyd ' is
thirty-three times as toxic as methyl alcohol. Thus, from the oxid-
ation of a toxic substance, products may result which are many times
more toxic. This has an important bearing in the interpretation of
the blood analysis.

1. Hunt, R.: The Toxicitv of Methvl Alcohol, Johns Hopkins Hosp. Bull.
13:2, 213, 1902.

2. Gettler. A. O., and George, A. V. St.: Wood .Mcohol Poisoning, J. A.
M. A. 70:145 (Jan. 19) 1918.

RABIXOVITCH— METHYL ALCOHOL POISOXIXG 823

KIDNEY FUNCTION

In experimental work with methyl alcohol changes in the kidney
have been noted. Tyson and Schoenberg ^ in work on dogs f oimd
at necropsy dark purple and congested kidneys. Gettler ^ in one case,
found marked parenchymatous degeneration of the kidneys. An
analysis of the chemical findings of the blood in our case with ref-
erence to the urea nitrogen, uric acid, creatinin and phosphorus shows
that rapid changes occurred in the kidney function. The uric acid
content increased from 3.1 to 9.3 mg. per hundred c.c. in less than
six days. The urea nitrogen content increased from 42 to 144 mg.
per hundred c.c, and the creatinin content from 1.6 to 4.5 mg. per
hundred c.c. blood in the same period. The acid soluble phosphorus
varied from 8 to 11 mg. per hundred c.c. blood calculated as phos-
phorus (P). So far as we know only one other case has been studied
from this viewpoint, that of Harrop and Benedict.-* Their findings
differ entirely from ours. These authors found that the urea con-
tent of the blood was normal at one period, but remarkably low
at another, 0.091 gm. per liter. This would correspond to 4.2 mg.
urea nitrogen per hundred c.c. blood, which is exceedingly low. The
blood phosphorus in their case was normal (3 mg. per hundred c.c.
blood). Their patient recovered.

The initial high findings in our case may have been due to a
previously existant chronic nephritis. There can, however, be no
doubt that the rapid changes noted daily were due to the action of
the poison. Such findings suggest a complete "renal block," and
correspond to those occasionally found in the acute retention as seen
in hypertrophy of the prostate, or the anuria of mercuric chlorid poison-
ing. That the kidney function was practically nil is also supported by
the rapid increase in the uric acid and creatinin content of the blood.
The patient took no food during these few days of illness. It may,
therefore, be assumed that all the uric acid found was of an endogenous
origin. If it is assumed that the average daily excretion (endogenous)
of uric acid is between 100 and 200 mg., and that this amount is not
excreted but is distributed throughout the blood, it will account for
the daily increase noted. The anatomic findings appear to corroborate
this view.

BLOOD SUGAR

Hyperglycemia was present throughout the course of the disease.
The lowest concentration of sugar, found at the first examination, was

3. Tyson, H. H., and Schoenberg, M. H.: Experimental Researches in
Methyl Alcohol Inhalation, J. A. M. A. 63:915 (Sept. 12) 1914.

4. Harrop, G. A., and Benedict, E. M.: Acute Methyl Alcohol Poisoning
Associated with Acidosis, J. A. M. A. 74:1 (Jan. 3) 1920.

824 ARCHIVES OF IXTERXAL MEDICIXE

0.182 per cent. This gradually increased to 0.228 per cent. These
findings seem difficult to interpret. Apparently, they can be attributed
to the impairment in the kidney function, for such findings are not
infrequent in advanced cases of chronic nephritis. It might, however,
be assumed that the figures do not represent glucose, for the reason
that on theoretical grounds methyl alcohol is oxidized to formaldehyd,
and the latter is a reducing agent. Part of the reduction of the cupric
oxid .in the test might therefore be attributed to the presence of this
agent. Virtually, however, it does not seem that this occurred. The
studies of Denis and Aldrich,'^ who employed liquor formaldehyd
for the preservation of blood specimens, show that the addition of
this drug in certain amounts does not alter the results obtained in
blood sugar estimation. Even if we assume that all the methyl alcohol
taken by this patient was completely oxidized to formaldehyd and
distributed throughout the body, its concentration in the blood would
not reach the percentage that these authors found could be added to
blood without interfering with the chemical estimation of sugar. An
interesting observation of these same authors is that liquor formaldehyd
prevents glycolysis for at least ninety-six hours in vitro. That this
should occur in vivo is only conjectural. The impairment of the
kidney function seems sufficient to account for the hyperglycemia
noted.

ACIDOSIS

The plasma carbcjn dioxid combining power on admission was 46
\olumes per cent. It eventually fell to 26 volumes per cent. Other
factors which may lower the carbon dioxid combining power of the
bk)od, such as increased pulmonary ventilation having beei) excluded,
it may be assumed that an acidosis existed. An acidosis has previously
been demonstrated * in the study of methyl alcohol poisoning." It
has. however, been attributed to the failure of the body to com-
pletely oxidize methyl, alcohol with the production of formic acid.
The acidosis has been found " to be associated with an increase in the
excretion of organic acids, lactic and formic. In our case the reten-
tion of phosphates in the blood would also explain part of the acido.sis.

It does not appear to be unreasonable, on theoretical grounds,
to suggest that the acidosis may. in large part, be due to the forma-
tion in the body of methylene derivatives, from the action of the
formaldehyd on the amino-acids present. These derivatives are more

5. Denis, W.. and .Mdrich, M. : Note on the Preservation of Specimen of
Blood Intended for Blood Sugar Estimation. J. Biol. Chem. 44:203 (Oct.) 1920.

6. Haskell. C. C; Hileman. S. P., and Gardner. W. R. : The Significance
of the .Acidosis of Methyl .Alcohol Poisoning. Arch. Int. Med. 27:71 (Jan.)
1921.

7. Van Slyke. D. D. : Studies of Acidosis. J. Riol. Chem. 41:.%7 (April)
1920.

RABISOIITCH— METHYL ALCOHOL POISOSISC 825

strongly acid in reaction owing to the destruction of the basic prop-
erties of the amino group, and, therefore, should effect the acid base
equilibrium of the blood. Such a reaction is readily demonstrated in
vitro. The Henriques-Sorensen formol titration of amino-acid nitrogen
is based on this principle * as is shown in the following equation :

K. CH. Ml; R— CH— N:CH:

COOH -6C1LO ~ COOH *H:0

Also the production of free acids from the action of liquor for-
maldehyd on neutral ammonium salts in the body does not seem
unreasonable and may explain part of the acidosis, as is shown in
the following equation :

4 XH.Cl + 6 CH;0 = \.(CH;), + 6 11,0 + 4 HCl

An interesting observation along these lines is that of Gregnolo
who found after the injection of methyl alcohol there was an increase
in the hydrogen ion concentration of the serum.

From the time of admission to the hospital the patient exhibited a
definite cyanosis. This was very slight at first, but became more
marked during the progress of the disease. Very little reference to
biochemical studies could be found in the literature on the relation
between methyl alcohol poisoning and cyanosis, although this relation
has frequently been noted clinically. An effort was made to determine
the cause in our case. An analysis was made of the oxygen content,
oxygen capacity and oxygen unsaturation of the blood.

It might be recalled that the oxygen content represents the total
oxygen combined with hemoglobin, and otherwise, circulating in the
blood at the moment and site of withdrawal. The oxygen capacity
represents the total oxygen the blood could hold if it were completely
saturated with o.xygen. The oxygen unsaturation^ represents the
difference between the oxygen capacity and content. The method
employed for the estimation of the oxygen was that of Van Slyke.'"
It has been shown" that if the blood is completely saturated with
oxygen in the lungs, the oxygen unsaturation of the venous blood may
increase from 13 to 14 volumes per cent, before cyanosis appears. If
it appears at less than this figure arterial unsaturation may be assumed.
It will be noted in the chart that at the first estimation the oxygen

8. Hawk: Practical Physiological Chemistry, Ed. 6. Philadelphia, P.
Blakistons Sons Co., p. 526.

9. Lundsgaard, C: Studies of Oxygen in Venous Blood, J. Biol. Chem.
33:133, 1918.

10 Van Slyke, D. D. : Gasometric Determination of the Oxygen and Hacmo-
glohin of Blood. J. Biol. Chem. 33:127. 1918.

11 Lundsgaard, C. : Studies on Cyanosis, J. Exper. M. 30:271 (Sept.) 1919.

826 ARCHIVES OF ISTERXAL MEDICIXE

unsaturation was 9 volumes per cent. Assuming, therefore, that
arterial unsaturation may have existed, the cause of this under the
circumstances (poisoning) was problematic.

An analysis of the daily clinical notes showed that no gross changes
occurred in the respiratory or circulatory systems. Although no gross
clinical changes need be evident, and still certain conditions may exist
which prevent complete oxidation of the blood, it seemed important
to determine whether any chemical alteration had occurred preventing
the blood from taking up its normal load of oxygen. Stadie ^- pointed
out that there are many substances which in vitro readily produce
methemoglobin. These include certain oxidizing agents, reducing
agents, organic bases, salts and bacteria. It will be noted in the chart
that during the first examination of the blood for methemoglobin
none could be found. The method employed was that of Stadie. ^^

The cyanosis gradually became more marked, and on the following
day definite changes (bronchopneumonia) were found clinically in the
lungs. The blood examination one half hour before death, at which
period there was a very marked degree of cyanosis, also did not show
the presence of methemoglobin. At this time the oxygen capacity
was practically normal. It may, therefore, be assumed that in the
case studied either no methemoglobin was formed, or that it was
eliminated as rapidly as it was formed and played no important part
in the production of the cyanosis.

REPORT OF POSTMORTEM EXAMINATION

Pathologic Report. — Acute parencliymatous nephritis ; cloudy swell-
ing of the heart and liver ; bronchopneumonia.

DETECTION OF METHYL ALCOHOL IN THE BODY TISSUES

Experimentally it has been found that when methyl alcohol is
given per rectum it is excreted by the stomach. It has also been
demonstrated that the stomach may excrete methyl alcohol unchanged
for a considerable time. It has been assumed - that the alcohol has
a selective action for brain tissue. The brain in every one of six
cases analyzed by this author was found to contain this alcohol. For
these reasons both the brain and stomach tissues were analyzed in
our case.

After a critical study of the fifty-eight methods proposed for
detecting methyl alcohol, Gettler " classified them in order of tlieir

12. Stadie, W. C. : Studies on Blood Changes in Pneumonoccic Infections,
J. Exper. M. 33:627 (May) 1921.

13. Stadie, W. C. : A Method for the Determination of Methaenioglobin
in the Blood, J. Biol. Chem. 41:237 (Feb.) 1920.

14. Gettler. A. O. : Critical Study of Methods for the Detection of Methyl
Alcohol, J. Biol. Chem. 42:311, 1920.

RABIXOriTCH— METHYL ALCOHOL POISO.XLXG 827

efficiency. Those accepted as reliable, extremely sensitive, and involv-
ing little technical difficulty were employed. The method employed in
this case is based on the oxidation of the methyl alcohol into for-
maldehyd and the detection of the latter by various color reactions.
Potassium bichromate and concentrated sulphuric acid was used as
the oxidizing agent. In detail the method was as follows :

Method. — In order to preserve the stomach contents the stomach
was tied off at the cardiac and pyloric ends and removed in toto This
was then passed through a meat grinder and minced to a fine pulp.
This pulp was then placed in an 800 c.c. Kjeldahl flask to which was
added 400 c.c. water and sufficient concentrated sulphuric acid until
a distinct acid reaction was obtained. This was then distilled and
200 c.c. of the distillate was neutralized to phenolphthalein with tenth
normal sodium hydroxid and acidified with 5 c.c. of concentrated
sulphuric acid, cooled, and 0.1 gm. potassium bichromate added and
dissolved. This was then redistilled. To this final distillate the various
color tests were applied.

1. To 3 c.c. distillate was added 5 c.c. concentrated sulphuric acid.
This was cooled. The addition of a few milligrams morphin sulphate
yielded a violet color. Test positive.

2. Test 1 was repeated with the morphin replaced by apomorphin.
A violet color resulted. Test positive.

3. To 3 c.c. of the distillate were added two drops of a 2 per cent,
solution of phenol. This was stratified on a layer of concentrated
sulphuric acid. A red ring was noted at the junction of the two fluids.
Test positive. Methyl alcohol was thus detected in the body tissues
six days after its ingestion.

SUMMARY

In a fatal case of methyl alcohol poisoning changes had occurred
in the renal and other functions as evidenced by variations from the
normal chemical composition of the blood. These, in themselves, dis-
regarding other well known factors, may account for the actual cause
of death. Methemoglobin played no imjwrtant part in the production
of the cyanosis noted. Methyl alcohol could be detected in the tissues
examined six days after the ingestion of the drug.

REVERSED RHYTHM OF THE HEART*
MORRIS H. KAHN, M.A., M.D.

NHW YORK

I purpose, in this communication, to emphasize the distinction that
should be made between the terms "reversed mechanism" and "reversed
rhythm" of the heart.

The term "reversed mechanism" should be confined to express a
reversal of the mechanism of impulse formation and conduction.
Normally, the impulse originates in the sinus and traverses the auricles
to be conducted through the auriculoventricular junctional tissues into
both ventricles. In reversal of the mechanism, the impulse first origi-
nates in the auriculoventricular node, or the junctional tissues, or at
a lower level in the heart and travels backward into the auricular
musculature. Thus the term "reversed mechanism" implies a direction
of stimulus conduction opposite to the normal.

Illustrative of such cases may be mentioned instances of auriculo-
ventricular rhythm. In these cases, a ventriculo-auricular (R-P)
interval may be exhibited with inversion or distortion of the auricular
(P) wave. Williams and James, ^ Heard and Strauss," White,^ and
Robinson and Draper ^ have recorded cases of this type. Cohen and
Eraser ° reported a case in which auricular contractions, represented
by an inverted P wave, resulted from the mechanical stimulus received
from the contracting ventricles. Wilson " reported a case of ventricular
extrasystoles transmitted to the auricles.

From the experimental side, a great deal of work has been done
on the subject by Lewis and his collaborators,' Eyster and Meek,**
Meakins," Wilson '" and others.

The term "reversed rhythm," on the other hand, does not attempt to
interpret the direction of stimulus conduction. It expresses only a
time relationship between the ventricular and auricular contractions,
i. e., the ventricular beat immediately precedes the auricular beat.

* From the Department of Cardiovascular Diseases, Beth Israel Hospital.

1. Williams, H. B., and James, H. : Heart 5:109, 1913.

2. Heard, J. D., and Strauss, A. E.: Am. J. M. Sc. 155:238, 1918.

3. White, P. D.: Arch. Int. Med. 16:517 (Nov.) 1915.

4. Robinson, G. C, and Draper. G.: Heart 4:97, 1912.

5. Cohen, A. E., and Eraser, F. R. : Heart 5:141, 1913.

6. Wilson. F. N.: Heart 6:17, 1915.

7. Lewis, T. : The Mechanism and Graphic Registration of the Heart Beat,
Chapters XV and XIX, 1920, Paul B. Hocber, New York; Lewis, White and
Meakins: Heart 5:289, 1913; Lewis and White: Heart 5:335, 1913.

8. Evster and Meek: Heart 5:227, 119. 191.?.

9. Meakins, J.: Heart 5:281, 1913.

10. Wilson, F. N.: Arch. Int. Med. 16:989 (Dec.) 1915.

(

KAHX— REVERSED HEART RHYTHM 829

This includes not only the cases of reversed mechanism, but also those
cases in wliich the mechanism is normal, yet in which the ventricular
contraction is delayed and, therefore, is grouped together with the
following auricular contraction. In these cases, the auriculoventricular
node discharges its impulse, in point of time, ahead of the sinus node,
yet without reversal of the mechanism.

The case reported by Xorrie and Bastedo " and entitled by thein
"reversed rhythm of the heart" is probably not, as they believed, a case
of reversed mechanism. It is a case of heart block due to digitalis
in which the beat of the auricle followed that of the ventricle. Their
explanation that in their case, "the impulse to beat arises in the ventricle
instead of at Keith's node and is conducted along the auriculoventric-
ular bundle in a direction the reverse of normal" leaves room for
question. Their published polygraphic curves might be interpreted
as delayed P-R conduction, the auricular waves being related to the
following ventricle. Thus their case is one of reversed rhythm, but
probably not of reversed mechanism.

Lewis " described a case of premature beats arising in the a-v
tissues in which the P wave was upright and of normal contour. From
this fact, together with the varying P-R intervals, he concluded that
the auricles were responding to the normal pacemaker while the
ventricles were responding to an impulse arising low in the auriculo-
ventricular bundle.

In Cushny's ^^ earlier observation that reversed mechanism can be
produced by injecting aconitin into the circulation of dogs, no electro-
cardiograms were recorded. In two instances, Cushny expressed the
possibility that reversal of the rhythm may have occurred without
reversal of the mechanism.

I desire now to give the observations in a case that presents periods
of reversed rhjthni without reversal of the mechanism and various
features of interest associated with transitory partial heart block.

REPORT OF C.A.SE
H. T., Russian, aged 46, gave a negative family history.
Past History.— He had measles and scarlet fever in childhood and occasional
attacks of tonsillitis since. When he was 36 years old, he passed minute
urinary calculi. The last attack occurred when he was 39 years old. II is
habits were good and he denied venereal disease.

Present Iltiiess.—Bh present illness began when he was about 40 with a
feeling of "heartburn" after meals which was relieved by sodium bicarbonate.
For s'ix months he had slight precordial pain which was worse with psychic
depression. He had hunger pains during sleep and a pressing sensation in the

11 Norrie V H , and Bastedo, W. A.: St. Luke's Hosp. M. & S. Rep. No. 2

12. Lewis.T,. and -■Mien: Am. J. M. Sc. 145:667, 1913.

13. Cushny: Heart 1:1, 1S09.

830 ARCHIVES OF INTERNAL MEDICINE

chest. At times, he had a feeUng of choking and a feeling of dizziness lasting
fifteen minutes together with a tingling sensation, as he described it, in the
precordium and below the angle of the left scapula.

Physical Examination.— He had large follicular tonsils. His heart showed
no enlargement, and no murmurs or accentuations were audible. The muscular
quality of the first sound was deficient at the apex. The apex beat and the

— - -tvf=::r--*:— ■

'^^^''%

i

P-P

1

.76

.76

.76

.76

f-A"- V Y" Y ^

1 R-R

1 .74

.7^

.76

.76

.76

Fig. 1. — Showing a period of
normal rhythm.

iricular and ventricular sequence and

first sound, on the first examination, were distinctly and audibly alternans in
character, and visibly so over the jugular. (This is explained later by the
electrocardiograms.) Respiration increased the heart rate. Exercise increased
it with some irregularity, the nature of which it seemed impossible to ascertain
from physical signs alone and even from polygraphic tracings alone. Strenuous

KAHS'— REVERSED HEART RHYTHM 831

exercise was tolerated without discomfort and produced no abnormal pulse or
blood pressure reactions.
The urine was negative.

Cardiographic Studies. — The cardiograms ;.ho\v, at times, a regular
heart with the normal auricular and ventricular sequence and normal
rhj^thm (Fig. 1). The sinus rate is 79 a minute and the conduction

[/irii-|-iiM^fiini_iji'ifiiiiiiii h

Jl^i*^

w

.68

.68

.68

.66

1 .66

P>N^\^

N^ N^

w

!

R.R
.68

.68

.68

.68

.66

.68

Fig. 2. — Showing delayed P-R conduction, with grouping of the R, P and T
waves. Jugular tracing above and radial below.

between auricles and ventricles is 0.26 of a second. The only significant
abnormality is the inversion of the T wave in Lead III.

The normal rhythm, however, is not permanent and can be easily
disturbed by various influences which act on the circulation, on the
myocardium and on the vagus nerves.

832

ARCHH'ES OF JXTERXAL MEDICI. \E

The conspicuous change that at times seems to occur spontaneously,
and can easily be induced, consists of a moderate degree of sinus
irregularity and a delay in conduction of the stimulus from the auricle
to the ventricle. This is illustrated in the first cardiogram taken
(Fig. 2). In this record, the action is regular. The auricular (P)
wave is followed by the ventricular stimulus after a delayed conduction
of 0.60 of a second. As the interval between beats is only slightly
more than this (0.68 of a second), the P wave immediately succeeds
the R and gives the picture of reversed rhythm where the R, P and T
are grouped together in this respective order. The P wave is upright
and normal in its contour.

The carotid wave is 0.18 of a second after the beginning of the
Q R S wave. The auricular wave is seen 0.18 of a second after P and
is very prominent because it contracts during ventricular systole and

i«/^M,.i«iWA%M.fWVW

^ \ y V N^^^N^

.76

Fig. 3. — Showing the effect on t!ie heart mechanism of taking a deep breath.
The two vertical lines are the signal.

forces the blood back into the jugular veins. From the jugular tracing
alone, the record could not be interpreted correctly. The carotid (c)
wave would be thought to be the auricular wave and the following
auricular wave would be considered ventricular because of the correct
time relation between them and of their relative promiiience. This
record peculiarly demonstrates the indispensability of the electrocardio-
graphic record in the interpretation of certain cases.

Effect of Vagus Stimulation. — The record following is a continu-
ation and shows the effect on the heart mechanism of taking a deep
breath indicated by the .signal in Figure .v The sinus is promptly
slowed and there is a delay in conduction; the latter efTect. on the
junctional and bundle tissues, predominating and producing transient
heart block. This is illustrated by the figures in the subjoined table
showing the time relations between the various waves. The first

) >

1

1

§

7

^^i

?

r

i^ ■

J =:i

<o

/

g

8

r

^

^

r

oi

. 1 1^

a<

ik

^

•^

l/l, '#

^

[r

^

~

•?^

^

/

a<

X

».

^

834 ARCHIVES OF ISTERSAL MEDICINE

^^tr "p^e^sr htd a very si.Har, but .uch .ore .arKed effec.
The sinus Sowing and the heart block increased -th the ".creas. o
pressure. One ectopic ventricular escape occurred alter a per.od of

° The cycle of events during continued ocular pressure is illustrated
in the series of strips in Figure 4, the original reversal of rhythm
returning after the effect of ocular pressure subsided.

E§ect of Exercise.— Dm'ing a period of normal rhythm (Fig. 1).
the patient was instructed to exercise, the exercise consisting of
hopping on one foot one hundred times. After this, the electrocardio-
gram showed a slight increase in the rate of the beat. About three

A. ^<\ vvt~v>

Fig. 5. — Showing the progressive increase of the block after exercise.

minutes after the exercise, while the patient was seated, the property
of conduction suddenly began to show easy fatigue. A normal cycle
took place with a normal F-R interval. The next cycle showed an
increased P-R interval ; the next might show a still further increase in
the block and finally an auricular beat would be blocked or an auricular
beat might occur simultaneously with a delayed ventricular beat ; after
this a repeated auricular beat would be promptly transmitted, conduc-
tion having recovered. This is illustrated in Figure 5.

Effect of Amyl Nitnte.-\my\ nitrite was administered by inhala-
tion after a sufficient amount of rest following the exercise A lone
record was taken during the entire period of inhalation. The rhythm
just before inhalation was found to be normal and the amyl nitrite did
not affect this rhythm. There was practically no increase in heart
rate. As is usual, it depressed the height of the T wave With the

KAHX— REVERSED HEART RHYTHM 835

elimination of its effect, the T wave returned to its previous height
and a distinct U wave became evident. There was no evidence of
difificult conduction.

Effect of Atropin. — When the effect of amyl nitrite had subsided
and the rhythm was still normal, atropin, 1/150 grain, was given
hypodermically. The first record was taken twenty-five minutes after
this. There became evident a sinus irregularity and a marked difficulty
of conduction, a normal conduction period being followed by a pro-
longed period or by a blocked cycle. No a-v nodal beats developed.
Sixty minutes after the atropin was given, a record showed less
difficulty in conduction than before. In Lead I, taken first, a normal
cycle was followed by cycles in which conduction was progressively
prolonged and every fourth cycle was blocked. In Leads II and III
taken immediately after, the prolonged conduction became constant at
0.46 of a second, the heart rate being eighty-three a minute; i.e., the
P-P and R-R intervals were 0.72 of a second.

This case is evidently an early case of heart block with periods of
normal conduction and an instability of the normal rhythm. The
important question to determine is whether the defect in conduction
is due to nervous influences or is more definitely organic, due to dis-
turbances in circulation or disease of the myocardium.

The features that incline one to the latter supposition are : ( 1 )
The stimulation of the vagus produces a temporary moderate, but not
constant, increase of the block: i.e., the vagus influence superimposes
itself on some other basis which continues to act after the vagus
stimulation has subsided ; (2) the depression or paralysis of the vagus
by atropin does not completely relieve the difficulty in conduction ;
(3) exercise, accelerating the heart, seems to have a beneficial effect
on conduction, perhaps by improving the cardiac circulation; (4) the
electrocardiogram shows inversion of the T wave in Lead III.

The tracings and the record of this case emphasize the distinction
that should be made between reversed mechanism and reversed rhythm
and show the importance of electrocardiographic analyses.

SOME OBSERVATIONS ON PAROXYSMAL RAPID

HEART ACTION WITH SPECIAL REFERENCE

TO ROENTGEN-RAY MEASUREMENTS OF

THE HEART IN AND OUT OF

ATTACKS *

SAMUEL A. LEVIXE, M.D.. and ROSS GOLDEN. M.D,

BOSTON

With the appHcation of the more accurate means of clinical investi-
gation in the past decade, some of the views previously held concerning
pathologic processes necessarily have needed revision. This has been
true particularly with regard to the disturbed mechanism of the heart
beat ever since the recent interest in cardiographic work. There must
take place also a similar revision of our views concerning cardiac
dilatation when one applies, as a check to findings from percussion and
palpation, measurements determined by roentgen-ray examination. Any
clinician, who has confirmed the results as to the size of the heart
determined by percussion and palpation, with roentgen-ray measure-
ments, must feel somewhat doubtful in detecting changes in the 'heart
size of less than 1 cm., and not infrequently he will find that his bedside
figures are at a considerably greater variance from the roentgenogram.

A further point that seems rather hazy in the minds of many
students is the matter of dilatation of the heart. Possibly, the com-
parison to an elastic sac expanding and contracting unconsciously has
led us to think of the heart in similar terms. For some reason there
is a common belief that the heart frequently dilates, and as the clinical
condition improves this dilatation disappears. The absolute proof that
this is a common occurrence in cardiac patients is lacking, for most
of the observations are made by the ordinary methods of inspection,
palpation, percussion and auscultation and not confirmed by roentgen-
ray examination. No doubt, the former methods are quite sufficient
in the great majority of cases for the proper clinical management of
the patient, but they are insufficient for the establishment of reliable
scientific data on which to base our knowledge.

It is with this purpose in mind that roentgen-ray observations were
made on eleven patients during attacks of paroxysmal rapid heart
action and after the heart had returned to normal rhythm. These
conditions have frequently been regarded as instances of acute dilatation

* From tlie Medical Clinic and tlie Roentgenological Department of the
Peter Rent Rrigham Hcispital.

LEVIXE-GOLDE\-!-RAPlD HEART ACTIOS 837

of the heart, especially before cardiograpliic studies enabled us to
classify and describe them properly. Mackenzie ' says :

Even more striking because more sudden and violent, are the changes
(meaning dilatation) that take place in certain cases of paroxysmal tachycardia.
. . . I have seen these cases on several occasions shortly before an attack
and watched the steady progress of the change. The hearts were normal in
size, but in three hours' time the transverse diameter had increased by two
inches, the face had become livid and the lips swollen. ... In the course
of twenty-four hours edema of the legs appeared and the liver became large.
Alter some days the dropsy extended up the legs, the abdomen became distended
and the urine scanty. With the cessation of the attack of paroxysmal tachy-
cardia, the patients at once experienced relief, and in a few hours every vestige
of heart failure had disappeared, and the heart itself returned to a normal size
and rhythm. ... I have seen many cases in which the inception of the
nodal rhythm was followed by these changes.

Lewis, ^ writing on the effect of paroxysms of fast heart on the
circulation, says :

When the heart commences to beat more rapidly it decreases in size. Where
the acceleration is great the arterial pressure falls considerably while the venous
pressure rises ; but with lesser degrees of acceleration the arterial pressure may
remain steady or may actually rise a little. . . . But in long continued
paroxysms, especially where the reserve power of the ventricular muscle is
imperfect, the heart dilates, the tall of pressure is more profound and the
blood stagnates in the heart and venous systems.

In this study are included eleven patients with paro.\ysnia! rapid
heart action, the clinical data of which are appended. Five had
paroxysmal auricular tachycardia, three had paroxysmal ventricular
tachycardia, one had paroxysmal auricular flutter and two had parox-
ysmal auricular fibrillation. In all cases the diagnosis of the type of
disturbance was confirmed by electrocardiographic tracings. In order
to determine whether the hearts dilated during the various upsets,
roentgen-ray examinations were made while the rapid heart action was
in progress and while the heart was beating at a normal rate. Measure-
ments of the transverse diameter of the heart shadows were made,
which with other details are given in Table 1. These examinations
were made after the attacks had lasted varying lengths of time, in some
instances days after the onset, in others an hour or so (Table 2).

It is not the purpose here to discuss the technic and details of the
roentgenological methods of heart examination. The method most
used in this country at the present time for the measurement of the
heart shadow is the teleroentgenogram, or the seven-foot plate. The
determination of slight or moderate degrees of cardiac enlargement
presents some difficulty even by the roentgenological method, because
the shadow on the plate is merely a silhouette and because of individual
variations associated with sex, height, weight, habitus, etc. The "cardio-

1. Mackenzie, J.: Diseases of the Heart. Oxford University Press, 1908. pp.
199-200.

2. Lewis, T.: Osier and McCrae's Modern Medicine. 1915. p. 87.

838 ARCHIVES OF ISTERNAL MEDICIXE

thoracic ratio" has been proposed by Danzer ^ as a convenient standard
for determining whether the heart shadow is larger than it should be
for the particular individual. In a study of more than five hundred
cases he found that normal hearts cast a shadow the transverse diameter
of which is 50 per cent, or less of the internal diameter of the chest
shadow measured at its widest point, usually about the level of the
fifth interspace anteriorly. This relation of the shadow of the heart
to that of the thoracic cavity he terms the cardiothoracic ratio. Allowing

TABLE 1.— ROENTGEN-

:Measurements of the Heart Shadow

Case

Diagnosis

Pat.-

.ipprox.
Target-
Plate

Dl8t.

Betore,
During,
After^
Attack

Measurements in Cm.

Change
in Mm.
with
Attack

MRIMLj TDjIDC|C-TH

1

Paroxysmal auricular
tachycardia

12/23/14
1/ 8/15

30 in.
30 m.

During
After

5.4 : 10.2 1 15.6 30.5 '. 64%
2.8 : 10.0 I 12.8 ! 32.0 40%

28

2

Paroxysmal auricular
tachycardia

r/ 6/16
7/ 8/16"

.4 in.
84 in.

During
After

5.9
5.6

9.2 ! 15.1 26.0 1 58%
10.1 15.7 26.0 j 60%

—6

3

Paroxysmal auricular
tachycardia

12/13/16
12/13/16

S4in.
84 in.

During
After

5.1

11.1 16.2 27.5 59%
10.8 15.2 27.5 1 55%

10

*

Paroxysmal auricular
tachycardia

11/ 6/19
11/ 7/19
11/10/19

84 in.
S4in.
S4in.

Before
During
After

3.8
3.5
3.9

5.5 ! 9.3
6.3 i 9.8

5.6 1 9.5

2o!5 48%
20.6 1 46%

5
3

5

Paroxysmal auricular
tachycardia

6/9/20
6/12/20

84 in.
84 in.

During
After

6.8
5.9

9.0 j 15.8
10.0 15.9

23.0 ! 69%
23.0 69%

—1

6

Paroxysmal ventricular
tachycardia

9' 5/15
12 2 15

30 in.
30 in.

During 1 3.2 ! 6.7
After 3.0 , 7.6

9.9
10.6

28.0 ; 39%
30.0 33%

-7(f1

7

Paroxysmal ventricular
tachycardia

11/ 3/16
11/ 9/16

30 in.
30 in.

During 5.0 1 U.4
After 3.8 11.9

16.4 31.5 1 52%
15.7 ! 30.6 ! 51%

7

8

Paroxysmal ventricular
tachycardia

1 5/20
3 9/20

84 in.
84 in.

During 1 4.8 ' 13.1 17.9 27.0 1 66%
After : 5.2 , 11.3 ^ 16.5 27.0 , 61%

14

9

Paroxysmal auricular

10/21/14
10/29/14
2/ 5/15
2/ 8/15

30 in!
30 in.
30 in.

During ' 7.3 i 9.6 16.9 32,0 53%
After , 7.1 ■ 9.4 i 16.5 I 32.0 1 52%

During i 6.5 ■ 9.3 15.8 30.5 i 52%
After j 6.4 ' 9.3 : 15.7 ; 80.5 51%

4
1

10

Paroxysmal auricular
fibrillation

3/12/20
3/15/20

84 in.
84 in.

During
After

4.5 ; 8.6 1 13.1 21.5 ! 61%
4.0 8.6 12.6 21.5 59%

5

11

Paroxysmal auricular
fibrillation

12/14/20
12/16/20

84 in.
84 in.

During
After

sil M U'.S u'.B 60%

-6

farthest

I D O = Internal diame

C-T B = cardiothoracic

■iinsverse diameter of tlie lie
the chest shadow at about
: T D divided by IDC.

a margin of 2 per cent, for error, he believes .the heart to he enlarged
when this ratio is more than 52 per cent., but points out that enlarge-
ment cannot be ruled out when the figure is less than 52 per cent.,
particularly in the ptotic habitus. As both the size and shape of the
chest and the size and position of the heart vary with the above
mentioned factors, especially with the habitus, this seems to be the best

3. Danzer, C. S. : Cardiothoracic Ratio, .^n Index of Cardiac Enlargement,
Am. J. M. Sc. 157:513, 1919.

LEflXE-GOLDEX— RAPID HEART ACTIOS 839

method at hand for the determination of slight or moderate degrees of
enlargement. It has been used in this study as a check on the actual
measurements in centimeters of the heart shadow, especially in those
cases taken at a distance of about thirty inches, in three of which there
was apparently enough variation in the distance to make a difference of
1 or 2 cm. in the internal diameter of the chest. In Case 6, for
example (Table 1), the heart shadow itself measured 0.7 cm. more

TABLE 2.— Clinic.\l D.^ta

Case

Med.
No.

2092

Age

IT

Sex

IT

Diagnosis

Was-
ser-

Symptoms

Failure
During

Attack

Bate

in

Attack

loura-
Dura- tionat
tionoj Timeol
Attack Roent-
genos-
copy

Dilata-
tion

1

Paroxysmal auricular
tachycardia

-

Severe
Severe

242

VhT'

3 days iMarked

2

4940

50

F

Paroxvsmal auricular
tachycardia, syphilis

++

Moderate

220

4% da.

4 da.

None

3

5742

21

M

Paroxysmal auricular
tachycardia, aortic in-
sufficiency, mitral in-
sufficiency, chronic
myocarditis

Severe

162

Several
hours

Few
hours

Slight

♦

12207

24

F

Paroxysmal auricular

tachycardia, mitral

stenosis

-

Moderate

233

2hr8.

^4hr.

f

6

13577

50

F

Paroxysmal auricular
tachycardia, chronic
infectious arthritis

-

None

171

Several
hours

2 hrs.

None

B

3324

64

^

Paroxysmal ventricular

tachycardia, chronic

myocarditis

~

Moderate

160

Several
hours

Few
hours

None

.,503

50

M

Paroxysmal ventricular

tachycardia, chronic

myocarditis

Severe

160

Few
hours

Ihr.

None

8

12846

64

M

Paroxysmal ventricular

tachycardia, chronic

myocarditis

-

Severe

165

36 brs.

12 hrs.

Slight

9

1791

35

M

Paroxysmal auricular

flutter, aortic stenosis

and insufficiency

liilSf

1^1

'\1i^

11 da.
3 da.

None
None

10

13074

38

F

1

Paroxysmal auricular
flbrlllation. mitral
stenosis and In-
sufficiency

Moderate

136

hours

Ihr.

None

11

•

26

M

Paroxysmal auricular
fibrillation

Slight

120

2 da.

18 hrs.

None

• This was a private case o! Dr. Paul D. White, Boston, who kindly offered it lor use In
this series, lor which we wish to express our appreciation.

after than during the attack ; but as the internal diameter of the chest
shadow also was larger in the second plate, the cardiothoracic ratio
was the same in both, showing that no definite change in the size of the
heart shadow had taken place.

In the estimation of the size of the heart from its shadow on the
plate three sources of error are to be considered: (1) magnification
due to the divergence of the rays; (2) changes in the position of the

840 ARCHIVES OF IXTERXAL MEDICIXE

heart during the respiratory cycle, and (3) changes in the profile of the
heart due to the fact that the chest is not pressed symmetrically against
the plate during the exposure and spoken of as "rotation." It seems
advisable to discuss these points very briefly.

The magnification due to the divergence of the rays is practically
compensated for by increasing the distance between the target of the
roentgen tube and the plate to approximately seven feet. The mathe-
matically figured correction for the transverse diameter of the heart
shadow taken at a target-plate distance of two meters, as given by
LeWald and Turrell,^ is 4 per cent., i. e., the true transverse diameter
is 96 per cent, of that of the silhouette.

During ordinary quiet respiration the position of the heart changes
very slightly, if at all. With full inspiration or forced expiration,
however, there is a very marked change in the relation of the heart
to the chest wall and in the physical axis of the heart, and consequently
in the form and size of its silhouette. The transverse diameter of the
heart shadow of one of us taken during quiet respiration at approxi-
mately 7 feet measured 14.3 cm., while that taken at the same distance
but with full inspiration measured only 12.8 cm. At a thirty-inch
distance with full inspiration the shadow was 14.1 cm. in diameter.
Here the change in the heart shadow due to the descent of the diaphragm
was slightly greater than that due to the magnification from the
divergence of the rays. For this reason the standard teleroentgenogram
is made during quiet respiration. Unfortunately the plates made in
four of the cases in 1914-1916 were taken at a thirty-inch distance at
full inspiration, the technic used for the study of the lung fields. This
study, however, is not concerned so much with the exact size of the
heart as with an estimation of the possible changes in its size which
might occur during the course of paroxysmal rapid heart action.
Therefore, we believe that two plates taken at approximately the same
distance at the same period in the respiratory cycle will furnish data
sufficiently accurate for this purpose, i. e., the comparison of the rela-
tive sizes of the heart shadow in the same patient on two diiierent
occasions.

The presence of rotation can very easily be determined by obser-
vation of the relation of the shadows of the sternoclavicular joints to
that of the spine. Very slight degrees of rotation produce errors which
are insignificant and which do not vitiate the observation. In this series
it occurred in very few plates and in such slight degrees that it has
been ignored.

4. LeWald, L. T.. ami Tiirrell, G. H.: The .Aviator's Heart. Roentgen Ray
Studies Under Conditions Simulating High .Altitude?. .\ni, J. Roentgenol. 7:67,
1920.

LEIIXE-GOLDEX—RAPID HEART ACTION 841

It is generally accepted that changes in the measurements of the
heart shadow up to 0.5 cm. and in the cardiothoracic ratio of 2 per cent,
are within the limits of error and may be discounted.

In estimating the change in the heart shadow we have used the
actual difference in the transverse diameters and the difference in
the cardiothoracic ratios. In a review of Table 1, one is readily
impressed by the absence of any distinct dilatation in most of the cases.
\\'hen it is remembered that changes of 0.5 cm. or less in the transverse
diameter of the heart shadow and of 2 per cent, in the cardiothoracic
ratio are within the limits of error, there remain only three cases in
which an appreciable dilatation of the heart occurred. Case 1 shows
the most marked change in the series. Here the heart dilated suf-
ficiently to produce an increase of 2.8 cm. in the transverse diameter
and of 14 per cent, in the cardiothoracic ratio. This patient was also
seen in a similar attack some months later and a dilatation was again
observed with an increase of 2.5 cm. in the heart shadow. Unfor-
tunately, one of these plates was lost. The marked dilatation in
this case probably was the result of the extremely rapid heart rate
and the comparatively long duration, for nothing abnormal in the heart
itself could be detected between attacks. Case 3 was a patient with
severe valvular disease who showed clinical evidence of a markedly
damaged myocardium. In this case there was an increase of 1 cm.
in the transverse diameter and of 4 per cent, in the cardiothoracic ratio.
Despite the short duration and the rather low heart rate during tachy-
cardia (162), distinct dilatation occurred and may be explained by the
poor condition of the heart that was evident when the rate was normal.
Case 8 was an elderly colored man with a severe chronic myocarditis
who had attacks of ventricular tachycardia. A roentgenogram taken
during an attack, twelve hours after the onset, while the rate was 165,
showed an increase of 1.4 cm. in the transverse diameter of the hear'
shadow and of 5 per cent, in the cardiothoracic ratio over the measure-
ments made after the heart had returned to normal rate. This dilatation
may be explained as in the previous case. In seven of the remaining
eight cases the changes did not exceed the limit of error. In Case 4
questionable increase in the size of the heart occurred, the cardio-
thoracic ratio being 3 per cent, larger during than before the attack, and
2 per cent, larger during than after the attack. It is interesting that
this patient, who had mitral stenosis, had considerable dilatation of the
left auricle during the attack as is shown in Figure 1. In Cases 2 and
11 there was an apparent decrease in the size of the heart shadow
during the attack.

The main interest in these roentgen-ray findings is the absence of
definite dilatation in the majority of the cases. It is fair to say that, with
the exception of Case 1, and possibly of Case 8. it would be impossible

842 ARCHU'ES OF IXTERXAL MEDICIXE

to detect by percussion and palpation any of the changes that occurred
in this series. For it must be appreciated that when a change in the
transverse diameter of the heart shadow occurs, it may be divided into
two portions, one to the right and one to the left of the midline. For
example, in Case 3, to detect the difference of 1 cm. that occurred, it
would be necessary to percuss an increase of 7 mm. to the right and
3 mm. to the left. It must not be understood that even the slight
dilatation that escapes clinical detection is not significant, for if the
transverse diameter of the heart increase 0.5 cm. it would mean
considerable stretching of the muscle fibers and likewise a considerable
increase in the heart volume. But the important point is that the
dilatation is so slight in most cases that it cannot be detected on ordinary
bedside examination, and that only in isolated cases is it very appreciable
even when roentgen-ray examinations are made. It appears to us that
the three factors that determine whether dilatation will occur are the
duration of the attack, the rapidity of the ventricular rate during
the attack and the general health of the heart. The longer the attack,
the more rapid the heart rate and the more severely damaged the heart,
the more apt it is to dilate.

Observations on the systolic, diastolic and pulse pressures were
made on seven of the prtients during the course of rapid heart action
and again while the heart mechanism was normal. Our interest was
aroused by Case 1. The patient, who had suffered serious accidents
with three attacks, always showed a remarkably small pulse pressure
during those in which he was observed. With such a sluggish circula-
tion, it is not surprising that he de\eloped gangrene of an arm.
hemiplegia and aphasia.

In Case 1 (Table 3) the systolic pressure fell and the diastolic
rose, resulting in a pulse pressure of 8 mm. as soon as eight hours
after the onset of the attack although the patient had no demonstrable
, heart disease. This occurred because the rate was extremely fast, about
250 to the minute. Case 2, on the other hand, showed a very low
pulse pressure of 10 mm. when the attack had lasted three days, but
the change was only slight during another attack when the readings
were made after it had lasted only sixteen hours. Case 4 illustrates
the rapid development of a small pulse pressure, for within fifteen
minutes after the onset of a rate of 233 it was only 12 mm., while
nine minutes after the attack had ended the pulse pressure had risen
to 18 mm. and three hours later to 24 mm. This rapid change took
place because of the very fast heart rate and because the patient was
already suffering from mitral stenosis. Similarly Case 7 showed a
decrease in the pulse pressure from 31 mm. to 18 mm. and from 46
mm. to 14 mm. only a few hours after the onset of two attacks,
although the heart rate was only 160. Tliis is explained by the fact

LEVINE-GOLDEX— RAPID HEART ACTION 843

that the patient was already suffering from chronic myocarditis. Like-
wise in Case 8. a chronic myocarditis, a small pulse pressure developed
within one hour after the onset of tachycardia, although the heart rate
rose to only 165. The last patient (Case 9), who had aortic stenosis
and a low systoUc pressure of about 90 mm., when the heart was beating
normally, showed a pulse pressure of only 12 mm. three days after the
attack of auricular flutter had begun, while the ventricular rate was

TABLE 3. — Blood Pressure Re.adings

Case

Diagnosis .

Date

Heart
Rate

Blo

Sys-
tolic

Dias- 1 Pulse
tolic

Time of Blood Pressure

Readings in Relation

to Attacks

1 1 Paroxysmal auricular
tachycardia

12/28/14
1/ 315
7/14/15

250
8-2
242

94
lOO
84

66
76

8
34

8

2 days after cnspt

3 days after end

8 hours after onset

2 j Paroxysmal auricular
tachycardia, syphilis
(positive Wassermann)

7/ 6/16
7/ 7/16
7/10/16
7/13/16
7/15/16

IS
86

192
68

148
9i,
135

114
75
116

95

10

i

fo

3 days after onset
1 day after end
1 day after end
16 hours after onset
1 day after end

3

Paroxysmal auricular

tachycardia, aortic

and mitral insufBciency,

chronic myocarditis

12/ 9 16
1213/16

162
90

148
200

30 ?
0?

118
200

10 hours after onset
1 day after end

Paroxysmal auricular

tachycardia, mitral

stenosis

11/ 5/16
11/ 7 Ifi
11/ 7/16
11/ 7/16
11/ 7/16
11/ 7/16

130

233

140
132

120
136
138
134
126
132

95
124
126
122
108
106

25
12
12
12
18
24

2 days before attack
15 minutes after onset
17 minutes after onset
19 minutes after onset

9 minutes after end

3 hrs. 22 min. after end

'

Paroxysmal ventricular

tachycardia, chronic

myocarditis

10/25/16
10/30/16
11/ 3/16
11/ 7/16

160
90
160

86

96

^^
114

80
72
83
68

18
31
14
46

1 hours after onset
3 days after end
Few hours after onset
3 days after end

8

Paroxysmal ventricular

tachycardia, chronic

myocarditis

1/ 3 20
1/ 4/20
1/ 5/20
1/ 5/20
1/ 8/20
1/19/20

S»
164
165
163
165

82

145
115

n2
no

110

124

95
94
94
90
92
92

50
21
18

32

1 day before onset
1 hour after onset

24 hours after onset

25 hours after onset
12 hours after onset
12 hours after end

9

Paroxysmal auricular

flutter, aortic stenosis

and insufBciency

2/ -5/15
2/ 6/15

176
74

92
90

SO
69

12
21

3 days after onset, aver
age of 6 readings

6 hours after end, average
of 3 readings

,

In

Out

194.3
94.2

112.5
126.7

90.4
81.4

22.1
45.1

176; six hours after the attack had stopped the pulse pressure had
risen to 21 mm.

This discussion, although it includes only seven cases with a com-
paratively small number of blood pressure readings, indicates that dur-
ing attacks of rapid heart action there is a tendency for the pulse
pressure to decrease. The extent of this diminution, like the occurrence
of dilatation, will depend on the duration of the attack, the rapidity of
the heart and the state of health of the heart muscle. It also shows that
the pulse pressure may occasionally become dangerously low when

844 ARCHIVES OF IXTERXAL MEDICIXE

these three factors are sufficiently antagonistic to an efficient circulation.
The figures at the bottom of Table 3 indicate that the decrease in
pulse pressure is brought about by a fall in systolic pressure and an
increase in diastolic pressure, for although there are some variations
in the isolated observations, the average of all readings show^s a fall
of systolic pressure from 126.7 mm. to 112.5 mm., and a rise of
diastolic pressure from 81.4 mm. to 90.4 mm. during attacks. Readings
made on other patients with paro.xysmal rapid heart action, not included
in this study because no roentgen-ray observations were made, showed
similar changes.

In addition to the roentgen ray and blood pressure determinations,
a study of these cases showed that a leukoc\'tosis and slight fever were
not uncommon findings during the attacks of rapid heart action. In
some instances they were quite striking, the temperature and leukocyte
count falling to normal promptly after the attack was over. Six of
the patients had a leukocytosis of from 13,000 to 22,000 during the
upsets and two had a temperature of over 100 F. It does not seem
likely that an actual infection had taken place which brought on the
tachycardia, but rather that the fever and leukocytosis were the result
of the cardiac upset.

SUMM.\RY

Roentgen-ray examinations and blood pressure studies were made
on eleven patients with paroxysmal rapid heart action. Five had
paroxysmal auricular tachycardia, three had paroxysmal ventricular
tachycardia, one had paroxysmal auricular flutter and two had parox-
ysmal auricular fibrillation. ( )bservations were made during the
attacks and while the heart was beating normally.

It was found that in eight cases no appreciable dilatation of the
heart occurred, in two it was definite but slight and in one it was
considerable. These results indicated that in ten out of eleven cases
it would have been impossible to detect with any certainty by percussion
and palpation a change in the size of the heart.

In seven of the cases blood pressure readings showed that the
systolic pressure was apt to fall and the diastolic pressure to rise,
resulting in a low pulse pressure which in rare instances became very
small, i. e., 8mm. This low pulse pressure may explain some of tlie
symptoms that occur during the severe attacks.

It is suggested that the amount of dilatation and the decrease in
pulse pressure are dependent on three factors; the duration of the
attack, the rapidity of the ventricles during the attack and the state
of health of the heart before the attack occurs.

In several cases a leukocytosis, even as high as 20,000, and a
temperature of 100 F. developed with the attacks and quickly subsided
as the heart returned to normal.

LEVINE-GOLDEN— RAPID HEART ACTIOS 845

REPORT OF CASES

Case 1 (Med. No. 2C&2). —History.— A baker, age 38, was seen at the hospital
on numerous occasions for attacks of palpitation. The first attack occurred
about four years before his first admission to the hospital. It lasted four days
and during it he suddenly developed loss of meniorj- which lasted for a year.
The second attack occurred about one year later and continued for eight
and one-half days. During the seventh day he developed a right hemiplegia
which cleared up after six months, leaving a slight weakness of that side of the
body. Two and a half years before the first admission he had the third attack.
This lasted ten days, but on the eighth day he developed drj* gangrene of the
left arm which required amputation just below the shoulder. Dec. 26, 1914,
following a fall downstairs, he noticed his heart beating irregularly and later
rapidly. .'\t times there were sharp gripping pains over the heart and a
feeling of dizziness. He was short of breath after the onset but had no
orthopnea, cough or edema. All the previous attacks began and ended sud-
denly. He entered the hospital Dec. 28, 1914.

Physical Examination. — This showed a well developed man with a very
anxious expression on his face. His skin was covered with a profuse perspira-
tion. Very rapid oscillatory pulsations were seen over the jugular bulbs. Heart
examination showed an extremely rapid regular rate, 250 to the minute. No
murmurs were heard. The second heart sound could not be heard. Lungs were
negative, liver edge not felt, no edema of the legs. The fingers and lips were
distinctly cyanosed.

The urine was negative. The blood showed 21,300 leukocytes, with 80 per
cent, polymorphonuclears during the attack, and 7.700 two days later during
normal rate. The temperature ranged around 101 F. during the attack and
then fell to normal. The blood Wassermann was negative. Electrocardiograms
taken during the attack showed paroxysmal tachycardia of auricular origin.

Course. — Numerous attempts to stop this attack by ocular and direct vagal
pressure were unsuccessful. It stopped spontaneously two and a half days
after admission. During the following two years the patient was observed
in several similar attacks all of which were successfully ended by pressure
over the carotid artery. During one of these further blood counts were made
and a leukocytosis of 25,400 was found ; there was no fever. The following
day when the heart rate was normal the white count was 6,000. Between attacks
examination showed the heart to be normal.

Blood pressure readings on the first admission to the hospital were: during
tachycardia, systolic, 94: diastolic, 86; five and seven days later during normal
rate: systolic, 100; diastolic, 66, and systolic, 114, diastolic, 84, respectively.
Several months later during an attack: systolic, 84; diastolic, 76; one-half hour
after the attack was ended by vagal pressure: systolic, 106; diastolic, 86.
The following morning: systohc, 130; diastolic, 90. Roentgenograms were
taken at a distance of about thirty inches during and after two attacks, and
showed an increase in the transverse diameter of about an inch while the
tachycardia was in progress. During recent years the patient has had occasional
attacks which he has been able to stop by vagal pressure or holding a deep
inspiration.

Diagnosis. — Paroxysmal auricular tachycardia.

Cash 2 (Med. Xo. A9AQ) .—History.— .\ woman, 50 years old, entered the
hospital June 5. 1916, complaining of palpitation. Twenty-five years previously
she had typical rheumatic fever. Thirty-five years previously, while stooping
to pick up something from the floor, her heart suddenly began to beat very
rapidly. This attack lasted five minutes. Similar attacks recurred every eight
or twelve months, gradually becoming more frequent and of longer duration.
Things would become black before her eyes at the b.'ginning and end of each
attack. She would always lie down during the attacks. The longest attack

846 ARCHIVES OF IXTERXAL MEDICIXE

was of seven days' duration. Recently she had been having an "aura" of
indigestion, a heavy feeling in the abdomen, nausea, gas and nervousness for
two or three days ending with precordial pain just preceding the attack. Riding
in a buggy on a rough road used to stop the attacks, later running up and down
stairs and applying ice to the precordium were successful, but now the only
thing that works is vomiting for which she uses ipecac. During the attacks
she had dyspnea on slight exertion, a tight sensation in the chest and tremendous
palpitation. The upset which brought her to the hospital began June 2. 1916.

Physical Examination. — This showed a somewhat enlarged heart with an
extremely rapid and regular rhythm but no murmurs. There was an area
of hyperesthesia in the right hypochondrium and a palpable and very tender
liver edge. Vagal pressure was tried without success. The attack stopped
spontaneously the night following admission, i. e., four days after the onset.
During her stay she had several other shorter attacks, some stopping spontane-
ously while others vfere ended by vomiting induced by ipecac.

Thee white count was from 18.000 to 21.000. The urine showed a large trace
of albumin on admission which later completely disappeared. The phenolsul-
phonepthalein output in two hours was 48 per cent. Wassemiann reaction was
strongly positive. The basal metabolism during the attack was + 27 per cent. ;
during normal rate, + 6 per cent. Electrocardiograms showed paroxysmal
tachycardia of auricular origin with a rate of 217.4. There was typical alterna-
tion of the ventricular complexes. Blood pressure readings were : during first
attack, July 6: systolic, 124; diastolic, 114; July 7, normal rate: systolic, 110;
diastolic, 75 ; July 13, during the third attack (duration one day) : systolic, 95 :
diastolic, 65 ; two days later with normal rate, systolic, 135, diastolic, 95.
Roentgenograms taken at seven feet during and after the attack showed no
appreciable change in the transverse diameter of the heart.

Course. — She was discharged July 26 improved.

Diagnosis. — Paroxysmal auricular tachycardia : syphilis ( positive Wasser-
mann).

Case 3 (Med. No. 5742).— History.— A man, aged 21, entered the hospital
Dec. 9, 1916, complaining of pain in the heart and palpitation. He had rheumatic
fever in 1913 and was quite sick for four months, after which he had frequent
attacks of sore throat. For four years he noticed shortness of breath, and
in 1915, after climbing a flight of stairs, he had severe palpitation. He had
numerous attacks of palpitation since then, at first coming on about once a
month but more recently every day. Generally, the attacks were of short
duration, sometimes lasting minutes or a few hours, but one attack continued
for twenty-four hours. Patient stated that at the commencement of an attack
the heart became irregular, weak beats and forceful thumps interspersed with
pauses and then the heart would finally "start its race." At the time of admission
and for several days afterward he had frequent short paroxysms of tachycardia
causing marked subjective discomfort.

Physical Examination. — This showed an enlarged heart with signs of aortic
and mitral insufficiency. There was no congestion of the lungs or edema of the
legs. The liver was not enlarged. The pulse rate generally was around lOU or
120, and would suddenly rise to 160 during attacks.

Electrocardiograms showed paroxysmal tachycardia of auricular origin. Pulse
tracings showed pulsus alternans during attacks. The urine was negative.
Blood Wassermann was negative. There was a leukocytosis varying from
15,400 to 22,400. The temperature was essentially normal throughout. During
attacks systolic blood pressure was 148 while at other times it ranged around
200; the diastolic figure could not be obtained. Roentgenograms were taken
at a distance of seven feet during and after an attack on December 13. The
plate taken during the attack showed the transverse diameter of the heart shadow
to be 1 cm. greater than that of the plate taken after the attack.

LEVINE-GOLDEX— RAPID HEART ACTIOS' 847

Course.— Dnmg his stay in the hospital the patient was extremely sick,
suffering with severe precordial pain, and having frequent paroxysms of tachy-
cardia which were generally stopped by vagal or ocular pressure. One day he
had an attack of unconsciousness with gasping respirations. The general
condition gave the impression of severe myocardial damage in addition to the
valve defects. He gradually improved so that during the last two weeks of his
stay in the hospital he was absolutely free from attacks. He was discharged
improved Jan. 4, 1917.

Diagnosis. — Aortic and mitral insufficiency ; chronic myocarditis, paroxysmal
auricular tachycardia.

The patient died about one year later.

C.A.SE 4 (Med. No. \22Q7).— History.— .\ seamstress, aged 24, came to the
hospital Dec. 5, 1919, complaining of "heart spells." She had the ordinary
children's diseases, frequent sore throats and typical rheumatic fever lasting
one month at the age of 10. One year later, while jumping rope, she suddenly
had a queer sensation in her throat "as if her heart had jumped up." She felt
faint and found that her heart was beating very rapidly. She remained in
bed two weeks, at the end of which time she vomited and the heart slowed
down as suddenly as the attack had begun. Following that similar attacks
recurred about twice a year, always ending with a vomiting spell which was not
voluntarily induced. Later the attacks were much shorter in duration and
more frequent, appearing every few weeks and lasting about twelve hours ;
generally the heart rate was around 200. She thought that her upsets were
brought on by dietary indiscretions, excitement or exertion.

Physical Examination. — At the time of admission the patient was not having
an attack, but came in to be studied. She had typical signs of mitral stenosis,
with a pulse rate of 120, but without evidence of decompensation. Otherwise
the physical examination was negative.

The urine and blood Wassermann were negative. White blood count on
admission was 13,200.

Course. — Two days after admission the patient developed one of her attacks
of rapid heart action, the rate jumping from 90 to 233. Electrocardiograms
confirmed the diagnosis of paroxysmal auricular tachycardia, .'\fter numerous
attempts to stop the attack by having the patient hold a forced inspiration
and by pressure over the eyeballs and over the carotids, left vagal pressure
finally ended the attack, the heart rate returning to its previous level. Blood
pressure readings were made as follows : before the attack : systolic, 120,
diastolic, 95; during the attack: systolic, 136, diastolic, 124; nine minutes after
the attack: systolic. 126; diastolic. 108: three hours later: systolic, 132;
diastolic, 108; five days later: systolic, 125; diastolic, 85. Roentgenograms were
taken before, during and after the attack, and showed a definite dilatation
of the left auricle during the paroxysm but no appreciable increase in the
transverse diameter of the heart shadow. She had no further attacks in the
hospital and was discharged on Nov. IS.

Diagnosis. — Mitral stenosis, rheumatic in origin, paroxysmal auricular tachy-
cardia.

C.\SE 5 (Med. No. \ii77) .—History.— .\ woman, aged 50, entered the hospital.
May 18, 1920, complaining of painful swelling of the hands and feet. Seven
weeks before admission she suddenly noticed painless swelling of the sole of the
left foot. Three weeks later a similar swelling appeared in the right foot.
Each lasted about a week. In the meantime, the joints of both hands and
feet began to ache and swell.

Course.— Dmm^ her stay of four months in the hospital she ran a typical
course of chronic infectious arthritis. Blood pressure on admsision was:
systolic, 180; diastolic, 115. .'^t times she had a fever, and her urine showed
many pus cells which were thought to be due to a pyelitis. The leukocyte
count, except on one occasion, was normal. Blood Wassermann was negative.

3 > s

— 1- -c

15 o

'E.E

o o
2 £.

LEVIXE-GOLDEX— RAPID HEART ACTIOS 849

Incidental to her main illness, on two days she had transient attacks of
tachycardia lasting a few hours. These caused no inconvenience to the patient,
and in fact were discovered on routine examination. During the attacks the
heart rate was 175 to 180. One was observed to end spontaneously and the
other was stopped by vagal pressure. No observations on blood pressure or
leukocyte counts were made during these attacks. Electrocardiograms showed
paroxysmal auricular tachycardia. Pulse tracings during the attacks showed
pulsus alternans. Examination of the heart during the slow rate showed definite
enlargement. Roentgenograms taken at seven feet during and after one of the
paroxysms showed no change in the size of the heart. She was discharged
Sept. 19, 1920, with her joint condition unimproved.

Diagnosis. — Chronic infectious arthritis: paroxysmal auricular tachycardia.

Case 6 (Med. No. 3324"). — History. — A woman, age 64, entered the hospital
Sept. 10, 1915. complaining of palpitation. Her father died at 51 of apople.xy.
Her mother died suddenly at 34 presumably nf heart disease. Two lirothers
died of heart disease (one suddenly). Patient had been a school teacher for
forty years and had never missed a day's work until the week before admission.
During April, 1915, she began to have palpitation and oppression in the chest.
She kept at work but often had to lie down during attacks of palpitation.
For a year she noticed slight shortness of breath and swelling of the ankles at
night. For three months walking would bring on a feeling of faintness and
palpitation. She took tincture of digitalis and some strychin oflf and on during
the month before admission. The morning of entry to the hospital she awoke
with a sense of oppression in the chest and a rapid heart.

Physical Examination.— 'NegAt'we. except for rapid heart action and poor
heart sounds. The rate was 160 per minute and regular. Electrocardiograms
showed paroxysmal tachycardia of ventricular origin. .Mter the resumption of
the normal rate the heart examination was al.so negahve.

The urine and the leukocyte count were essentially normal. The blood
Wassermann was negative. Temperature remained normal throughout.

Course.— The patient had numerous attacks of tachycardia lasting from a
few minutes to several hours, and although frequent attempts were made to
stop the attacks by ocular and vagal pressure none were successful. In addition
there were times when the heart mechanism showed transient auricular fibrilla-
tion. The admission blood pressure, which was the only one obtained, was:
systolic. 116; diastolic, 70. Roentgenograms taken at a distance of about thirty
inches during tachycardia and during normal rate showed no appreciable change
in the size or outline of the heart. She gradually improved with rest in bed
and small doses of digitalis and was finally discharged Dec. 16, 1915. improved.

Diagnosis.— Chronic myocarditis; paroxysmal ventricular tachycardia.

Patient died suddenly Oct. 1, 1916.

Case 7 (Med. No. 5503) History.— A machinist, .50 years old, entered the
hospital Oct. 25, 1916, complaining of indigestion and palpitation. He had con-
sumed large amounts of tobacco and was a moderate drinker. One year previously
he had an attack of indigestion lasting ten days. He belched a great deal of gas,
vomited after meals and was constipated. During this time he had dull pams
down both arms, palpitation and rapid heart. One and a half months before
admission he had a similar attack. Since then attacks of palpitation have been
the most troublesome complaint. A few days before entering the hospital he
thought he was going to die during an attack. These attacks lasted about
fifteen minutes and occurred daily for three weeks. For one month he had
increasing dyspnea and orthopnea. There was no edema, cough or precordial
pain. For a few days he had dizzy spells requiring him to hold on to nearby
objects for support.

Physical Examination.— This showed a very sick-looking, orthopneic man,
with a pasty yellowish appearance to his face. The lips were cyanosed. The

850 ARCHIVES OF IXTERXAL MEDICIXE

heart was considerably enlarged, the sounds distant, rapid and regular, rate
160, and no murmurs. The lungs were negative. The liver edge was felt 4 cm.
below the costal margin and was tender. There was slight edema of the ankles.

The urine showed a very slight trace of albumin and a few granular casts.
Subsequently the urine became normal. There was a leukocytosis of 19,000. The
Wassermann reaction was negative. There was a temperature of about 100 F.
for a few days. Electrocardiograms taken during the attack showed paroxysmal
ventricular tachycardia.

Course. — The attack which brought the patient to the hospital continued for
about three days. Numerous attempts at vagal and ocular pressure failed to
produce any slowing of the heart rate. Three grams of digitalis leaves were given
during the course of the first eight days in the hospital. Shortly before the attack
ended the patient had a spell of hiccoughing. During the second week in the
hospital he had another attack lasting about twenty-four hours. Blood pressure
readings were made at various intervals during and after attacks. October 25.
the heart rate was 160; systolic pressure, 98; diastolic, 80. October 30, with the
heart rate normal, systolic pressure was 103 : diastolic, 72. Xovember 3, the
heart rate was 160; systohc pressure, 97; diastolic, 83. November" 7, with a
normal heart rate, systolic pressure was 114; diastolic, 68. Roentgenograms
taken at a distance of about thirty inches during and after the second attack
showed no appreciable change in the shape or in the transverse diameter of
the heart. Discharged November 16, improved.

Diagnosis. — Chronic myocarditis ; paroxysmal ventricular tachycardia.

Case 8 (Med. No. 12646) .—History.— A colored man, aged 64, entered the
hospital Jan. 3, 1920, complaining of dizziness. During the previous year he
had five or six spells of dizziness and fainting, in which he fell to the ground
and lost consciousness for five or ten minutes. After this he felt weak and
unsteady. For the previous five or six wrecks he had shooting pains in his legs,
and for one week there w-as increasing dyspnea and weakness so that he was
compelled to remain in bed.

Physical Examination. — This showed cyanosis of the lips and finger tips.
The heart was considerably enlarged and there was a blowing systolic murmur
at the apex. The heart rate on admission was 90. There was some congestion
at the base of the left lung. The liver edge was just felt below the right
costal margin. There was no edema of the legs.

The urine on admission showed a trace of albumin which cleared up sub-
sequently. There were no casts. The leukocyte count on admission was
13,600; the temperature was 99.6 F. The Wassermann reaction was negative.

Course.— The: day after entry his heart rate jumped from 74 to 160 and
electrocardiograms showed paroxysmal tachycardia of ventricular origin. Pulse
tracings at this time showed pulsus alternans. This attack lasted thirty hours.
During his stay in the hospital he had several such attacks lasting generally two
hours. Numerous attempts to stop the paroxysms by ocular and vagal
pressure were unsuccessful ; they always ended spontaneously. The blood
pressure on admission with a normal heart rate was : systolic, 145 ; diastolic. 95.
The next day during tachycardia it was: systolic. 115; diastolic, 94 On
two other days during tachycardia three readings were obtained as follows:
(I) systolic, 112; diastolic. 94; (2) systolic, 110; diastolic, 90; C3) systolic, 110;
diastolic, 92. Jan. 19, 1920, during normal heart rate, the readings were : systolic,
124; diastolic, 92. Roentgenograms were taken at a seven-foot distance during
and after one attack. The transverse diameter of the heart shadow was
slightly but definitely increased during the attack.

Course. — During the first two weeks, the patient was dangerously ill, but
with rest in bed and digitalis therapy he gradually improved and became free
from attacks. He was discharged March 10. 1920. improved.

Diagnosis: Chronic myocarditis: paroxysmal ventricular tachycardia.

LE\-1\E~G0LDES— RAPID HEART ACTIOX 851

Case 9 (Med. No. \791).— History.— A man, age 35, entered the hospital
Oct. 21, 1914, complaining of shortness of breath. Two years previously he had
an attack of acute tonsillitis with some stiffness of the elbows. On October 9
he began to have pain in his shoulders and arms and on the following day he
felt that his heart was heating rapidly. He kept about his work but on the
sixteenth he began to feel short of breath. After staying two days in the
hospital the patient was displeased with his treatment and left. His course,
however, was followed in the Outpatient Department from time to time. The
attack of rapid heart action and shortness of breath ended some time between
October 22 and 29. After this he was well for three weeks when a similar
attack occurred lasting three weeks. He remained in good health until Feb. 2,
1915, when the third attack began, associated with pain in the left ankle and
both shoulders.

Physical Exaniiiiatioiu — During the two attacks which were observed the
physical findings were essentially the same. The heart was definitely enlarged
and the action rapid and regular, 176 to the minute. No murmurs were heard.
There were prominent pulsations in the veins of the neck. During the first
attack the urine was negative, the white count was normal, the temperature
w-as 99.2 F. and the blood Wassermann was negative. During the second attack
the temperature was 99.4 F. At this time he was given 1 gm. sodium salicylate,
every three hours, and digitalis leaves, 0.1 gm., three times a day. Electrocardio-
grams taken during both attacks showed that the auricles were fluttering with
a rate of about 350 and the ventricles w'cre contracting regularly to every other
auricular impulse. February 6 he awoke and found that his heart was beating
normally. Examination during this normal rhythm disclosed signs of aortic
stenosis and slight aortic insufficiency, i. e., there was a prominent systolic thrill
in the aortic area, a loud systolic murmur and a faint but definite diastolic
murmur. The blood pressure Febrnan,^ 2. during an attack, was systolic, 92 ;
diastolic, 80; February 6, during normal rate: systolic, 90; diastolic, 69. Roentgen-
ograms taken at about thirty inches during and after both the first and third
attacks showed no appreciable change in the outline or in the transverse diameter
of the heart shadow. The first plate was taken when the attack had lasted
twelve days. The roentgenogram during the other attack was taken when
it had been in progress three days.

Course.— The patient returned to Bulgaria, his native country, and was
reported to have joined the army and to have died during the succeeding year
or so.

Diagnosis. — .\ortic stenosis, slight aortic insufficiency; paroxysmal auricular
flutter.

Case 10 (Med. No. 13074).— //i.!/ory.— A woman, aged 38, entered the
hospital March 2, 1920, complaining of pain in the heart and stomach. Following
curettage two years previously she developed precordial pain and palpitation.
This passed away and she had no further trouble until five weeks before
admission when she began to have severe epigastric and precordial pain;
palpitation, nocturnal dyspnea and orthopnea.

Physical Examination.— This showed an enlarged heart with an absolutely
irregular rhythm and a loud blowing systolic murmur at the apex. There was a
pulse deficit of 20 beats. A tender liver edge was felt. There was no edema
of the legs.

The urine on admission showed a trace of albumin which cleared up
subsequently. The leukocyte count was normal. The Wassermann reaction
was negative. The basal metabolism was plus 39 per cent. The temperature
was jiormal throughout. The day following admission the cardiac rhythm
became regular.

Course. — During her stay in the hospital she had several attacks of paroxysmal
auricular fibrillation, which diagnosis was confirmed by electrocardiograms.

852 ARCHIVES OF IXTERXAL MEDIC I\E

Roentgenograms were taken at a distance of seven feet during and after an
attack of fibrillation and showed no definite change in the size or contour of
the heart shadow. There was evidence of cardiac hypertrophy. The patient
improved slowly with rest in bed and with digitalis. She impressed some of those
who saw her as suffering from hyperthyroidism and others thought she had
mitral stenosis. She was discharged May 17, 1920, improved.

Diagnosis.— ^i'\ir^\ stenosis and regurgitation; paro.Kysmal auricular
fibrillation.

C.'^SE \\.— History.— \ man. aged 26, came to Dr. Paul D. White at the
Massachusetts General Hospital Dec. 14, 1920, complaining of palpitation. The
first attack occurred during the spring of 1916 while he was playing tennis.
For one hour his heart beat very tumultuously, rapidly and irregularly. Two or
three months later there was a second attack. He had numerous similar attacks
after this and the usual interval between them was two or three months, except
while he was driving a motor cycle in France, when they came weekly. He
was sent home as a case of effort syndrome. These upsets started and stopped
suddenly and lasted from one hour to more than a day. During the attack he
had slight dyspnea, weakness and marked palpitation, but he never had to stop
work.

Physical Examination. —The attack which brought him to Dr. White began
the previous day about 11 p. m. Electrocardiograms showed auricular fibrilla-
tion with a ventricular rate of from 110 to 130.

Course.— t^o effect was produced by right or left vagal or right ocular
pressure. December 16 he was seen again after the attack had stopped.
Examination of the heart disclosed no evidence of valvular disease. Roentgeno-
grams were taken at a distance of seven feet during and after the attack and
showed no appreciable change, although there was evidence of hypertrophy.

Diagnosis. — Paroxysmal auricular fibrillation.

A STUDY IX EXPERl.MEXTAL DIABETES

THE EFFECT OF INTRAVENOUS INJECTION OK I'ANCREATIC

PERFUSATES ON THE D/N RATIO FOLXOWING

PANCREATECTOMY *

HERBERT E. LAXDES, A.B., LESTER E. GARRISON, S.B.,
AND JAMES I. MOORHEAD. M.D.

The relation of the pancreas to diabetes mellitus was proved in
1889 when Von Mering and Minkowski discovered that fatal diabetes
always followed complete pancreatectomy. Two principal hypotheses
have been advanced to explain how the pancreas is linked with carbo-
hydrate metabolism, viz., the detoxication and the internal secretion
theories. In the former the pancreas is supposed to remove a toxic
substance from the blood stream, the presence of which interferes
with the oxidation of carbohydrates by the tissues ; in the latter the
pancreas is said to contribute something to the blood stream which
acts as a necessary link in the process of carbohydrate assimilation.
.\ttempts to establish either of these theories have met with little suc-
cess although the preponderance of evidence to date seems to favor
the theory of an internal secretion. Notwithstanding a vast amount
of experimentation the exact role of the pancreas in carbohydrate
metabolism remains unknown.

The parabiosis experiments by Forschbach,' the transplantation and
cross circulation experiments by Hedon.= pancreas feeding experiments
and many others have yielded results either negative or else contributory
alike to the detoxication and the internal secretion theories.

The blood transfusion experiments by Carlson and Ginsberg;* and
by Drennan; the study of experimental diabetes in pregnant dogs by
Carlson * and the experiments by Clark '■ in which the surviving mam-
malian heart and pancreas were perfused in circuit, all support the
theory of an internal secretion. These conclusions are also supported
bv the in vitro studies carried out by Colinheim " and Levene.'

♦From the Hull Laboratories of Physiological Chemistry and Pliarmacology.
Lniversity of Chicago.

1. Forschbach: Arch. f. exper. Path. u. Pharm. 9:131. 1908.

2. Hedon: Arch, intern, di. Physiol. 13:4. p. 255. 191.?.

3. Carlson and Ginsberg: .Am. J. Physiol. 36:280. 1915.

4. Carlson: .\m. J. Physiol. 28:391, 1911.

5. Clark, A. H.: J. Exper. Med. 24:621. 1916.

6. Cohnheim: Ztschr. f. Phvsiol. Cliem. 39:, 1903.

7. Levene. P. A., and Meyer. G. M. : .1. Binl. Chei<i.. WA.

854 ARCHIVES OF IXTERXAL MEDICIXE

Of these methods of attack, perfusion experiments seem most
promising, for if it can be shown that the pancreas contributes some-
thing to a physiologically inert solution passing through its vessels by
which utilization of sugar by the tissues is augmented, the theory of
an internal secretion would be made tenable. Besides the perfusion
experiments of Clark already mentioned, only two references to the
direct perfusion of the pancreas could be found. Hustin perfused
the pancreas in a study of its external secretion and de Meyer found
that by adding a solution previously perfused through a pancreas to
one subsequently perfused through a liver there was an increase in
liver glycogen.

In the experiments reported herewith an attempt has been made
to determine the effect of pancreatic perfusates upon dextrose utiliza-
tion in depancreatized dogs.

Fig. 1. — 1. Woodyatt pump. 2. Mercury manometer for measuring tile pres-
sure of the perfusate. 3. Air cushion. 4. Perfusion chamber. 5. Supporting
rod to which the pancreas is fastened. 6. Cannula for the pancreaticoduodenal
artery. 7. Water, bath maintained at 42 C. 8. Water bottle for saturating air
with moisture. 9. Bath for maintaining second water bottle at 75 C. In this
way the air entering the chamber was heated to from 30 to 35 C. 10. Cotton
filter for air entering apparatus. 11. Artificial lung composed of glass beads.

12. Small manometer for registering negative pressure in perfusion chamber.

13. Receiving bottle containing 150 c.c. of perfusion medium. 14. X'cssel con-
taining warm water used to warm the perfusate to body temperature previous
to injection. 15. Attachment for vacuum pump. 16. Woodyatt pump by which
perfusate was injected. 17. Cannula for the saphenous vein. The black tubing
carries air.

Method of Procedure. — The D/X ratio was taken as a criterion
for the carbohydrate consuming power of the animal. Dogs were
depancreatized and the D/N ratios were determined daily. When
these figures became rather constant for each animal, it was the plan
to inject intravenously a quantity of pure dextrose dissolved in
Tyrode's solution. Subsequent determinations of the D/N ratio

LA\DES-GARRIS0.\-M00RHEAD-EXPER1ME.\TAL DIABETES 855

should show the amount of dextrose retained after such injections.
It was then proposed to perfuse the pancreas of a normal animal with
Tyrode's solution containing dextrose, and to inject this perfusate
intravenously, at the same rate, into an experimentally diabetic ani-
mal, with the expectation that the subsequent D/N ratios would show
some change in dextrose utilization.

Females under morphin-ether anesthesia were used for removal
of the pancreas. The technic of pancreatectomy was developed so that
the complete extirpation of the gland could be assured. Asepsis was
given particular attention. The postoperative recovery of the ani-
mals was satisfactory and thev survived from five davs to five weeks.

dextr

n<=i

■

,075

.070

.065

\

.060

A

\

.056

\

\

.050

.045

,

.040

a

2

• b

/

f^

4

.0:^5

/

\

/

/

\

.630

/

\

/'

/

^

.0E5

y-

^

/

\/

-,

/

\

.020

"•'

/-

V

...

.015

i/

\

/

.010

\

/

.005

/

,

r J

> r

. .

V/

r 1

f 1

iCii

0 *

ijri

fSU

7. I

U^

Hours following panereatectotnj'.

Fig. 2. — Showing the effect on the sugar and nitrogen excretions of repeated
injections of pancreatic perfusates. Black line indicates dextrose. Dotted line
indicates nitrogen. Figures represent time of injection of perfusate.

Immediately following pancreatectomy the animals were placed in
clean metabolism cages and the urine was collected every twenty-four
hours. The dogs were kept on a constant daily diet of 480 gm. raw
lean meat. Water was given freely. The urine was collected at the
same time every day, alcoholic thymol being used as a preservative.
The volume was measured and the sugar and nitrogen determinations
were made. The total nitrogen was determined by the Kjeldahl-

856 ARCHIVES OF IXTERXAL MEDlCiXE

Gunning procedure. The Munson-Walker-Bertrand method was used
for the determination of the urine sugars.

The perfusion medium used in all experiments was Tyrode's solu-
tion modified so that it contained no dextrose. The formula was as
follows : Sodium chlorid, 0.7 per cent. : potassium chlorid, 0.02 per
cent. ; magnesium chlorid, 0.02 per cent. ; calcium chlorid, 0.02 per cent. ;
sodium bicarbonate, 0.01 per cent., and monobasic sodium phosphate,
0.005 per cent. Fresh solutions were always used.

The pancreas was perfused as follows: The peritoneal cavity of
a dog under morphin-ether anesthesia was opened aseptically by a
right rectus incision so as to expose the pancreas. The superior pan-
creatico-duodenal artery and vein were isolated and ligated close to
the origin of the former. The vessels were then cannulated so as to
send the perfusion medium through the pancreas in the direction of
the blood flow. To render the subsequent steps of the experiment
practically bloodless, the thorax was rapid!)- opened, the aorta clamped

"V"

Fig. J. — Showing effect of the intravenous injection of 1 c.c. of the freshly
prepared perfusate on the blood pressure of an etherized animal.

above the diaphragm and the chest cavity quickly closed. The appar-
atus for maintaining the circulation through the gland was designed
so that it could be sterilized in an autoclave. It consisted of a motor
driven syringe (Woodyatt transfusion pump) connected through a
manometer to the cannula in the pancreatico-duodenal artery. The
solution as it came from the pancreatico-duodenal vein poured over
glass beads through which a current of moist air was being drawn.
The perfusate then flowed into a closed bottle from which it was
drawn by the syringe and again sent through the gland. During the
procedure the dog was kept under light anesthesia. The perfusate
was maintained at 37 C. and a pressure of 120 mm. of mercury.
The gland at first became mottled and later white in color as the
perfusate continued tn pa<^ tlinnisjli iis \-e-sels. After fifteen or

LAXDES-GARRISOX-MOOKHEAD—EXPERIMEXTAL DIABETES 857

twenty minutes moisture was noticed, due to the escape of the per-
fusate from the surface of the pancreas, and in this way a few cubic
centimeters of the solution were lost. By the addition of small quan-
tities of Tyrode"s solution at intervals to compensate for this loss
on the surface of the gland, a continuous flow of Tyrode's solution
through the vessels could be maintained indefinitely.

One hundred and fifty cubic centimeters of Tyrode's solution con-
taining 10 gm. dextrose was allowed to pass through the gland for
one hour. At the end of this period a small sample was withdrawn
for analysis and the remaining portion injected by means of a Wood-
yatt apparatus at the rate of 10 c.c. per minute, directly into the
saphenous vein of a diabetic animal. Local anesthesia was used to
cannulate the vein. The concentration of dextrose in the perfusate
was determined in the sample withdrawn for that purpose, and the

Fig. 4. — Showing effect oi the intravenous injection uf 1 c.c. of the per-
fusate allowed to stand for forty hours at room temperature on the blood pres-
sure of an etherized animal.

quantity of sugar injected into the diabetic animal calculated. It was
the plan that dogs so treated should show a fairly constant D/N
ratio for a period of at least three days preceding the injection. Also
the quantity of sugar excreted in the urine following injections of
known quantities of pure dextrose was to be determined before injec-
tion with the perfusate containing dextrose. It was thought that the
pancreatic factor found present by Clark in his perfusates might
enable the diabetic animal to utilize all or a portion of the injected
dextrose.

Such experiments were attempted on a series of about fifteen
depancreatized dna< Xecronsy on the-^^ aiiinr>N revealed cnmplet-

858 ARCH WES OF IXTERXAL MEDICINE

removal of the pancreas in almost every instance. Death in a large
percentage of cases was caused by pneumonia. Table 1 shows that
Dog 8 was injected with 3.27 gm. pure dextrose seventeen days fol-
lowing pancreatectomy. For some unknown reason the daily sugar
excretion, preceding the day the animal was injected, rose from 10.54
gm. to 20.35 gm. The following day only 15.15 gm. sugar were
excreted in spite of the injection of 3.27 gm. dextrose. In no instance
did we feel justified in drawing conclusions as to dextrose utilization
from such fluctuating figures.

Having found this procedure useless in the solution of the prob-
lem, it was decided to attempt to influence the onset and course of
experimental diabetes by the intravenous injection of pancreatic per-
fusates alone. Such an experiment, if successful, might also throw
some light on the hypothetical internal secretion of the pancreas.

T.ABLE 1.-

—Experimental Results from

Observation

OF Dog 8,

Receiving

480 Gm. Raw Meat per D.\y, Plenty

OF

Water

.imouDt of Sugar,
Time Urine per Cent

Sugar

Total

Date*

:. Total

Nitrogen

D/N

April 3

10:00 a.m. 350 4.70

16.15

April 4

10:00 a.m. 470 2.30

11.28

'aoes

i.86

April 6

10:00 a.m. 490 2.40

11.76

11.270

1.05

April 6

10:00 a.m. 527 2.00

10.54

12.648

April 7

10:00 a.m. 260 Contaminated urine

April 8t

10:00 a.ui. 370 5.50

20.S5

9.250

2.20

April 9

10:00 a.m. 605 3.00

15.15

9.695

1.56

April 10

10:00 a.m. 265 4.80

12.72

7.950

1.60

April 11

10:00 a.m. 200 4.80

9.60

6.000

April 12

10:00 a.m. 185 4.60

7.79

7.326

l!06

♦ Complete pancreatectomy March
t Intravenous injection at 3:00 p.
Tyrode s solution. Fifteen minutes wi
cannidated under local anesthesia.

Clarke's work as previously quoted, indicates the existence of such
secretion in pancreatic perfusates.

The first line of experiments were conducted as follows : Pro-
ceeding on the theory that the animal would most likely respond to
treatment while the cells were in a normal physiologic state rather
than after they became altered by an abnormal carbohydrate metab-
olism, intravenous injections of pancreatic perfusates were made
immediately following pancreatectomy. In this way the onset of
experimental diabetes might be delayed.

The following technic was used in preparing these perfusates. The
abdomen was opened by an upper right rectus incision. The splenic
end of the pancreas was first liberated as far as the pylorus. The
duodenal end was then cut from the mesentery up to the point of its
attachment to the bowel wall. The sujicrior pancreatico-duodenal
artery was isolated and ligated close to its origin. The thorax was
opened and the aorta clamped immediatelx- above the diaphragm so

LANDES-GARRISOS-MOORHEAD—EXPERIMESTAL DIABETES 859

as to make the subsequent steps practically bloodless. The gland with
the attached piece of duodenum, 8 cm. in length, was quickly extir-
pated and washed in a warm, glucose- free, Tyrode's solution. The
pancreatico-duodenal artery was then cannulated and the gland sus-
pended by passing the supporting rod of the perfusion chamber through
the lumen of the attached bowel. The splenic end of the pancreas
was then tied by a thread to a loop in the supporting rod, and the
gland placed in the perfusion chamber. The cannula was connected
and the perfusion started. This procedure was carried out in from
five to seven minutes. A pressure not to exceed 120 mm. of mercury
was maintained and about 15 c.c. fluid was passed through the gland
per minute. The perfusate was maintained at i7 C.

g. dex
.065

trc

LHe^

—

—

—

—

—

—

—

—

—

—

—

—

—

—

.060

A

055

/

\

1

.050

/

^

s

.045

/

/

\

.040

(

\

f—

^

•n.

/

V

030

020

.015

_

y

_

,.

n:

"1.

^

21;

--

_^

^

—

—

—

—

.010
.005

—

j_

—

—

—

—

—

—

—

—

—

—

:!:.

=-^

1

H

:ur

2
s

J

fol

l07

'6
7in

J
S

oar

4
or

U

Dat

ec

toe

y.

e

^

Fig. 5. — Curve showing dextrose and nitrogen excretion for a period of
sixty-two hours following the injection of 175 c.c. pancreatic perfusate. The
perfusate was prepared by allowing the solution to flow through the gland once.
It was injected immediately into the animal following pancreatectomy. Solid
lines indicates dextrose. Broken line indicates nitrogen.

For these experiments a special eflfort was made to devise a per-
fusion apparatus which would approximate, as nearly as possible,
physiologic conditions. A diagram of this is shown and explained in
Figure 1. Following each experiment the apparatus was carefully
washed and filled with 75 per cent, alcohol. Immediately preceding
each experiment the alcohol was removed and repeated washings of
Tyrode's solution were pumped through the system. One hundred

860 ARCH 1 1 -ES OF JXTERXAL MEDICI XE

and fifty cubic centimeters of fresh, glucose-free Tyrode's solution
were then placed in the receiving bottle.

Twenty-seven dogs were used in this series of experiments. None
of the animals were fed after operation. The perfusate was allowed
to pass through the vessels of the pancreas for one hour. Small, white
areas appeared on the surface of the gland almost immediately after
the perfusion fluid had entered its vessels. These regions gradually
enlarged as the perfusion proceeded until the body and head of the
pancreas became white. Aside from this whiteness, due to the wash-
ing out of the blood, the tissue remained entirely normal in gross

mg. (
065

A

^

.060

/

y

.055

/

y

050

y

^

y'

.045

/

\y

.040

j^

.035

J

— ■

"

■~-

~~

___

1

.030

'

---

-

, 0?,5

\

.020

.015

.010

.005

10 20 30 40 50 60 70
Hours following pancreatectomy.

Fig. 6. — Curve showing dextrose and nitrogen excretion for a period of
seventy-six hours following injection of 100 c.c. pancreatic perfusate, prepared
as described below. Solid line indicates dextrose. Dotted line indicates
nitrogen. The perfusate was prepared by passing once through a pancreas
100 c.c. of defibrinated blood drawn from a diabetic dog. It was injected intra-
venously into a second animal immediately following pancreatectomy.

appearance. Peristaltic waves were often observed traveling along the
attached bowel, and these persisted throughout the time of perfusion.
While the gland was being perfused, a pancreatectomy was completed.
The perfusate was then injected, intravenously, into the depancreatized
dog by the Woodyatt apparatus at the rate of 10 c.c. per minute. This
operation was done asejjtically while the dog was still under ether

LA.\DES-G.-iRR]SO\-MOOKHEAD—EXPERlME.\TAL DIABETES 861

anesthesia. Following this the animal was at once placed in a metab-
olism cage and the urine collected as soon as possible after being
voided. The time of the appearance of glycosuria was noted and
quantitative sugar and nitrogen determinations were made from this
point until the death of the animal.

Tt

,080

075

I

1

065

j

.060

1

055

\

.050

\

.045

.040
.035

V

,

/

,

\

\

;

\

I

1—

'

.030

\

\

\

i'

.020

\

J

.015

\j

.010

005

15 30 45 60 75

)ur6 following
pancreat

jctomy.

Fig. 7.— Curve showing the dextrose and nitrogen excretion for a period of
ninety-two hours following injection with 150 c.c. pancreatic perfusate. The
perfusate was allowed to circulate through the gland for one hour and injected
immediately following the completion of pancreatectomy. Solid line indicates
dextrose. Dotted line indicates nitrogen.

It was found that the course of diahetes following complete removal of the
pancreas was fairly constant. Sugar invariably appeared in the urme from six
to eighteen hours after complete pancreatectomy and reached its maximum (trom
5 to 8 per cent.) in from twenty-four to forty-eight hours, .^side from the
glvcosuria, the usual diabetic .syndrome was noted, viz.: excessive thirst, poly-
phagia polyuria, failure of wounds to heal, rapid loss ot weight, progressive
weakness, and death from complications or extreme inanition.

8. Drennan. F. M.: .^m. J. Physiol. 2fr:.m 1911.

862

ARCHIVES OF INTERXAL MEDICIXE

Dogs injected immediately following pancreatectomy with 150 c.c.
of perfusate prepared as described, showed no delay in the onset of
glycosuria. There were, however, two exceptions in which sugar did
not appear in the urine of animals so injected until forty and fifty-
eight hours, respectively, following complete removal of the pancreas.
Glycosuria, on appearing, followed its usual course until a few days
later when the dogs died of pneumonia. Necropsy on these animals
revealed the duodenum to be smooth and free from bits of pancreas
visible to the naked eye. Both dogs were in early pregnancy, fetuses
measuring from 1 to 1.5 cm. being found.

]np n

^t1

rn

nP

055

050

/

\

045

/

\

1

—

^

040

\

/

\

/

0?55

\

/

\

/

030

\

/^

\

/

^'

025

\

/

r

/

020

>>

\

■

-^

=^

.^'

-/"

'-

^

/

015

\

/

'

'

010

'

005

12 13 14 15 16 17 18 19
Days following pancreateetom:-.

20 21

Fig. 8. — Showing glucose and nitrogen output per twenty-four hours on a
constant protein diet. Solid line indicates dextrose. Dotted line indicates
nitrogen. Arrow indicates intravenous injection of 3.27 gra. pure dextrose.

Glycosuria developed within from six to eighteen hours following
pancreatectomy in the remaining ten dogs of this series. As is noticed
in Table 2, subsequent injections after glycosuria had developed pro-
duced no consistent change in the D/N ratio. Figure 2 shows that
the urine sugar did not reach its maximum until 158 hours following
pancreatectomy. During this period the dog was injected three times
with pancreatic perfusates. The result is typical of those obtained in
this series. The delayed sugar excretion is probably apparent rather
than real, since it is always followed by a compensatory rise lasting
through a period of several hours. The total quantity of glucose
excreted for the given period remains unchanged.

LAXDES-GARRISOX-.MOORHEAD—EXPERIMEXTAL DIABETES 863

This interpretation seems justified by examination of the tracing
(Fig. 3) which shows the action of these perfusates on the blood
pressure of an etherized animal. Their marked vasodilator effect
indicates that they are in this respect similar to tissue extracts. It
is known that a fall in blood pressure resulting from the injection
of tissue extracts into experimentally diabetic animals is usually accom-
panied by a decrease in glycosuria followed by a compensatory rise.

T.'\BLE 2. — Experimental Results fro.m Observatiox of Doc 20*

July 12-13
July 13-14
July 14-15
July 15-16
July 16-17 §
July 17-18

of

Sugar.

Sugar.

Total

Nitrogen.

Total

Urine

per Cent.

Gm.

Sugar

Nitrogen

17.-.

4.568

170

4.29

7.298

4.875

30

2.83

0.849

8.142

0.843

5.718

7r,

3.31

2.483

118

4.64

5.475

7.434

50

2.74

1.370

6.*(5

3.150

10.584

65

4.42

2.873

, J-,,

105

4.33

4.517

3.476

3.757

2.814

50

6.72

3.360

14.537

10.097

80

4.80

3.84

3.84

2,176

2.176

* Pancreatectomy complete 12:45 p. m. July

5, 1919.

Injected immediately

following

pancreatectomy with 150 c.c.

pancreatic perfusate

prepared a

t Second injection 12:30 p. m. July 6.

: Third injection 12:45 p

m. July 7.

injection 9:00 p

m. July 16.

TABLE 3.-E.XPERI

MENTAL Results

FROM Observation

OF Dog

26*

A

imount of Sugar,

Sugar,

Total

Total

Date

Time

Urine per Cent.

Gm.

Sugar

Nitrogen

July 17

10:00 p.m.

450 1.30

5.896

5.896

4.158

1.42

m

395 4.33

17.1(M

July 18

8:35 p

m.

18.'-. 0.40

8.45 a

340 3.67

^^■SS

July 19

11:00 a

July 19

4:00 p

m

50 4..51

2.255

18.224

July 19

1:30 p.m.

35 4.06

pancreatic perfusate directly as it

The marked depressor action of the perfusates may be a factor in
explaining their apparent influence on glycosuria (E. (',. Kirk)." Fig-
ure 4 shows an increase in this property if the perfusate is allowed
to stand thirty hours, probably a result of autolysis caused by pan-
creatic digestive enzymes present in the .solution. However, throughout
the experiments no rise in temperature or visible signs of toxemia or
depression were noted.

9. Kirk. E. G.: .Arch. Int. Med. 15:39 (Jan.) 1915.

864 ARCHIVES OF ISTERSAL MEDIC IKE

It was suggested that the apparent negative results were due to a
decomposition of the internal secretion contained in the perfusate, and
that decomposition rather than concentration resulted from an hour's
perfusion of the gland. The findings of Clark as to the stability of
the secretion made this theory probable. Consequently, the solution

TABLE

4.— Experimental

Results

FROM Ob

,ERVATION

OF Dog

30

*

Amount ol

Sugar,

Sugar,

Total

Total

Date

Time

Crine

per Cent.

Gm.

Sugar

Xitrogeo

D/N

July 2.3

9:30 a.m

90

4.87

4 183

4.183

3.384

1.26

120

8.124

July 24

9:15 p.m

115

4.64

.5.336

13.460

9.823

1.37

80

6.68

5.344

July 25

8:15 p.m

40

2.61

1.014

6.388

4.392

1.45

• Pancreatectomy on Dog 30 was completed at 3:45 p. m., July 22, 1919, anil the animal
immediately injected with a perfusate prepared by passing deflbrinated blood of a diabetic
dog through the normal gland once.

-E.XPERIMENTAL RESULTS FROM ObSERV.^TIOX OF DoG 22''

of

Sugar,

Sugar,

Total

Nitrogen,

Total

Crine

per Cent.

Gm.

Xitrogen

Via

80

2.70

2.100

2.296

98

4.47

4.380

3.900

8.338

10.276

0.89

210

3.27

6.867

4.767

35

3.89

1.362

0.795

TABLE 6.— Dog 2 Was Fed 480 Gm. Raw, Lean Meat per D.>
AND Received Plenty of Water *

Amount of Sugar, Total
Urine per Cent. Sugar

Total

Date

Time

Xitrogen

D/N

April 26

10:00 a.m.

645 2.60 16.77

12.90

1.30

April 27

10:00 a.m.

427 Spilled by accident

10:00 a.m.

350 3.00 9.11

April 29

10:00 a.m.

227 2.30 5.22

4.994

April 30
May 1

270 3.10 8.37

6.B70

1.48

10:00 a.m.

70 2.92 2.04

May 2

10:00 a.m.

240 3.20 7.68

* Complete pancreatectomy April 16.

was passed once through the gland and injected immediately into
the depancreatized animal. The same technic and precautions were
observed as previously described.

Figure 5 shows the typical curve from sucJi an c.xiKrimciit.
(ilycosuria appeared si.x hours following ]KuicrcatecU)my and riisc to
its maximum within thirty hours. .As will be seen in 'I'ablc .S, the

LAX DES-GARR1S0\-M00RHEAD— EXPERIMENTAL DL4BETES 865

D/N ratios rose from 1.42 to 2.69 during the sixty hour period
immediately following pancreatectomy.

A third series of experiments was then attempted in which defi-
brinated blood from a depancreatized animal was used instead of
Tyrode's solution. It was hoped that this would be a better medium
for collecting the internal secretion from the gland. Defibrinated
blood from depancreatized animals had been previously shown by
Drennan to have no effect on the D/N ratio in experimental diabetes.

Under light anesthesia, blood was drawn aseptically from the
jugular vein of an' animal depancreatized three days previously. The
animal was diabetic, having a glycosuria of 3.36 per cent and a hyper-
glycemia of 1.22 mg. dextrose per c.c. of blood. The Munson-Walker
method for blood sugar determination was used. One hundred cubic
centimeters of this blood, after being defibrinated, was passed through
a normal pancreas and then injected into an animal immediately fol-
lowing pancreatectomy.

Figure 6 shows the dextrose nitrogen curves for this experiment.
There was apparently no effect on the onset of glycosuria. Table 4
shows a uniform rise of the D/N from 1.26 to 1.45 during a period
of seventy-five hours following injection.

CONCLUSIONS

1. As compared with the figures reported in the literature, espe-
cially by Lusk, the D/N ratios are uniformly low.

2. The percentage of sugar in the urine and the total sugar excre-
tion per day show wide variations both on a constant diet and during
starvation.

3. Although the total nitrogen excretion in some animals was
relatively constant, others under identical conditions showed marked
variations.

4. All factors within our control, such as diet, water, etc., did
not maintain a D/N ratio sufficiently constant to be of use in carrying
out the first series of experiments.

5. Perfusates prepared by passing glucose- free Tyrode's solution
for one hour through the blood vessels of a normal pancreas did not
delay the onset of diabetes.

6. Two exceptions to the above conclusion are recorded. It seems
unlikely, in view of the work of A. J. Carlson and co-workers on the
Control of Diabetes in Pregnancy, that the fetuses described in these
animals could have altered the course of diabetes. Since we were
unable subsequently to confirm these results, they are merely of
suggestive value.

866 ARCHirES OF IXTERNAL MEDICINE

7. Maximum elevation of glycosuria is delayed by injection of
pancreatic perfusates.

8. The influence on glycosuria is probably apparent and not real,
since it is followed by a compensatory rise. This effect may be
explained by the action of depressor substances shown to be present
in the perfusates.

9. Solutions passed once through the gland when injected intra-
venously showed no effect either on the onset or course of experimental
diabetes.

10. Defibrinated blood drawn from pancreatectomized animals and
passed once through the normal pancreas had no effect on experimental
diabetes.

11. The internal secretion of the pancreas, if it can be obtained by
the perfusion experiments described, may be too labile or in too low
concentration to influence the sugar metabolism of the diabetic animal.

We wisli to record our thanks to Dr. F. C. Koch for his many valuable sug-
gestions and his untiring interest in our work.

BOOK REVIEWS

THE EVOLUTION OF MODERN MEDICINE. A Series of Lectures
Delivered at Yale University on the Silliman Foundation in April, 1912,
by Sir William Osler, Bt., M.D., F.R.S.

Nearly ready for publication, the proofs partially corrected by Sir William,
the great war interrupted the final correction and completion of this volume.
The work was completed by Fielding H. Garrison. Harvey Gushing, Edward P.
Streeter, and Leonard L. Mackall, who have carried out the author's plans and
wishes in every respect. The volume represents Osier at his best — not only as
a physician but as a scholar, a man of letters and an historian — while the pub-
lishers have achieved a setting worthy of the content. Gomposed originally for
a lay audience and popular consumption, it will furnish inspiration to physicians
as well, tracing as it does the devious course of medical progress and struggle
to the present, and one needs no medical knowledge to follow the golden thread.
In the introduction he considers Egyptian, Assyrian, Babylonian and Oriental
medicine, with evidence of early operative and therapeutic procedures and
directions. Then Greek medicine with the contributions of both mythology and
the philosophers — medieval medicine with the isolated centers in universities
throughout Europe — the renaissance and development of anatomy and physiology
— modern medicine, and the rise of preventive medicine. Osier himself called
it "an aeroplane flight over the progress of medicine through the ages." But, in
spite of the necessary brevity, one gets an amazingly clear picture of the
individuals and the events, due partly to the numerous cuts and engravings,
themselves, the author's choice from an enormous bulk of extant material.

The volume is a delightful addition to the physician's library. With it he
can journey for a time away from his twentieth century setting to bygone times
and other places, returning with a happier perspective and a clearer understand-
ing of his own position in the evolution of medicine.

HUMAN PARASITOLOGY. By Damaso Rivas. Philadelphia and London:
W. B. Saunders Co., 1920.

This book is Written primarily as a textbook of the animal forms parasitic
in man, and while a large part is devoted to the classification and description
of these parasites, the author has added brief chapters on certain phases of
bacteriolog>', serology and laboratory diagnosis. These are not especially
attractive, as they, in themselves, are sufficiently comprehensive to be considered
in separate texts rather than as addenda to a textbook of the animal parasites
of man.

The first chapters of the book deal with the historical features of animal
parasitology, the relation of parasite to host and the effect of animal parasitism
on a host. Descriptions of the parasitic forms are complete. The classification
is strictly zoologic, and somewhat confusing, at least for the average physician.
However, it is readily understood by those with a broad zoologic training, and
the book, therefore, will appeal more to such physicians. The literature ref-
erences of a book now being offered to physicians as a text of the animal
parasites of man should include recent work in parasitology, especially as such
a book is intended for more than a laboratory guide.

PRECIS DE PARASITOLOGIE, par le Pr Guiart, professeur a la Faculte de
medecine de Lyon et a la Faculte de medecine de Cluj. Ed. 2, 1922, 575
pages, with 462 illustrations. (Bibliotheque du Doctoral en Medecine,
Gilbert et Fournier) (Librairie J.-B. Bailliere el fils, 19, rue Haulefeuille,
Paris).
This volume is one of the thirty-five composing the "Bibliotheque du

Doctorate en Medecine" collected under the supervision of Gilbert and Fournier.

868 ARCHIFES OF IXTERXAL MEDICIXE

In undertaking the work the editors state their reasons frankly. The sum of
the knowledge required today of the student and practitioner of medicine is
large, and increasing daily, and in the relatively short time devoted to the
acquisition of the knowledge necessary, there is a large amount of information
which can neither be covered in the classes nor remembered afterward, especially
the minutiae and less essential details of the subdivisions or specialties of
medicine. In undertaking to present the series, the editors have endeavored
to obtain the outstanding authorities in each division. The present second
edition is the volume on parasitology, the result of twenty-four years' specializa-
tion by the author. Professor Guiart. It is written for the student and physician,
contains only material useful to them, and retains the medical point of view in
the study of the parasites, followed by concise descriptions of the diseases
produced by them. The volume is profusely and clearly illustrated, well indexed,
and in every respect attains the editors' desire to present all that is indispensable
in the knowledge of parasitology in a clear, concise volume.

ZUR THERAPIE DES KARZINOMS MITT RONTGENSTRAHLEX. Vor-
lesungen Ueber Die Physikalischen Grundlagen Der Tiefentherapie. Von
Prof. Dr. Fr. Dessauer, Direktor Des Instituts Fiir Physikalische Grund-
lagen Der Medizin An Der Universitat Frankfurt A. M. 30 Illustrations.
Dresden and Leipzig: Theodor Steinkopff, 1922.

In the form of four lectures the author outlines the roentgen-ray therapy
of deep lying cancer by the use of the Roentgen rays of extremely short wave
length combined with copper filters, large areas and increased distance, so as
to obtain a high dose quotient at the depth of the disease. Only by this means
is an adequate dosage of rays administered to the cancer cells without injuring
the overlying tissues. The depth of ten cm. is used as a standard for com-
puting ray absorption, and in order to generate a beam rich in the rays of high
penetrating character a larger capacity tube must be used combined with an
electrical transformer delivering a current of much higher potential than has
been the practice in the, past. Two hundred thousand volts excites the tube,
and the beam is filtered through a copper filter of a thickness the major fraction
of a millimeter. This absorbs much of the heterogeneous discharge and allows
the passage of short wave lengths only. To obtain a sufficient dose the time
must be prolonged. Scattered radiation within the tissues is taken advantage
of by increasing the area of tissue treated. It is found that multiple small
portals of entry and many angles of cross-fire are not necessary. Much atten-
tion must be paid to the percentage of rays absorbed at various levels, however,
so that when cross fire dosage is given the proper summation of effects is
obtained throughout the pathologic mass. In malignancies located in the
middle of the body a heavy exposure is made from the front, back and sides
to build up in the center a combined dosage sufficient to destroy carcinoma. The
ratio of skin to depth dose, or dose quotient, is further increased by increasing
the tube-skin distance. The physics underlying these technical considerations
is explained and illustrated by many cuts, tables and graphs. Abstracts of
numerous other articles relating to deep therapy are appended.

THE LETHAL WAR GASES: PHYSIOLOGY AND EXPERIMENTAL
TRE.\TMENT. Frank P. Underhill, Ph.D., New Haven, Yale University
Press. 1920.

This octavo volume of 310 pages embodies a report of work conducted during
the late war under the author's direction by the section on Intermediary
Metabolism of the Medical Division of the Chemical Warfare Service, organized
originally by the Bureau of Mines. The work of- this section consisted largely
in exposing dogs during single or repeated periods of different time lengths to
chlorin, phosgene or chlorpicrin in known and varying concentrations. Thorough
and systematic examinations were made of the gassed animals under exactly
specified conditions as to the clinical symptoms and their course, of the mor-

BOOK REVIEWS 869

phologic and volume changes in the tissues and blood, the chemical changes in
the blood and urine, the respiratory function, the acid base balance, etc. The
findings are tabulated in detail together with the writer's analysis and inter-
pretation. Having established a standard for control, experiments were
performed with different therapeutic procedures.

One gathers from the data that the most notable gross changes observed con-
sisted of edemas, especially of the lungs, with a fall in the volume of the blood,
a rise in the blood concentration, and a failing circulation, this accounting,
according to Underbill, for those observed changes of the respiratory function
referable to oxygen starvation of the tissues with a fall of temperature and
finally suspension of vital activities. The immediate cause of death must, he
believes, be assigned to blood concentration. Acidosis, as measured by the
methods selected, was missed except in the later stages of the picture. The
therapeutic measures that yielded the most favorable results were those
directed toward increasing the blood volume and dilution, namely, bleeding
properly carried out and followed by a restoration of water, with alkali admin-
istration when acidosis was demonstrable. The work is notable for its thor-
oughness and accuracy. Apart from the value that it would have in the case
of future warfare with gases it represents a piece of inductive research of
exceptionally high order, and contains a mass of data and the description of
riiethods that may well find their application in the study of toxicology in
general.

THE MECHANICS OF THE DIGESTIVE TRACT. By Walter C.
Alvarez, M.D.. Assistant Professor of Research Medicine University of
California Medical School. Pp. 200. 22 Illustrations. Paul B. Hoeber,
New York City.

This monograph is written to defend or establish the following theses from
experimental and clinical data: (1) The normal and pathologic activities of
the intestine are due to a polarized gradient in the intestinal musculature.
There are no local reflex actions in the intestine itself. Auerbach's ganglionic
plexus functions only in connection with the extrinsic nerves. All the local
variations in action shown in different regions of the intestine are due to
change in the rhythmicity of the muscular gradient. (2) That many, if not
all, of the symptoms usually referred to a disordered intestine are caused by
reversal of the gastro-intestinal peristalsis.

The first seven chapters deal with the facts and speculations on which the
theory of muscular gradient of intestine action is based. Chapter IX deals
with the practical application of this theory to conditions of gastro-intestinal
disease in man. Chapter X deals with the supposed symptoms of reversed
peristalsis. Qiapter XI contains a critique of the current notions of "vagotonia"
and "sympatheticotonia." The last chapter deals briefly with experimental
methods, and the volume ends with a good bibliography of 450 titles.

The author is both a clinician and a laboratory worker. He has enthusiasm,
and a good knowledge of the literature. The book is, therefore, a valuable
compilation and treatise on the intestine, and is of interest to all medical men
and physiologists. While, in general, the author is scientific in the analysis of
facts and theories, in a few cases he becomes a special pleader and miscon-
strues facts in favor of his two main theses. He speaks of pieces of the
stomach and intestine as "strips of muscle," when, as a matter of fact, they are
strips of muscle, ganglia and nerves. On page 4 he says: "Most of us think
of the pupillary response to light as a complicated cerebral reflex." That
view may be held by the ignorant; the others know that in the mammal it is a •
midbrain, not a cerebral reflex.

On page 14 he says : ".Another function of the .\uerbach's plexus is probably
to make the intestinal muscles respond proi)erly to stimuli coming from the
underlying mucous memljrane. These stimuli are taken up by Meissner's
plexus and transmitted to Auerbach's by connecting fibres." This is in all
essentials a reflex act, and yet the author devotes a large part of the book to
disprove reflex action in the intestine itself.

870 ARCHIVES OF IXTERXAL MEDICIXE

Alvarez's recurring objection to local reflexes and the nervous action in the
motor activities of the intestine is that "nervous action is not really under-
stood." Granted! But is muscle contraction, or cell growth, or, in fact, any
fundamental physiologic phenomenon really understood today? We do not
understand the mechanism of memory, but must we, therefore, exclude memory
as a factor in animal behavior? Nobody denies today that visceral as well as
skeletal muscles, placed in certain artificial solutions in vitro, may be made
to contract rhythmically. But these activities of dying tissues in artificial
mediums have probably no relation to the normal activities of these tissues in
the intact animal. Theories built on such nonphysiologic experimentation may
help to explain, e. g., paralysis agitans or St. Vitus dance, but not normal
neuromuscular coordination.

The following symptoms, according to the author, are due to reverse peris-
talsis in the intestine: vomiting, regurgitation, heart burn, belching, nausea,
biliousness, coated tongue, foul breath, feeling of fulness soon after beginning
a meal, globus, hiccough. The interested reader must judge for himself how
successful the author is in these explanations. How is nausea and vomiting
started by the smell or sight of disgusting objects to be explained by reverse
peristalsis in the gut without admitting that the reversal of the peristalsis is
itself a nervous action? In the paragraph on "Globus," the author offers this
bit of physiologic speculation : "A few times in my life I have happened to
swallow while a wave of regurgitation was on its way up the esophagus, and
when the two waves met there was a painful tearing feeling. It may be that
globus is brought about in that way." Alvarez meets the objection that reverse
peristalsis is rarely seen even in marked gastro-intestinal motor disturbance by
the assumption that such peristalsis, too feeble to move the intestinal content
toward the stomach, or to be detected by the Roentgen ray, still produces
nausea and allied mental states. He fails to make clear why reversal of
peristalsis should cause central symptoms, while normal peristalsis does not,
or why the reversed peristalsis which is normal (.colon; duodenum) fails to
cause untoward effects. The medical practitioner will probably be most inter-
ested in the "practical application" in Chapter IX. At the end of this chapter
we read: "Some one may ask: In what way does the idea of a gradient altered
by disease influence our method of treatment? The answer is that so far little
has been done because therapeutists have not been thinking along these lines "
The author hopes for drugs to restore the normal gradient, but at present can
suggest only a smooth diet, a remedy that helps or fails, irrespective of the
theories of intestine mechanism.

INDEX TO VOLUME 29

PAGE

Adams, R. D., and Pillsbury, H. C. : Position and activities of diaphragm

as affected by changes of posture 245

Addis, T. : Blood pressure and pulse rate levels; levels under basal and day-
time conditions 539

Ames, O., Wearn, J. T., and Warren, S. : Length of life of transfused

erythrocytes in patients with primary and secondary anemia 527

Anemia, circulatory compensation for deficient oxygen carrying capacity of
blood in severe anemias ; G. Fahr and E. Ronzone 331

Aneurysm, mycotic embolic aneurysms of peripheral arteries ; deW. G.
Richey and W. W. G. Maclachlan 131

Aorta, a hitherto undescribed tumor of base of aorta ; G. R. Herrmann and

M. T. Burrows 339

Arsphenamin, chemical studies of blood and urine of syphilitic patients
under arsphenamin treatment, with a note on mechanism of early ars-
phenamin reactions ; C. Weiss and A. Corson 428

Arthritis, metabolism of; R. L. Cecil, D. P. Barr, and E. F. Du Bois 583

Auricular fibrillation, use of quinidin in; J. A. E. Eyster and G. E. Fahr.. 59

Auriculoventricular rhythm and digitalis; H. B. Richardson 253

Bacillus acidophilus and its therapeutic application ; L. F. Rettger and

H. A. Cheplin 357

Barker, B. I., and others: Clinical studies on respiration; effect of exercise
on metabolism, heart rate, and pulmonary ventilation of normal sub-
jects and patients with heart disease 277

Barr, D. P., Du Bois, E. F., and Cecil, R. L. : Clinical calorimetry;

metabolism of arthritis 583

Clinical calorimetry; temperature regulation after intravenous injection

of proteose and typhoid vaccine ■ ■ ■ 608

Du Bois, E. F., and Coleman, W. : Clinical calorimetry; metabolism

in erysipelas 567

Bell, E. T., and Hartzell, T. B.: Etiology and development of glomerulo-
nephritis ■ 768

Bigelow. G. H. : Intracutaneous reactions in lobar pneumonia 221

Blood, alkali reserve in pulmonary tuberculosis; D. S. Hachen 705

circulatory compensation for deficient oxygen carrying capacity of blood

in severe anemias ; G. Fahr and E. Ronzone 331

efHux of blood from carotid artery of dog and its expression by a gen-
eral empirical formula; H. L. Dunn 368

pigment metabolism and its relation to liver function; C. M. Jones 643

pressure and pulse rate levels; levels under basal and daytime condi-
tions ; T. Addis 539

pressure, efflux of blood from carotid artery of dog and its expression

by a general empirical formula ; H. L. Dunn 368

sugar curves following a standardized glucose meal ; W. H. Olmsted and

L. P. Gay .- • 384

transfusion, length of life of transfused erythrocytes in patients with
primary and secondary anemia; J. T. Wearn, S. Warren and O. Ames 527
Blumgart, H. L. : Antidiuretic effect of pituitary extract applied intra-

nasally in a case of diabetes insipidus 508

Boas, E. P.; Nature of so-called "capillary pulse" 763

Book Reviews :

Acute Epidemic Encephalitis (Lethargic Encephalitis) 401

Blood Supply of the Heart ; L. Gross 142

Diabetes: A Handbook for Physicians and Their Patients; P. Horowitz 401

Diagnostische Winke fiir die tagliche Praxis; E. Graetzer 554

Diseases of Digestive Organs, With Special Reference to Their Diag-
nosis and Treatment ; C. D. Aaron 402

Evolution of Disease ; J. Danysz 534

Evolution of Modern Medicine; W. Osier 867

IXDEX TO VOLUME 29

PAGE

Book Reviews — Continued

Human Parasitology : D. Rivas 867

La Genese de I'Energie Psychique : J. Danysz 141

Lethal War Gases : Physiologv and Experimental Treatment ; F. P.

Underhill '. 868

Mechanics of Digestive Tract ; W. C. Alvarez 869

Nutrition and Clinical Dietetics; H. S. Carter 141

Precis de Parasitologic ; P. Guiart 867

The Heart and the Aorta: Studies in Clinical Pathology; H. Vaquez

and E. Bordet 276

Zur Therapie des Karzinoms mit Rontgenstrahlen ; F. Dessauer 868

Buell, M. Van R. , Gibson. R. B., and Martin. F. T.: Metabolic study of
progressive pseudohypertrophic muscular dystrophy and other muscu-
lar atrophies 82

Burrows, M. T.. and Herrmann, G. R. : Tumor of base of aorta 339

Calorimetry, clinical; metabolism in ervsipelas: W. Coleman. D. P. Barr

and E. F. Du Bois 567

clinical ; metabolism of arthritis ; R. L. Cecil. D. P. Barr and E. F.

Du Bois 583

clinical ; temperature regulation after intravenous injection of proteose
and typhoid vaccine; D. P. Barr, R. L. Cecil and E. F. Du Bois 608

Capillary circulation, clinical observations on ; S. O. Freedlander and

C. H. Lenhart 12

pulse, so-called, nature of ; E. P. Boas 763

Cecil, R. L. ; Barr. D. P.. and Du Bois, E. F. : Clinical calorimetry ; metabo-
lism of arthritis 583

Clinical calorimetry; temperature regulation after intravenous injection
of proteose and typhoid vaccine 608

Cheplin, H. A., and Rcttger. L. F. : Bacillus acidophilus and its therapeutic

application 357

Christie, C. D.. and Stewart. G. N. : Study of diabetes insipidus, with
special reference to detection of changes in blood when water is taken
or withheld 555

Coleman. W. , Barr. D. P., and Du Bois, E. F. : Clinical calorimetry;

metabolism in erysipelas 567

Corson, A., and Weiss, C. : Chemical studies of blood and urine of syphi-
litic patients under arsphenamin treatment 428

Cutler, E. C, and Hunt, A. M. : Postoperative pulmonary complications.. 449

Diabetes, experimental diabetes ; effect of intravenous injection of pan-
creatic perfusates on D/N ratio following pancreatectomy; H. E.

Landes, L. E. Garrison and J. J. Moorhead 853

insipidus, antidiuretic effect of pituitary extract applied intranasally in

case of diabetes insipidus; H. L. Blumgart 508

insipidus, studies in diabetes insipidus, water balance and water intoxi-
cation ; J. F. Weir, E. E. Larson and L. G. Rowntree 306

insipidus, study of some cases of diabetes insipidus with special reference
to detection of changes in blood when water is taken or withheld ;

C. D. Christie and G. N. Stewart 555

nitrogen requirement for maintenance in diabetes mellitus ; P. L. Marsh,

L. H. Newburgh and L. E. Holly 9i

Diaphragm, position and activities of, as affected by changes of posture;

R. D. .^dams and H. C. Pillsbury .245

Digitalis and auriculoventricular ; H. B. Richardson 253

seat of emetic action of digitalis bodies; R. A. Hatcher and S. Weiss... 690
Du Bois, E. F., Barr, D. P., and Cecil, R. L. : Clinical calorimetry; tem-
perature regulation after intravenous injection of proteose and typhoid

vaccine 608

Clinical calorimetry ; metabolism of arthritis 583

—Coleman, W., and Barr, D. P.: Clinical calorimetry; metabolism in

erysipelas 567

Dunn, H. L. : Efflux of blood from carotid artery of dog and its expression

by a general empirical formula 368

Duodenum ulcer, studies of cause of pain in gastric and duodenal ulcers;
peristalsis as direct cause of pain in gastric ulcers with achylia and in
duodenal ulcers ; L. L. j. Hardt 684

JXDEX TO rOLUME 29

PAGE

Dyspnea, relation of dyspnea to maximum minute-volume of pulmonary
ventilation; C. C. Sturgis, F. W. Peabody, F. C. Hall and F. Fremont-
Smith, Jr 236

Dystrophy, muscular, a metabolic study of progressive pseudohypertrophic
muscular dystrophv and other muscular atrophies ; R. B. Gibson, F. T.
Martin and M. Van R. Buell 82

Electrocardiography, a convenient electrode for experimental electrocardio-
graphic work : C. S. Williamson 274

studies in variation of length oi Q-R-S-T interval; G. K. Fenn 441

Electrode for experimental electrocardiographic work; C. S. VVilliamson. . 274
Emphysema, tracheal and bronchial stenosis as causes for emphvsema ;

C. F. Hoover ' 143

Epplen, F. : Pathology of cirrhosis of liver 482

Erysipelas, metabolism in: W. Coleman, D. P. Barr and E. F. Du Bois... 567
Erythrocytes, length of life of transfused erythrocytes in patients Vi-ith

primary and secondary anemia; J. T. Wearn, S. Warren and O. Ames 527
Exercise, effect of exercise on metabolism, heart rate, and pulmonary
ventilation of normal subjects and patients with heart disease ; ¥. W.

Peabody, C. C. Sturgis, B. 1. Barker and M. N. Read 277

Eyster, J. A. E., and Fahr, G. E. : Quinidin in auricular fibrillation 59

Fahr, G. E., and Eyster, J. A. E. : Quinidin in auricular fibrillation 59

— and Ronzone, E. : Circulatory compensation for deficient oxygen carry-
ing capacity of blood in severe anemias 331

Fenn, G. K. : Variation of length of Q-R-S-T interval 441

Fineman, S. : Microlymphoidocytic leukemia 168

Freedlander, S. O.. and Lenhart, C. H. ; Capillary circulation 12

Fremont-Smith. F.. Jr.. and others: Clinical studies on respiration; rela-
tion of dyspnea to maximum minute-volume of pulmonary ventilation 236

Garrison, L. E. , Moorhead, J. J., and Landes, H. E. : A study in experi-
mental diabetes; effect of intravenous injection of pancreatic perfusates
on D/X ratio following pancreatectomy 853

Gay, L. P., and Olmsted, W. H. : Blood sugar curves following a stand-
ardized glucose meal 384

Gibson, R. B., Martin, F. T., and Buell, M. Van R. : Metabolic study of
progressive pseudohypertrophic muscular dystrophy and other muscu-
lar atrophies 82

Glomerulonephritis, etiology and development of; E. T. Bell and T. B.
Hartzell 768

Glycosuria, renal; D. S. Lewis 418

Golden, R., and Levine, S. A. : Paroxysmal rapid heart action with special

reference to roentgen-ray measurements of heart in and out of attacks 836

Greene, L. W.. and Whitman, R. C. : Disseminated miliary tuberculosis in
a still-born fetus 261

Hachen. D. S. : Alkali reserve in pulmonary tuberculosis 705

Hall, F. C, and others : Clinical studies on respiration ; relation of dyspnea

to maximum minute-volume of pulmonary ventilation 236

Hambrecht, L.. and Xuzum, F. R. : Correlated study of indications for

tonsillectomy and of pathology and bacteriology of excised tonsils... 635

Hardt, L. L. J.: Studies of cause of pain in gastric and duodenal ulcers;
peristalsis as direct cause of pain in gastric ulcers with achylia and
in duodenal ulcers 684

Hartzell, T. B., and Bell, E. T. : Etiology and development of glomerulo-
nephritis. 768

Hatcher. R. A., and Weiss, S. : Seat of emetic action of digitalis bodies.. 690

Heart disease, effect of exercise on metabolism, heart rate, and pulmonary
ventilation of normal subjects and patients with heart disease; F. W.

Peabody. C. C. Sturgis, B. I. Barker and M. N. Read 277

reversed rhythm of ; M. H. Kahn 828

tuberculosis of. with report of 2 cases; E. Weiss 64

Hemoglobin metabolism in paroxysmal hemoglobinuria, with observations
on extrahepatic formation of bile pigments in man; C. M. Jones and
B. B. Jones 669

IXDEX TO I'OLUME 29

PAGE

Hemoglobin — Continued
pigment metabolism and regeneration of hemoglobin in body; G. H.

Whipple 711

Hemoglobinuria, paro.xy.smal, hemoglobin metabolism in, with observations

on extrahepatic formation of bile pigments in man ; C. M. Jones and

B. B. Jones 669

Herrmann. G. R., and Burrows, M. T. : Tumor of base of aorta 339

Hewlett, A. W., and Jackson, N. R. : Vital capacity in group of college

students 515

Holly, L. E., Marsh, P. L., and Newburgh, L. H. : Nitrogen requirement

for maintenance in diabetes mellitus 97

Hoover, C. F. : Tracheal and bronchial stenosis as causes for emphysema 143
Hunt, A. M., and Cutler, E. C. : Postoperative pulmonary complications.. 449

Jackson, N. R., and Hewlett, A. W. : Vital capacity in group of college

students 515

Jones. B. B.. and Jones, C. M. ; Study of hemoglobin metabolism in parox-
ysmal hemoglobinuria, with observations on extrahepatic formation of
bile pigments in man 669

Jones, C. M. : Blood pigment metabolism and its relation to liver function 643

— and Jones, B. B. : Study of hemoglobin metabolism in paro.xysmal hemo-
globinuria, with observations on extrahepatic formation of bile pig-
ments in man 669

Kahn, M. H. : Reversed rhythm of heart 828

Kline, B. S., and Oppenheimer, B. S. : Ochronosis 732

Landes, H. E. , Garrison. L. E., and Moorhead. J. J. : A study in experi-
mental diabetes: effect of intravenous injection of pancreatic perfu-
sates on D/N ratio following pancreatectomy 853

Larson. E. E. , Rowntree, L. G., and Weir,' J. F. : Studies in diabetes

insipidus, water balance and water intoxication 306

Lenhart, C. H., and Freedlander, S. O. : Capillary circulation 12

Leukemia, microlymphoidocytic, with report of case: S. Fineman 168

Levine, S. A., and Golden, R. : Paroxysmal rapid heart action with special
reference to roentgen-ray measurements of heart in and out of attacks 836

Lewis, D. S. : Renal glycosuria 418

Liver, cirrhosis of, pathology of, an historic-pathologic study: F. Epplen 482

function, blood pigment metabolism and its relation to; C. M. Jones... 643

Lung, postoperative pulmonary complications ; E. C. Cutler and A. M. Hunt 449

McCann, W. S. : Protein requirement in tuberculosis 33

McClure, C. W.. and Reynolds, L. : Motor phenomena occurring in nor-
ma! stomachs, in presence of peptic ulcer and its pain, as observed
fluoroscopically 1

Maclachlan, W. W. G., and Richey, deW. G. : Mycotic embolic aneurysms
of peripheral arteries 131

Marsh, P. L. . Newburgh. L. H.. and Holly, L. E. : Nitrogen requirement
for maintenance in diabetes mellitus 97

Martin, F. T. , Buell, M. Van R.. and Gibson, R. B.: Metabolic study of
progressive pseudohypertrophic muscular dystrophy and other mus-
cular atrophies 82

Marvin, H. M., and White. P. D. : Paro.xysms of tachycardia 403

Metaliiilism. aids to basal metabolic rate determinations; H. S. Newcomer 748

in erysipelas; W. Coleman, D. P. Barr and E. F. Du Bois 567

of arthritis; R. L. Cecil, D. P. Barr and E. F. Du Bois 583

Mcthvl alcohol poisoning, biochemical studies in a fatal case of; L M.
Rabinovitch 821

Moorhead. J. J. , Landes, H. E., and Garrison, L. E. : A study in experi-
mental diabetes ; effect of intravenous injection of pancreatic perfu-
sates on D/N ratio following pancreatectomy 853

Muscular atrophy, a metabolic study of progressive pseudohypertrophic
muscular dvstrophy and other muscular atrophies; R. B. Gibson, F. T.
Martin and M. Van R. Buell 82

Nephritis, glomerular, etiologv and development of : E. T. Bell and T. B.
Hartzell .' 768

IXDEX TO rOLUME 29

PAGE

Nervous system syphilis, observations following intravenous injections of

hypertonic salt solutions in cases of neurosyphilis; J. Wynn 72

Newburgh. L. H., Holly, L. E., and Marsh. P. L. : Nitrogen requirement

for maintenance in diabetes mellitus 97

Newcomer, H. S. : Aids to basal metabolic rate determinations 748

Nitrogen requirement for maintenance in diabetes mellitus; P. L. Marsh,

L. H. Newburgh and L. E. Holly 97

Nuzum, F. R., and Hambrecht, L. ; Correlated study of indications for

tonsillectomy and of pathology and bacteriology of excised tonsils... 635

Ochronosis, with a study of an additional case ; B. S. Oppenheimer and
B. S. Kline 732

Olmsted, \V. H., and Gay, L. P. : Blood sugar curves following a stand-
ardized glucose meal 384

Oppenheimer, B. S., and Kline, B. S. : Ochronosis 732

Pancreatectomy, effect of intravenous injection of pancreatic perfusates
on D/N ratio following pancreatectomv ; H. E. Landes. L. E. Garrison
and J. J. Moorhead 853

Peabody, F. \V., and others ; Clinical studies on respiration ; relation of

dyspnea to maximum minute-volume of pulmonary ventilation 236

— and others : Clinical studies on respiration ; effect of exercise on
metabolism, heart rate, and pulmonary ventilation of normal subjects
and patients with heart disease 277

Pigment metabolism and regeneration of hemoglobin in body; G. H. Whipple 711

Pillsbury, H. C, and Adams, R. B. : Position and activities of diaphragm

as affected by chan.ges of posture 245

Pituitary extract, antidiuretic effect of pituitary extract applied intra-

nasally in case of diabetes insipidus; H. L. Blumgart 508

Pneumonia, lobar, intracutaneous reactions in; G. H. Bigelow 221

Protein requirement in tuberculosis : W. S. McCann 33

therapy, temperature regulation after intravenous injection of proteose
and typhoid vaccine; D. P. Barr, R. L. Cecil and E. F. Du Bois.... 608

Pulse, blood pressure and pulse rate levels; levels under basal and day-
time conditions ; T. Addis 539

nature of so-called "capillary pulse"; E. P. Boas 763

Q-R-S-T, studies in variation of length of Q-R-S-T interval ; G. K. Ftnn 441
Quinidin in auricular fibrillation: j. A. E. Eyster and G. E. Fahr 59

Rabinovitch, I. M. : Biochemical studies in a fatal case of methyl alcohol

poisoning 821

Read, M. N.. and others: Clinical studies on respiration; effect of exercise
on metabolism, heart rate, and pulmonary ventilation of normal sub-
jects and patients with heart disease 277

Respiration, clinical studies: relation of dyspnea to maximum minute-
volume of pulmonary ventilation ; C. C. Sturgis, F. W. Peabody, F. C.

Hall and F. Fremont-Smith, Jr 2.36

clinical studies, effect of exercise on metabolism, heart rate, and pul-
monary ventilation of normal subjects and patients with heart disease ;
F. W. Peabody, C. C Sturgis, B. I. Barker and M. N. Read 277

Rettger, L. F., and Cheplin, H. A.: Bacillus acidophilus and its therapeutic
application 357

Reynolds, L., and McClure, C. W. : Motor phenomena occurring in normal
stomachs, in presence of peptic ulcer and its pain, as observed
fluoroscopically :■;■■■. '

Richardson, H. f^. : Auriculoventricular rhythm and digitalis 253

Richey, deW. G., and Maclachlan, W. W. G.: Mycotic embolic aneurysms
of peripheral arteries ' 131

Ronzone, E., and Fahr, G. : Circulatory compensation for deficient oxygen
carrying capacity of blood in severe anemias 331

Rowntree, L. G. , Weir, J. F., and Larson, E. E. : Studies in diabetes

insipidus, water balance and water intoxication 306

IXDEX TO VOLUME

PAGE

Stewart, G. N., and Christie. C. D. : Study of diabetes insipidus with special
reference to detection of changes in blood when- water is taken or
withheld 555

Stomach, motor phenomena occurring in normal stomachs, in presence of
peptic ulcer and its pain, as observed fluoroscopically ; L. Remolds

and C. W. McClure .' " 1

ulcer, studies of cause of pain in gastric and duodenal ulcers ; peristalsis
as direct cause of pain in gastric ulcers with achylia and in duodenal
ulcers ; L. L. J. Hardt 684

Sturgis, C. C. and others : Clinical studies on respiration ; relation of

dyspnea to maximum minute- volume of pulmonary ventilation 236

— and others: Clinical studies on respiration: eflect of exercise on
metabolism, heart rate, and pulmonary ventilation of normal subjects
and patients with heart disease 277

Surgery, postoperative pulmonary complications; E. C. Cutler and A. M.
Hunt 449

Syphilis, chemical studies of blood and urine of syphilitic patients under
arsphenamin treatment, with a note on mechanism of early arsphena-

min reactions ; C. Weiss and A. Corson 428

observations following intravenous injections of hypertonic salt solu-
tions in cases of neurosyphilis ; J. Wynn 72

Tachycardia, observations on paroxysmal rapid heart action with special
reference to roentgen-ray measurements of heart in and out of attacks ;

S. A. Levine and R. Golden 836

paroxysms of, observations on; H. M. Marvin and P. D. White 403

Temperature regul - 'on after intravenous injection of proteose and typhoid

vaccine; D. ^ , ,rr, R. L. Cecil and E. F. Du Bois 608

Tonsils, correlr.'i -iidy of indications for tonsillectomy and of pathology

and bactf: i , .,:' excised tonsils; L. Hambrecht and F. R. Nuzura 635
Tuberculosif . iit, , ii-seminated, in still-born fetus; R. G. Whitman and

L. W. 261

of hear' ? cases ; E. Weiss 64

proteii X. S. McCann • 33

pulm> :. n ; D. S. Hachen 705

Turn' ■ I tumor of base of aorta ; G. R. Herrmann

339

T egulation after intravenous injection of

L. Cecil and E. F. Du Bois 608

students; A. W. Hewlett and N. R.

Weir, '•

insi
Weiss. (.

litic
Weiss, E.
Weiss. S..
Whipple, G.

bodv

White. P. D..
Whitman, R. L

in still-born
Williamson, C. .

graphic work
Wvnn, f.; Intravei

syphilis

T, : Length of life of transfused

y and secondary anemia

' ). : Length of life of transfused

and secondary anemia

■e, L. G. : Studies in diabetes

ioxication

of blood and urine of syphi-

tion of digitalis bodies.,
•ration of hemoglobin in

ichycardia

•, tj miliary tuberculosi

>ntal electrocardio-
utions in ncuro-

527

527

306

428
64
690

711
403

261

274

72

Arctaires of Internal medicine

STOi^AGE