Full text of "American journal of physiology"

See other formats

rg)
inte:
rei

‘

af ay
fet suit
F wtadstet rye

pedyey
Lp md ve
TP be pocgd oe
oe ie Saree)

hy
Ws ‘ La
Wi aeduete Bani 4

SON

i Wine kat a
it,

i
iy,

oa WOE

rt
it wih
Mies i fA “

Lath

tf Gy me
¢ oH is oe

.
Apgrouee rt
fy} Wena

Fe kt o §
Alte,

ake

r+, ai

a
;
i

oe oa

THE AMERICAN JOURNAL
OF

PHY SIOLOGY-

7p Sie

QU > aaa oe

ia}

Bs ee dhe a

—e

; i se ae

= smn
a i

o wa ‘
ae ope ‘eh if ae
@ ‘ » a
é ¢ ee : : bi te ; i
»
a” cn ' th pers 1
‘ Ae
4? ’
*. *
4
ha
; _
’

EDITED FOR

Dds

THe American Physiological = T

‘EPLE

AMERICAN JOURNAL

Pry SLO weezy

VoLuME XXVIII.

BOSTON; U2 SA.

IQIO-IQII.

Ce Nh EN bs:

PAGE
PROCEEDINGS OF THE AMERICAN PHYSIOLOGICAL SOCIETY ....... 1x
\ No. I, NOVEMBER 1, IgI0O.
Tue INFLUENCE OF ALCOHOL UPON NITROGENOUS METABOLISM IN MEN
AND Animats. By Lafayette B. Mendel and Warren W. Hilditch . . . I
A StupDy OF THE ISOLATED KIDNEY. — THE INFLUENCE OF PULSE PRESSURE
Poo ENAL MONCHON: BY DOR. He0RG os me ss ye es OK
SENSORY CHANGES IN THE SKIN FOLLOWING THE APPLICATION OF LOCAL
ANESTHETICS AND OTHER AGENTS. —I. EtHyt CHLorIDE. By Shep-
herd tnory Pranzand William C. Ruédiger 3. woe ww ee ee RG
CONCERNING THE SECRETION OF THE INFUNDIBULAR LOBE OF THE PITU-
ITARY BoDY AND ITS PRESENCE IN THE CEREBROSPINAL FLuIp. By
Humey Cushing and Emil Gocisch. . 0 6 ee ew ee «| O
OBSERVATIONS ON AURICULAR STRIPS OF THE Cat’s HEART. By Joseph
ei eRe a eras ty Col St hy A aie oid 2s) gat ede Oe
A COMPARISON OF THE TOTAL NITROGEN EXCRETION OF EITHER KIDNEY
IN NorMAL INDIVIDUALS DURING VARYING PERIODS OF Time. By
Theodore B. Barringer, Jr., and Benj. S. Barringer... . ©. « » . . «= TI9
SOME OBSERVATIONS ON THE PRODUCTION OF LIGHT BY THE FIREFLY. by
wosene HH. Kase ond. Alex. Welermoi 2 ce. 2g. 4 se a Be 2 E22
ACAPNIA AND SHOCK. — VII. FAILuRE OF THE CircuLaTION. By Yandell
ECU ILER SOIL Emre a MCG ee a Np eae hy ee ore ee | ED
METABOLISM OF DEVELOPMENT. — II. NITROGEN BALANCE DURING PREG-
NANCY AND MENSTRUATION OF THE Doc. By John R. Murlin. . . 177
No. II, DECEMBER 2, IgI0.
On THE NERVOUS MECHANISM OF THE RIGHTING MOVEMENTS OF THE STAR-
intent NR OUEr = rt Ye eee A ee gag

THE Errect oF LESIONS OF THE DorsAL NERVE Roots ON THE REFLEX
EXCITABILITY OF THE SPINAL Corp. By Clyde Brooks ...... 212

v1 Contents.

PAGE

A QUANTITATIVE STUDY OF FARADIC STIMULATION. — V. THE INFLUENCE OF
TIsSuE RESISTANCE AND OF KATHODE SURFACE ON STIMULATING
Errectiveness. By E.G: Maritim... . 3... \.. 2) ee

ON THE Dynamics or CELL Diviston. —II. CHANGES IN PERMEABILITY
oF DEVELOPING Eccs To ELtEctroLtytTES. By J.F. McClendon .. . 240

Tue RELATION OF AFFERENT IMPULSES TO THE VASOMOTOR CENTRES. By
WT. Porter, . 26 6 ons ns BS ls

No. III, JANUARY 2, IgrI.

Tue Puysrotocy oF CreLtt Division. —III. Tue Action oF CALctium
SALTS IN PREVENTING THE INITIATION OF CELL DIVISION IN UN-
FERTILIZED EGGS THROUGH ISOTONIC SOLUTIONS OF SopIUM SALTS.

By RS. dalle. a ee es er

EFFECTS OF PRESSURE ON CONDUCTIVITY IN NERVE AND Muscre. By
Walter J. Meck and W. E. Leaper . . ... . . ... . re

THE EFFECTS OF STRETCHING THE NERVE ON THE RATE OF CONDUCTION
OF THE NERvouS Imputse. By As J. Carlson. . : |. 30) See

THE PRODUCTION OF GLYCOSURIA BY ADRENALIN IN THYROIDECTOMIZED
Docs. By Frank P.Underal 9.) 5... ss re

No. IV, FEBRUARY I, IQII.

STUDIES IN EXPERIMENTAL GLycosuRIA. — VI. THE DISTRIBUTION OF
GLYOGEN OVER THE LIVER UNDER VARIOUS CONDITIONS. Post Mor-

TEM GLYCOGENOLYsIS. By J. J. R. Macleod and R.G. Pearce... . 341

THE MeErTAsorism oF Docs witH FUNCTIONALLY RESECTED SMALL IN-
TESTINE. By Frank P.Underlill .. . . . . ... . eee

THE INFLUENCE OF THE PRECEDING DIET ON THE RESPIRATORY QUOTIENT
AFTER ACTIVE DIGESTION HAS CEASED. By Francis G. Benedict, L. E.
Emmes, and J. A. Rithé. . 2... 2 ow. ee

THE RESPIRATORY EXCHANGE AS AFFECTED BY Bopy Position. By L. E.
Emmes and J. A. Riche. . 5... . . 2

No. V, MARCH 1, Torr.

CONTRIBUTIONS TO THE PHYSIOLOGY OF REGENERATION. — V. REGENERA-

TION OF ISOLATED SEGMENTS AND OF SMALL Pieces OF Worms. By
Sergius Morgulis of PAS

Contents. Vil

PAGE

THE INFLUENCE OF COLD BATHS ON THE GLYCOGEN CONTENT OF MAn. By
ie area IC Mths wey 2 EIA aye rand lO tar, ae ieee Bi nen

On Nucremn METABOLISM IN THE Doc. By P. A. Levene and F. Medi-
name eres See ALS a oe tis Me ge Sport els, NARA aes tS MSS

THE Errects oF VARIOUS ForMS OF EXERCISE ON SysToLic, DIASTOLIC,
AND PULSE PRESSURES AND PULSE RATE. By Oswald S. Lowsley . . .° 446

ON THE QUESTION WHETHER DEXTROSE ARISES FROM CELLULOSE IN
PIBUSRON by Granam Lush. SSF 6 ke dba dlte ele « 2) AGF

eRe ME Mew avec, sh ol sl cisdban. 1318 sa eae. were 2 ea, ss «ACO

Peer E DINGS OF THE AMERICAN PHYSIO-
LOGICAL, SOCIETY,

TWENTY-THIRD ANNUAL MEETING,

NEw HAVEN, DECEMBER 28, 29, and 30, IgI0.

PROCEEDINGS OF THE AMERICAN PHYSIOLOGICAL
SOCIETY.

ON LOCALIZED, CONTRACTION IN SKELETAL MUSCLE.
By GERTRUDE FRANCES BARBOUR AND Percy G. STILES.

WE have attempted to determine whether the gastrocnemius of the
frog contains any localized reserve of inactive tissue when it is moder-
ately contracted, the stimulation being reflex or spinal. To this end
we have secured simultaneous records of the movement of the ten-
don of Achilles and of various points on the surface of the muscle. It
seems clear that in smaller contractions there is relatively less partici-
pation by the upper third of the muscle than is the case in maximal
movements, or, in other words, that there is, near the femoral at-
tachment, a reserve of fibres rather difficult to excite through the
spinal centres.

THE SOURCE OF THE IMMUNE BODIES IN THE LYMPHS.
By F. C, Brest anp A. 8B. LuckKHArpr.

AFTER having established the facts as regards the concentration of
antibodies in normal and in actively and passively immunized animals,
the problem of the passage of the antibodies from the blood to the
lymphs and other body fluids was begun.

The method employed was to immunize a dog by intravenous in-
jections with a foreign blood, until a high state of immunity was
established. Samples of cervical and thoracic lymph were collected,

- then the animal was cross-circulated for from seven to fifteen minutes
xi

xii Proceedings of the American Physiological Society.

with a normal animal, from which samples of both the lymphs and
blood had already been drawn. A sample of blood was drawn from
each immediately after cross-circulation and then in most cases the
actively immunized animal was killed. Samples of lymph and blood
were then collected from the passively immunized animal under light
ether anesthesia at regular intervals for several hours. The blood
was centrifugated after being defibrinated by whipping, and the
lymphs were defibrinated immediately after collection. Cerebio-
spinal fluid and aqueous humor were collected at the end of the ex-
periment only. The hemolytic, agglutinating, and opsonic power of
the fluids was then tested.

We find that the antibodies — hemolysins, agglutinins, and opso-
nins — pass at about the same rate from the blood into the lymphs, but
they make their appearance in a shorter time in the thoracic than in the
cervical lymph. They are nearly always in higher concentration in
the former than in the latter, although the reverse may be true ccca-
sionally after the experiment has been in progress for several hours.
The antibodies hardly pass into the cerebrospinal fluid at all. The
same is true of the aqueous humor. The concentration of the anti-
bodies in the various body-fluids in the animal rendered passively
immune by this method soon reaches an equilibrium, which is the
same as that in the actively immunized animal of the same degree of
immunity. From our experiments we have concluded that the source
of the antibodies of the lymphs is the blood, and that the antibodies
obey the laws of lymph formation as do the other constituents of the
lymphs.

SOME OBSERVATIONS ON THE NATURE OF GASTRIC
PERISTALSSS.

By W. B. CANNON.

Gastric peristalsis is similar to antiperistalsis in the colon in not
being stopped by nicotine, in not being accompanied by a forerunning
inhibition, in being rhythmically repeated, and in starting at a pulsat-
ing ring. Pulsations start at once if the organ, while in a state of
tonic contraction, is distended, — they seem therefore to result from
tension. Intragastric pressure measures the degree of tension. If

Twenty-third Annual Meeting. Xili

the pressure is increased, the pulsating ring is moved towards the
pylorus; if diminished, towards the cardia. At first, therefore, the
waves do not originate in any special region; later they start at a pul-
sating tonus ring, as in the colon.

Interruption of the myenteric plexus by several incisions, through
both muscular coats, passing entirely around the stomach, does not
stop the passage of the waves.

The necessary tonus is probably given initially by the vagi, and
later maintained intrinsically by the stomach; for cutting the vagi
before digestion begins greatly delays the appearance of peristalsis,
whereas cutting the nerves after digestion has begun does not affect
peristalsis. The recovery of motor activity after vagus section is
attended by the development of independent tonus.

THE RECEPTIVE RELAXATION OF -THE STOMACH.
By W. B. CANNON AND C. W. LIEB.

From two to four and a half seconds after a cat swallows, the intra-
gastric pressure which prevails during gastric digestion begins to fall.
The lowest point of pressure, less than 1 cm. of water, is reached be-
tween six and ten seconds after the fall begins. In a few seconds
more pressure begins to rise again and is soon at the former height.
With a continued pressure in the gastric content between 3 and 4 cm.
of water, the increase of gastric volume may be as much as 8 or Io C.c.
during the period of relaxation. ‘Repeated swallowings keep the
pressure at the lowest point. The phenomenon disappears if both
vagi are cut.

The relaxation is at its maximum when the esophagus would natur-
ally propel a bolus into the stomach.

ATTEMPTS TO PRODUCE EXPERIMENTAL HYPER-
THYROIDISM.

By A. J. Cartson, J. F. Rooxs, ano J. F. McKie.

- THE work is primarily a search for tests for the thyroid secretions
in the body fluids in order to determine whether in hyperthyroidism

xiv Proceedings of the American Physiological Society.

experimental and clinical) there is excess of these secretions in the
body fluids or in the tissues.

Desiccated sheep and dog thyroids were fed to groups of pigeons,
chickens, ducks, mice, rats, guinea pigs, rabbits, cats, foxes, dogs,
and one monkey.

1. The different genera exhibit great variations in their resistance
to thyroid feeding. Dogs, cats, foxes, and ducks seem the most
resistant, doses of 20 gm. per day to dogs using from 3 k. to 7 k.
producing no symptom, save increased appetite.

2. When thyroid is fed in sufficient quantity, the most constant
symptoms produced are emaciation and diarrhea. Both of these
symptoms may, however, be absent even in fatal cases. Death in
convulsions or in depression.

3. The most constant post mortem findings are hyperemia of the
intestines with hemorragic infiltration of the intestinal mucosa.

4. Exophthalmus and nervousness cannot be detected. Many of
the animals continue to eat until a few hours before death, so that
the post mortem reveals a full stomach, or crop. In dogs and cats
the largest doses so far tried have not produced tachycardia.

5. Control experiments with feeding of desiccated testes, pancreas,
liver, and muscle seems to indicate that the above symptoms are
mainly due to the thyroid feeding.

6. It would seem, then, that in the case of normal healthy animals
there is a great variation in the resistance to thyroid substance in the
different genera, and that when the thyroid substance reaches the
toxic concentration the nervous,and cardiac phases of the symptom
complex in the birds and the lower mammals differ from those in man.

A SHADOW PUPILLOMETER FOR THE ACCURATE
STUDY OF PUPILLARY REACTIONS:

By G. W. Frez

THE principle underlying the apparatus is that an opaque object held
close to the eye and in its line of vision will not cast a shadow (wmbra)
upon the retina unless it is at least as large as the (apparent) pupil.

Twenty-third Annual Meeting. XV

It follows that an opaque object, as a slender cone or thin wedge
(pencil point), which can be made to vary in size by exactly determin-
able amounts, can be used to measure the diameter of the pupil. In
the shadow pupillometer this is accomplished by inserting a section
of thin steel ribbon 20 mm. by 6.5 mm., into the end of a rotatable
rod, which carries on its other end a protractor so adjusted that its
zero is at its index when the plane of the steel ribbon is in the line
of vision. For convenient application to the eye, the rod is suitably
supported (as by the head piece of a stereoscope).

The steel ribbon is rotated until a shadow line is projected across
the field of vision, and the angle of rotation is read. The diameter
of the pupil is then calculated by the formula, diameter = sine a X

d\2
6.5 mm., and the area of the pupil by the formula, area = .) Os

A graphic table which gives both of these values upon inspection, is
readily constructed. If preferred, a quadrant graduated to diameters
and areas may be substituted for the protractor. The apparatus is
adapted to the testing of either one or both eyes, alternately or simul-
taneously. It is easily accurate to 7>7 mm. in diameter of pupil, and
discloses quick variations of diameter of even less amount, such as
the arterial pulsations in the iris.

THE ALLANTOIN-PURINE EXCRETION OF THE MONKEY.
By ANDREW HUNTER AND MAvrRIcE H. GIVENS.

Our findings confirm the observation of Wiechowski,’ and support
the general conclusion of Wells.2, From 75 c.c. of the urine of one
monkey we have isolated by Wiechowski’s method 8.5 mgm. of allan-
toin in typical crystals. From 500 c.c. of the mixed urine of two
monkeys we obtained 172 mgm. of crystalline, though not entirely
pure, allantoin. In another 500 c.c. sample of mixed urine from the
same two animals we could detect no uric acid, and only 4.5 mgm.
of nitrogen in the form of purin bases. These samples cannot, un-

1 WIECHOWSKI: HoFMEISTER’s Beitrige, 1908, xi, p. Ior.
2 WELLs: Journal of biological chemistry, 1910, vii, p. 171.

xvi Proceedings of the American Physiological Society.

fortunately, be directly compared, but the analyses demonstrate at
least the great relative importance of allantoin as end-product of
purin metabolism in our animals.

THE PART PLAYED BY THE SPLEEN IN THE FORMATION
OF IMMUNE BODIES.

By Arno B. LucKHARDT AND FRANK C. BECHT.

As a working hypothesis we assumed that, other things being equal,
an animal possessed of a spleen would develop antibodies more rapidly
and in greater ultimate concentration than a splenectomized animal.
Splenectomy, therefore, was one of the procedures adopted to test
the validity of our assumption, active immunization of the animals
being effected by a single intravenous injection of antigen (goat’s or
rat’s corpuscles) administered at various intervals previous to or
following splenectomy. For each splenectomized animal we pro-
vided on the same day a control animal having same age, weight,
and size, and in order to make conditions as comparable as possible,
performed a laparotomy at which occasion the spleen was tempo-
rarily removed from the abdominal cavity and replaced (anesthesia,
mechanical irritation, shock). After immunization the animals were
bled daily for the first nine days and thereafter at regular intervals
for about three weeks. ‘The sera were kept surrounded by a freezing
mixture and were tested under the same conditions and on the same
suspension of corpuscles at the end of that period.

Our report on this phase of the work is based on the results obtained
from seven series of dogs comprising 21 animals which were immunized:
twenty-four hours previous to splenectomy and laparotomy (3 dogs);
immediately after splenectomy and laparotomy (4 dogs); three-
fourths of a year (2 dogs), twenty-one days (4 dogs), sixteen days
(4 dogs), eleven days (2 dogs), and five days (2 dogs) after splenec-
tomy and laparotomy. Briefly the results are as follows:

(1) The animals possessed of a spleen produced the specific anti-
bodies (hemolysins, hemagglutinins, and hemopsonins) more rapidly.

(2) The ultimate concentration of these antibodies in the serum
was usually much higher than in the splenectomized animal; never
was the concentration lower.

Twenty-third Annual Meeting. XVii

(3) In this relation of immunity there seems to be no compensa-
tion for the spleen, at least within a period of eight months.

2. Intraperitoneal introduction of spleen emulsion from dogs
immunized three to twenty-four hours previously by an intravenous
injection of antigen (goat or rat blood) resulted in the appearance
of the specific antibodies in the serum of the recipients. No increase
in antibodies was noted in the sera of those animals into whose peri-
toneal cavity normal spleen emulsion was introduced. The intro-
duction of “immune” heart muscle, liver, bone marrow, and lymph
glands did not give positive results.

3. The method of transplantation of the spleen zm toto has so far
not proved feasible in our hands.

THE OSMOTIC PROPERTIES OF SMOOTH MUSCLE.
By EE. 5B. Metes:

STRIATED muscle maintains its weight, as a general rule, only in salt
and sugar solutions isotonic with the blood of the animal from which
it was taken. Smooth muscle, when placed in isotonic or even hyper-
tonic sugar solution, gains in weight almost as fast as it does in dis-
tilled water. In Ringer’s solution having the formula NaCl, 0.65 gm.;
mCi c-02 gm:; CaCl, 0.025 gm.; NaHCO,, 0.02 gm.; H,O, 100 c.c.
the tissue usually gains slowly until at the end of twenty-four hours
it may be 20 per cent or 30 per cent heavier than originally, though
still quite irritable. These facts indicate that the fluid interchange
between smooth muscle and its surroundings is controlled by factors
widely different from those which operate in the case of striated
muscle and other cells.

THE EFFECTS OF EXTRACTS OF THE DIFFERENT PARTS
OF THE TAVPOPHYSIS.

By J. L. Miter, D. D. Lewis, anp S. A. MATTHEWS.

Extracts of the pars intermedia, freed of the depressor substance,
when injected intravenously gave a distinct pressor effect. Extracts

xviii Proceedings of the American Physiological Society.

of the pars nervosa, freed of the pars intermedia and its own
depressor substance, also give a pressor effect when injected in-
travenously. The pressor substance is apparently secreted by the
pars intermedia, but passes into the pars nervosa. It does not
need to be activated by the pars nervosa before it can exert the
pressor effect.

Extracts of the stalk of the ox hypophysis never gave any pressor
effect when injected intravenously. There is, therefore, a distinct
interruption in the path of secretion of the pressor substance from
the pars nervosa to the ventricle.

The portion of the anterior lobe immediately adjacent to the cleft
contains groups of cells belonging to the pars intermedia. Extracts
of this part of the gland freed of the depressor substance give the
same pressor effect as that obtained from extracts of the pars inter
media in the posterior lobe. This tissue probably accounts for the
secondary rise noted by Hamburger after injections of the anterior
lobe extracts and for the Ehrmann reaction occasionally obtained by
Franchini when using anterior lobe extracts. 3

We have obtained a decided pressor effect from the contents of a
cyst of the pars intermedia. The substance which slows the heart,
noted first by Howell, is confined mostly or at least in its most active
form to the pars nervosa. Extracts of the pars intermedia in their
purest form do not give this reaction.

THE HEART ACTION IN RELATION | @@s Toe
RESPIRATORY METABOLISM.

By J. R. Muriin anv J. R. GREER.

EXPERIMENTS on dogs were devised in which the absorption of oxy-
gen and the output of carbon dioxide were determined by means of a
small Benedict respiration apparatus attached directly to the dog’s
trachea. Simultaneously the blood pressure was recorded. The
effects of anesthesia were controlled.

Similar experiments on several different men in widely different
nutritive condition and in varying degrees of muscular activity (lying

Twenty-third Annual Meeting. KIX

on a bed, standing, standing and lifting weights, shivering, etc.) were
also done by means of the same respiration apparatus and the Er-
langer sphygmomanometer. The results show a fairly close corre-
lation in the same individual between the heart output expressed as
the product of the pulse pressure and the heart rate on the one hand,
and the absorption of oxygen and the elimination of carbon dioxide
on the other. The relationship between carbon dioxide elimination
and heart action is on the whole a little more constant than that be-
tween the oxygen absorption and heart action. Data will also be
given on the volume output of the heart determined both directly
and indirectly.

THE OLFACTORY SENSE OF FISHES:
By Go Hy Parxaer:

Tuar the olfactory surfaces of fishes are bathed with water has led
many physiologists to conclude that these surfaces are organs of
taste rather than organs of smell as they are in the air-inhabiting
vertebrates. The common catfish, Amiurus nebulosus, will attack
wads of cotton cloth containing hidden bits of earthworm, but is un-
stimulated by similar wads containing no such fragments. After
cutting the olfactory tracts in this fish, it treats both kinds of wads
with indifference. The killifish, Fundulus heteroclitus, will frequently
seize wads of cloth containing hidden dogfish meat, but it will seldom
touch similar wads containing no meat. When the anterior nasal
apertures of this fish are closed by being stitched up, the fish reacts to
wads containing meat as it does to those without meat. On reopening
the nasal apertures by loosening the stitches, the ability to distin-
guish between the two kinds of wads is again shown by the fish. As
the power to discover distant hidden food is lost by both kinds of
fishes when their olfactory organs are prevented from acting, even
though their organs of taste are in normal condition, it is concluded
that their olfactory organs are as truly organs of smell as are those
of the higher vertebrates, and that their olfactory organs are, there-
fore, properly classed as distance receptors.

xx Proceedings of the American Physiological Society.

MEASUREMENT OF THE BLOOD FLOW IN MAN.
(DEMONSTRATION.)

By G. N. STEWART.

Method. — The blood flow in the hand is calculated from the formula

Ts Wa
OS pai a
where O is the quantity of blood passing through the hand in the
time of observation, H the quantity of heat given off by the hand to
a calorimeter, S the specific heat of blood, T the temperature of the

Grams of blood per
100 c.c. of hand per Mean temperature of

minute. calorimeter.
Remarks.

Right hand. | Left hand. | Right hand. | Left hand.

10.8 : : M. C.,normal man, standing.

12.61 ; ; | : | M. C. sitting.

M. C. standing. Right hand

made to execute move-

/ ments in the calorimeter.

| N. M. standing. Boy with
left hand paralyzed (in-
fantile paralysis).

M. H. Man with left hand
affected by spastic paraly-
sis (birth palsy). Standing.

* Mean results of double observations with interchange of hands between the
calorimeters.

arterial blood entering the hand, and 7” the temperature of the venous
blood leaving the hand. The heat produced in the hand is negligible
in comparison with the heat exchange associated with the blood flow.
T is taken as approximately rectal temperature; TJ’ for a part like
the hand, where nearly the whole blood flow takes place through
quite superficial vessels, is, for a certain range of temperature, approxi-
mately the temperature of the calorimeter when the hand has been
immersed for a sufficient time in a large bath at the temperature of

Twenty-third Annual Meeting. bo.)

the water in the calorimeter. Outside of this range of temperature,
O, when calculated on the assumption that 7’ is the mean tempera-
ture of the calorimeter, gives us a minimum below which the blood
flow cannot lie.

FULSE-PRESSURE VARIATIONS IN THE PULMONARY
CHAC UIE:

By Cari J. WIGGERS.

Method. — A simple or T-cannula was inserted into the central
end of a pulmonary artery, the chest rendered air-tight, a negative
pressure equal to that previously existing in that animal was created,
and the animal allowed to breathe naturally. The cannulas communi-
cated by a tube passing through the chest wall either with a mem-
brane manometer or with two valved manometers recording maximal
and minimal pressures. ;

In normally breathing dogs with arterial pressures ranging from
roo to 112 mm., the maximal pressure in the pulmonary artery aver-
aged 36 mm., while the minimal averaged 3 mm. Both systolic and
diastolic pressures were lower in inspiration than in expiration. In-
creasing the negative pressure within the chest affected only the
systolic pressure during expiration, which was increased. Decreasing
the heart rate by weak vagus stimulation caused scarcely any change
of pressure during expiration, but acted to increase the systolic and
decrease the diastolic during inspiration.

While the mean pressure fell during inspiration, the pulmonary
pulse pressure increased, due to the fact that the diastolic pressure
was much lowered. The pulse-pressure quotient (ratio of pulse
pressure to systolic pressure) was increased, tending to show that
changes in the output of the right ventricle are not the only influences
causing a decrease in pressure.

The vagus nerve was stimulated and the heart stopped, causing
the pressure in the pulmonary arteries to fall to 2 or 4mm. With
each inspiration a decrease in pulmonary arterial pressure occurred,
and with each expiration an increase. Further eliminative experi-
ments showed that these results could be explained only by an effect
of intra-thoracic pressure on the lung vessels.

xxii Proceedings of the American Physiological Society.

THE FUNCTIONS OF THE CORPUS LUTEG)]
By Leo LoeEs.

THE corpus luteum has at least two functions:

1. To make possible the formation of the maternal placenta by
supplying a sensitizing substance to the uterine mucosa. .

2. To prolong the sexual cycle of the female organism in the preg-
nant as well as in the non-pregnant animal.

It can be shown experimentally that the former function is inde-
pendent of nervous connections between the uterus and the ovaries.

THE MECHANISM OF THE ASPHYXIAL RISE OF
BLOOD PRESSURE IN THE SPINAL ANIMAL.

By F; Hy Pixs.

EXPERIMENTS on animals (cats) after decerebration and transection
of the spinal cord or after paralysis of the brain, including the medulla
oblongata, by depriving it of its blood supply, show that the first
asphyxial use of pressure is due to the skeletal muscles.

A METHOD OF REMOVING GLYCOGEN FROM THE
HUMAN SUBJECT.

By GRAHAM LUSK.

In order to compare the results obtained upon phlorhizized dogs
through the influence of cold, experiments have been made upon
men with a view to causing the removal of glycogen from their bodies.
The procedure was as follows: The evening meal contained only
protein and fat; the breakfast was a cup of coffee. During the
morning the metabolism of a resting period was obtained by use
of a small Benedict respiratory apparatus. Then followed a cold
bath in a bathtub filled with ice blocks, the temperature of the bath

Twenty-third Annual Meeting. XXill

being 10°, and its duration from six to twelve minutes. The metab-
olism was then determined for a second time during the period of
shivering. ‘This procedure was repeated.

The results in one case showed a fall in the respiratory quotient
from 99 during the first period of rest to 75 during the period of
shivering, and quotient 75 was also found during the resting period
which followed. This See pond to the quotient found by Benedict
after prolonged fasting.

In a second case a thin man showed a respiratory quotient of 67
during a resting period following a period of shivering in which the
quotient was 85. This indicated the exhaustion of glycogen from the
body, and the low quotient is only to be interpreted by assuming
a production of glycogen from protein.

In a third case the individual experimented on was a muscular
athlete, and prolonged cold was not able to reduce his quotient at any
time below 80. The shivering was sufficient in these cases to increase
the metabolism from 100 to 200 per cent above the normal. |

EPPECT OF INTRAVENOUS INJECTION OF EXTRACTS
OF THE PINEAL BODY.

By J. A. E. Eyster (with H. E. Jorpan).

AN aqueous extract of the pineal body of the sheep causes, on in-
travenous injection, a fall of mean arterial blood pressure in the dog,
sheep, cat, and usually in the rabbit, greater than the fall obtained
from a similar extract made from other portions of the brain. The
fall of blood pressure is associated with a vascular dilatation in the
intestines, and since there is no important change in pulse rate and
no important effect on the excised mammalian heart, the dilatation
would seem to be the sole cause of the decrease in blood pressure.
Acidulated aqueous extracts of the pineal body cause on intravenous
injection in the rabbit a moderate and transitory diuresis. There is
no very definite effect on respiration in the anesthetized animal.
The results, as a whole, would seem to indicate a relatively low de-
_ gree of physiological activity of extracts of the pineal body so far as

xxiv Proceedings of the American Physiological Society,

the organs studied are concerned. One of us (Jordan) is making a
detailed study of the microscopic anatomy of the pineal body in
various animals. It is our intention to extend our physiological work
to a study of the effects of extirpation in the sheep and other animals.

ACUTE ANAPHYLACTIC DEATH IN RABBITS.
By J. Aum.

Ragesits highly sensitized by repeated injection of horse serum re-
spond in a characteristic fashion to the toxic injection when given
intravenously. The train of symptoms is ushered in by a marked
slowing of the respiration; without further premonitory symptoms
the animal suddenly falls over with clonic and tonic convulsions,
gives a few weak cries, and respiration ceases permanently. The
heart is usually no longer palpable a few minutes after cessation of
the convulsions. Autopsy shows the gut pale, without any hem-
orrhages; the splanchnic vessels are full. The lungs collapse well, but
show usually traces of pulmonary cedema. The heart is distended
with blood, especially the right ventricles. The ventricles usually do
not beat, nor do they as a rule respond with a contraction to mechan-
ical or electrical stimulation. The auricles usually beat regularly.

Blood pressure tracings show a fairly abrupt fall to the 10-20
millimetre level, the fall usually being preceded by a slight rise.

In order to rule out the central nervous system and the splanchnic
region from the production of this fall of blood pressure, a rabbit was
pithed including the medulla and the basal portions of the brain and
the aorta and vena cava inferior were clamped. Artificial respiration
was, of course, given. Injection of the toxic dose under these condi-
tions showed the same characteristic fall of blood pressure and the
same reduction or loss of irritability of the heart muscle.

It is therefore justifiable to conclude that the vital cause for acute
anaphylactic death in rabbits lies in the heart itself, the toxic injec-

tion producing a reduction or abolition of contractility of the
ventricles.

Twenty-third Annual Meeting. SEK

ON INTESTINAL PUTREFACTION DURING COPIOUS
AND MODERATE WATER DRINKING WITH MEALS.

By W. M. HATTREM AND P. B. HAWK.

1. THE drinking of copious (1000 c.c.) or moderate (500 c.c.) volumes
of water with meals decreased intestinal putrefaction as measured by
the urinary indican output.

2. Copious water drinking caused a more pronounced lessening of
the putrefactive processes than did the moderate water drinking.

3. In copious water drinking the total ethereal sulphate output
was increased coincidently with the decrease in the indican output.
This observation furnishes strong evidence in favor of the view that
indican has an origin different from that of the other ethereal sul-
phates, and that they cannot correctly be considered as indices of the
same metabolic process.

4. When Ellinger’s method is employed, the determination of
indican should be made on fresh urine before any preservative has
been introduced. Especially is this true when thymol is used as the
preservative.

5. The decreased intestinal putrefaction brought about through
the ingestion of moderate or copious quantities of water at mealtime
is probably due to an inhibition of the activity of indole-forming
bacteria following the accelerated absorption of the products of
protein digestion.

BIOLOGICAL ANALOGIES IN SOIL OXIDATION.
By OSWALD SCHREINER AND M. X. SULLIVAN.

THE soil is the seat of many biochemical activities which directly or
indirectly affect soil fertility. Many of the processes in the soil are
analogous to those occurring in plants and animals. Soils may show
fatigue under a one-crop system and likewise under unsanitary con-
ditions contain material which is retardative of plant growth. Many
other compounds, some of which are known to be products of proteo-

xxvi_ Proceedings of the American Physiological Society.

lytic digestion, occur in soils. The soil per se has oxidizing and catalyz-
ing powers which in cropped soils are due partly to activities of plant
roots, but in air-dried soils are due mainly to non-enzymotic soil
constituents, inorganic and organic, working separately, conjointly,
or in reinforcing and activating combinations. ‘The recently dis-
covered activating action of salts of organic hydroxyacids and the
discovery that alfalfa laccase is a mixture of salts of organic hydroxy-
acids have a close counterpart in soil oxidation studies.

THE ACTIVITY OF THE PANCREATIC FUNCTION
UNDER THE INFLUENCE OF COPIOUS WATER
DRINKING WITH MEALS.

By P. B. Hawk.

THE activity of the pancreatic function as measured by the fecal
amylase was found to be greatly facilitated when additional volumes
of water ranging from 1500 to 4000 c.c. were daily ingested at meal
by normal men maintained upon a uniform diet. Wohligemuth’s
method was employed in determining amylase. The reaction of the
feces may have been a disturbing factor in the determination. This
point will be further investigated.

FEEDING EXPERIMENTS WITH MIXTURES OF
ISOLATED FOOD SUBSTANCES.

By Tuomas B. OsBoRNE AND LAFAYETTE B. MENDEL.

THE experiments reported form part of an extensive study planned
primarily to throw light upon the significance of the individual pro-
teins in nutrition. The present series has been conducted with white
rats; the methods and devices in use were demonstrated. Observa-
tions on food intake, nitrogen balance, digestibility, and body weight
were made over long periods, and attention was devoted to such

Twenty-third Annual Meeting. XXVll

questions as the palatability of the ration, monotony of diet, and
inorganic salts, in addition to the nutrient rdle of the proteins used.
The investigation has taken account of the food requirement during
growth as well as the maintenance ration. The authors have already
succeeded in maintaining rats on diets containing a single, isolated
protein over a longer period than in any records heretofore published.
The experiments are being continued.

ARE THE PARATHYROIDS CAPABLE OF REPLACING
THE THYROIDS FUNCTIONALLY?

By SUTHERLAND SIMPSON.

In the course of some experimental work on the sheep the thyroids
were removed completely from nineteen lambs and twelve adults.
Most of them were allowed to live from five to six months after the
operation and were then slaughtered. Since during this period they
showed practically no symptoms of thyroid insufficiency, it was thought
that an examination of the parathyroids which had been left behind
might throw some light on the question of their compensatory func-
tion. Accordingly all the parathyroid tissue discoverable at the time
of death was removed, fixed in various fluids, sectioned in paraffin,
stained and examined microscopically. For comparison the para-
thyroids from a few normal sheep and lambs were fixed and stained
in the same way.

In the sheep the parathyroids are usually four in number — two
internal, included in the substance of the thyroid, and two external,
one on either side, imbedded in or closely related to the head of the
thymus. The latter are situated so far away from the thyroid that they
are not disturbed in any way by the operation of thyroidectomy.

The gland in the normal sheep is surrounded by a thin connective
tissue capsule from which delicate septa pass inwards subdividing the
organ into compartments that are occupied by the secreting cells.
These are usually arranged in solid columns, but in some parts of the
field small lumina can be made out in these columns and here the
gland is of the tubular or acinar type. In the human subject several

xxviii Proceedings of the American Physiological Society.

types of cells have been described, but in the sheep there appears to
be only a single type. The cytoplasm is granular, the nucleus rela-
tively large, rounded or oval and rich in chromatin. The glands are
extremely vascular

Microscopical examination of the two external parathyroids re-
moved from the thyroidectomized sheep and lambs when they were
slaughtered from five to six months after the operation, failed to
show that they differed in any respect from the normal gland. Ve-
sicular formation of the cells and the presence of colloid substance in
the acini were particularly looked for, but neither was found, and
the structure did not in any way appear to approximate to that of
the thyroid. In the sheep, therefore, there does not seem to be any
histological evidence in favor of the view that the parathyroid can
functionally replace the thyroid.

THE CONSTANTS OF PUPILLARY REACTION ae
PRELIMINARY REPORT OF EXPERIMENTATION
WITH THE SHADOW PUPILLOMETER.)

By (Ge W.. Freez:

Tue shadow pupillometer through its ability accurately to measure
instantaneous changes upon the experimenter’s own eyes, opens up
a new field in the study of pupillary reactions. Experiments with it
show that prevalent conceptions of these reactions require extensive
modification.

Pupillary movements. — The iris shows marked pulsations of some-
what irregular character, which are apparently associated with the
slight twitchings of the lids, although the latter are not always
perceptible. The resulting changes in the pupil may be as great as
25 per cent of the total area, and occupy a quarter of a second. They
are superposed upon much more regular and frequent variations of
a fairly uniform character, which are too irregular for pulse beats,
although they blend with them and assume the character of exag-
gerated beats. The pulse.beats themselves can be distinguished best
when the shadow of the pupillometer is reduced to its permanent

ie ee ——<

Twenty-third Annual Meeting. KIX

minimum for the illumination used. They amount to about 0.05 mm.
in diameter. The changes in the pupil due to ordinary respiration
are negligible. Enforced respiration, however, produces some change
in size.

Pupillary adjustments. — In studying pupillary reaction, the area
of the apparent pupil (7. e. of the pupil as seen from without, magni-
fied by the lens effect of the cornea) was adopted as the best basis of
comparison, since it gives more nearly the light-gathering power of
the eye, and therefore the intensity of the illumination upon the
retina, than does the area of the actual pupil. Much of the adjust-
ment of the eye to strong light (above ro metre-candle power) appears
to be due to a diminishing sensitiveness of the retina rather than to
an extreme narrowing of the pupil, which remains of nearly fixed
size (slightly under 1 mm.). The pupil dilates markedly in lights
below 4 metre-candle power. The adjustments of the pupil to sudden
changes of light, as, for example, when the eye is opened after a closure
of several seconds, are completed in approximately one half second.
Light which does not strike the fovea affects the size of the pupil in-
versely as its distance from the visual pole. Accommodation makes
a difference in the pupil’s area of from 10 to 30 per cent, depending
upon the strength of the illumination; hence all tests of pupillary
reactions must have as a condition an equal convergence of the eyes.

THE MIGRATION OF SOLUTIONS IN ANIMAL BODIES
DEPRIVED OF THEIR CARDIAC CIRCULATION.

By S. J. MELTZER.

SIx years ago the author showed that an injection of adrenalin into
one of the lymph sacs of a normal frog causes a maximal dilatation of
the pupil. In this case the adrenalin is carried to the iris through the
circulation and the dilatation appears a few minutes after the injec-
tion. Since the pupil of the frog reacts to the adrenalin for many
hours after the death of the animal, I used it as a definite reaction for
the study of the question whether there can be a migration of fluid
~ in an animal body deprived of its cardiac circulation. In one method

xxx Proceedings of the American Physiological Society.

the heart of frogs was exposed, firmly ligated and removed. In an-
other method a curved needle was passed through the intact wall
around the large vessels which were firmly ligated around the sternum,
and then a second ligature was applied in the same manner around the
middle of the heart. The results were positive. One cubic centimetre
of adrenalin injected into the dorsal lymph sac invariably brings on a
maximal dilatation of the pupils which sometimes appears as early
as half an hour after the injection. When injected into the lateral
lymph sac&, it may take two hours before the dilatation appears.
Positive results can be obtained also from an injection into the thighs
or legs. Positive results were also obtained in many instances, even
when the animal was suspended by the head. The adrenalin migrated
even against gravity.

The experiments with strychnine gave still more striking results.
One cubic centimetre of a 1 per cent solution of strychnine brings
out a definite tetanus when injected into the lymph sacs of the trunk,
or of the extremities, or into the abdominal cavity. A strong dose
of strychnine causes paralysis.

Since the firm ligation and removal of the heart excludes the par-
ticipation of the heart, blood vessels, heart lymphs and lymph vessels,
the migration can take place only through the lymph spaces. ‘The ex-
periments demonstrated that there is in the body a mechanism which
is capable of carrying on in a slow but fairly efficient manner the
migration of solutions in the body without the aid of the cardiac cir-
culation. It is not improbable that this mechanism is active also in
the presence of the cardiac circulation.

AN AUTOMATIC SHELLACKING DEVICE.
(DEMONSTRATION.)

By D. E. Jackson.

A TRANSVERSE supporting axle D permits the combination to revolve
back and forth through an angle of forty-five degrees. The can A is
six inches long by five inches in diameter. C is half-inch tubing. At
H a shelf one and one half inches wide placed across the back end of

Twenty-third Annual Meeting. XXxi

B forms a sort of reservoir in which the varnish can accumulate when
B is suddenly elevated. This prevents the varnish from spilling out
on the floor before it has had time to run back into A. B is eight and
one half inches long, three inches deep, and five inches wide. At I a
hole is made in A for equalization of the air pressure. G is a small
lead weight used to pull A down when the treadle is raised. (A simple
spring may bé used for this if desired.) The uprights M, M, are at-
tached to a wooden base (JV) twelve inches long by ten inches wide.
The ends of the axle D turn in holes in the upper ends of MW, M. N
is fastened to a table or shelf. At O a transverse bar of wood is nailed
across between the uprights, and serves as a stop for the pan B when
it has been pulled down to a horizontal position. The treadle E may
be hinged at an angle to B when B is lowered to prevent dropping of
varnish on the operator.

Diagram showing a lateral view of the apparatus.

INHIBITION OF THE DUODENUM COINCIDENT WITH
THE MOVEMENTS OF- THE PYLORIC PART OF THE
STOMACH.

By D. R. JosEPH AND S. J. MELTZER.

IN our present investigation we studied the normal contractions of
the pyloric part of the stomach and of the duodenum simultaneously,
by methods which we shall not describe here. We wish only to re-
port here briefly the well-established fact that during each contraction
of the pyloric part of the stomach the duodenum stops its rhythmic

xxxii Proceedings of the American Physiological Society.

activity and loses its tone, only to resume both again as soon as the
contraction of the stomach passes off. We wish to add that we look
upon this fact as a manifestation of the Law of Contrary Innervation.
The contraction of the duodenum is antagonistic to that of the pyloric
part of the stomach, and it is therefore a part of the normal function
that the movements of the duodenum should be inhibited while the
pyloric part of the stomach contracts. The phenomenon is identical
in design with the relaxation of the cesophagus and cardia during
deglutition as observed by Kronecker and Meltzer and with the ob-
servation by Cannon, reported at this meeting, of the relaxation of
the stomach during the contraction of the cesophagus.

THE STIMULATION OF THE GASTRIC SECRET
UNDER THE INFLUENCE OF WATER DRINKING
WITH MEALS.

By F. WILts AND P. B. HAwkE.

UNDER the influence of water drinking at mealtime by normal men
the ammonia content of the urine is increased. This increase is con-
sidered an indirect index of the activity of the gastric function. This
increased output of ammonia is directly proportional to the increase
in the water ingestion. Within limits the quantity of ammonia ex-
creted per 10oc.c. of ingested water is also practically the same whether
the water ingestion is large or small. The above facts indicate that the
animal organism attempts to maintain a constant acid concentration
in the stomach contents under the influence of water drinking with
meals.

FURTHER EXPERIMENTS ON THE ANTAGONISTIC
ACTION OF SALTS.

By Jacques Loers.

THE author showed, first, that a pure solution of KCl of the same
concentration as that in which this salt is contained in the sea water

Twenty-third Annual Meeting. XXXill

kills half-grown fundulus in two days or less, while in a sodium
chloride solution of the same concentration the fish live indefi-
nitely; second, that potassium chloride solutions which are toxic can
be rendered harmless through the addition of a definite quantity
of sodium chloride; third, that the ratio between the toxic con-
centration of potassium chloride and the concentration of sodium
chloride required to annihilate the toxic effect of the potassium
chloride is a constant one; and fourth, that the antagonism exists in
this case between the two kations, Na and K and not between K
and Cl. From these facts the author draws the conclusion that the
antagonistic action is ultimately due to a partition of the same pre-
sumably colloidal anion contained at the surface of the cells, espe
cially of the gills of the fish, between the two metals, K and Na.

The author points out that as long as the ratio ra remains below
the critical value found in these experiments, namely, x7 to a5, not
enough K ions get into the blood of the animal to produce a toxic
action.

It was further found that there exists an upper limit for the con-
centration of KCl beyond which its toxic action can no longer be
inhibited by NaCl.

=)
=

Ay Ss Ee

a i ns
7%

THE

American Journal of Physiology.

VOL. XXVII. NOVEMBER 1, Ig1o. NO. I.

THE INFLUENCE OF ALCOHOL UPON NITROGENOUS
METABOLISM IN MEN AND ANIMALS.

By LAFAYETTE B. MENDEL anp WARREN W. HILDITCH.

[From the Sheffield Laboratory of Physiological Chemistry, Yale University,
New Haven, Connecticut.]

HE action of alcohol in the body may be twofold, pharmaco-
dynamic and nutritive. Respecting certain features of its
behavior in metabolism there is considerable unanimity of opinion.1
When alcohol is ingested in moderate amounts, its influence on the
utilization of the ordinary foodstuffs in the alimentary tract is mini-
mal. Moderate doses are oxidized in the body and may thus exer-
cise a sparing action on the metabolism of other foodstuffs. There
is, further, an influence admittedly exerted by alcohol and alcoholic
fluids on the elimination of uric acid; but the significance of this
phenomenon is by no means clearly understood. Other incidental
features of the influence of alcohol on metabolic processes, such as,
for example, the oxidative functions, have been recorded from time
to time. ,
The appearance of intermediary metabolic products, such as gly-
curonates, in the urine in cases of acute alcoholic intoxication sug-
gested the possibility of discovering some details of the specific ac-
tion of alcohol by making a more refined analysis of the urine than is
commonly done in such instances. Just as peculiarities of purine

1 The literature of the subject has been compiled to 1904 by ABDERHALDEN:
Bibliographie der gesammten wissenschaftlichen Literatur iiber den Alkohol und
den Alkoholismus. The most recent general discussion of alcohol in relation to
metabolism is by RosEMANN: OPPENHEIMER’S Handbuch der Biochemie, 1909,

im, Part 1, p. 413.
I

2 Lafayette B. Mendel and Warren W. Hilditch.

metabolism have been revealed in this connection, it.was believed
that other aberrant features might come to light.? Attention has
been directed towards obtaining evidence of changes in the relative
participation of various processes in the nutritive functions, particu-
larly the partition of nitrogen in the urine. Heretofore the total
nitrogenous metabolism and the total energy transformations have
for the most part been the subject of study. |

We have made comparisons, in men and dogs, of the output of
urea-N, ammonia-N, creatine-N, creatinine-N, purine-N, and other
constituents of the urine under fixed dietary conditions with and
without intake of alcohol. The optical rotation of the urine was
also observed in order to detect perversions of metabolism. The
data for dogs supplement those recently presented by Salant and
Hinkel.? Earlier observations on man, by Jackson and Blackfan,*
are likewise of interest in this connection.

EXPERIMENTS ON MAN.

The human experiments were conducted on two young men not
accustomed to the use of alcohol: W. W. H., weighing 54 kilos; and
J. F. L., weighing 67 kilos. The daily diet consisted of

Breakfast, 8.30 A. M. Dinner, 1 P.M. Supper, 6 P. M.
Grape fruit, about 175 gm. Potato..... 100 gm. ‘“‘Uneeda” biscuit 25 gm.
Cereal (‘Force’) . 20 “ Hamburg steak . 100 “ Creamcheese .. 20 “
Banana, about .. 80 “ Pickles,about. . 30 “: Pickles;/about ] 9

Orange, about _. 125.
One egg.

Milk, 900 c.c., and sugar, 100 gm., divided between three meals.
Bread, 125 gm., and butter, 60 gm., divided between two meals.
‘This was calculated to contain 13.4 gm. N. and yield about 2600 calories.

* A brief report of these studies was presented to the American Physiological
Society, December, 1909. Cf. Proceedings, This journal, 1910, xxv, p. xi. The
experimental data are taken from the thesis presented by W. W. Hixprrcu for
the degree of Doctor of Philosophy at Yale University, June, 1909.

* SaLant and HInxkEL: Journal of pharmacology, ror10, i, p. 493.

* Jackson and BrackFan: Albany medical annals, January, 1907.

* These subjects also served in the experiments on the metabolism of purines
by Menver and Lyman: Journal of biological chemistry, 1910, viii, p. 115, which
may be consulted for comparisons.

Influence of Alcohol upon Nitrogenous Metabolism. 3

During the alcohol periods it was decided to make the intake of
alcohol equal to that selected by Atwater and Benedict ® in their
well-known experiments on the energy exchange. They adminis-
tered 72 gm. of absolute alcohol in divided doses, without noticeable
psychic effects. Our two subjects received 96 c.c. of 95 per cent
alcohol, furnishing about 500 calories, daily in six doses of 16 c.c.
each. Three were taken with meals, the others at 10.30 A. M., 3.30
P. M., and zo P. M., in milk or milk and water. The daily intake of
fluid, including the milk, was 1300 c.c. After each dose a feeling of
warmth was experienced, even at mealtimes. On several days when
the mid-morning or mid-afternoon dose was taken somewhat later
than usual, a slight and transitory dizziness was experienced, possi-
bly owing to the fact that the alcohol reached an almost empty
stomach. No other untoward features were noted. The diet was
introduced three days before any of the analyses were begun.

The results of the analyses are summarized in Tables I and II.’

DISCUSSION OF THE EXPERIMENTS ON MAN.

In considering the tabulated results it is important to bear in
mind that, aside from differences in body weight and individuality,
the conditions of experiment were duplicated in the two subjects.
This fact lends unusual significance to the deductions which are
permissible.

Food utilization. — The diet itself afforded little more than a main-
tenance ration, and it was well utilized. The evidence for this is
found in the study of the composition of the feces.

The very slight increases in air-dry solids, nitrogen, and purine
content during the alcohol period scarcely exceed the limits of ana-

6 ATWATER and BENEDICT: National Academy of Sciences, 1902, viii, sixth
memoir.

7 The analytical methods employed in all the work reported in this paper
were those of Forrn, for urea-, ammonia-, creatine-, and creatinine-N; of
KRUGER and Scum, for urinary purine-N and uric acid; of Kriicrer and Scutr-
TENHELM for purine-N in the feces; titration with uranium nitrate solution for
phosphorus. Nitrogen was estimated by the Kjeldahl-Gunning process. Tests
for sugar were made with the delicate reagent of BENEDICT: Journal of biological

chemistry, 1909, v. p. 485. Optical rotation is expressed in degrees on the Ventzke
scale (V°).

4 Lafayette B. Mendel and Warren W. [Hilditch.

EXPERIMENT I.

Man. — Composition

ne ae Sire ee sh a4
® |38| 5 s oS 2 Zz sg Zz
Sneek & =e: bo 3 5
Ow oa "a : 8 g 3
> | 22 ° = PS 3 ° : =
Bee ole ae eae p
= 3 Bo) n <x
id Vs ca (= 4
K. fm love; c.c; c.c, 1.0- gm. gm gm.
12 67.4 1 105 —30 11.45 55 9.55
13 2 1150 —30 12.96 55 10.98
14 67.5 3 1150 —21 12.91 57 10.77
15 67.4 96 1 1300 —28 10.83 58 8.87
16 aoe 96 2, 1300 —18 12.53 56 10.24
17 von 96 3 1300 —21 11.29 58 9.05
18 67.3 96 4 1350 - —22 11.29 53 9.19
19 aoe 96 5 1300 —24 11.96 22 9.82
20 Pow! 96 6 1300 —23 11.88 48 9.97
21 67.5 | 96 7 1300 —21 11.99 54 9.81
22 ae 96 8 1300 —29 11.61 54 9.48
23 aw 96 9 1300 —27 72 54 9.74
10 9.42
8.47
10.08
9.04
10.00
10.10
11.04

lytic error. If one were inclined, in view of their uniform occur-
rence, to assign a cause to them, it might be sought in the secretory
effect of alcohol along the alimentary tract.’ Such an application
scarcely seems justified on the basis of the data at hand. Our utili-

* Cf. CutrTENDEN, MENDEL, and Jackson: This journal, 1898, i, p. 164.

Influence of Alcohol upon Nitrogenous Metabolism. 5

EXPERIMENT I.
or Excreta. (J. F. L.)

si f < | % $2) 2¢] 23] & z gE
mye thie tee leel ee le) a |e
gm. gm. gm. gm. gm. gm. em. | gm. gm. ve
ene |) 52 A 6k fs er | 1216. | 4-108. osds | —2
156 | .014 | .59 67 1 | 1367 | -023| 0972 | —a
156 014 | 50) 81 71 | 13.62 | —0.18 | 10% | —2
210 020 | .59 56 79 |-Stlo2 |----1182-} oss | —2
230 027 | .59 88 79 | 13.32 | +012] 1.051 | —.1
205 025 | .59 84 7) | 12.08 | 41.36 | 0.988 0
ey OF |, 5a.) 020.) 75}. 90.-} Azos [4136-4 0972.) —2
5 | 01 | 59 | .013:| 81 | .79 | 12.75 | +.009 | 0.950 | —4
182 020 | .58 65 79) } 4967 |'-+0.77 | 0086 || 24
186 021 | .61 82 79 | 12.78 | +0.66 | 0.904 | —.2
198 018 | .58 80 79 | 1240 | +1.04 | 0.994 | —2.
208 022 | .59 62 79 | 12.51 | +093 | 0.994 | —1
193 020 | .61 76 7) \ (wat | 11g e osteo) 3
191 021 | .62 68 79: hd. 27 4 A An ose. |) 202
190 019 | .60 67 79 | 12.83 | +061 | 0.965 | —.2
162 015 | .62 63 69 | 11.65 | +0.79 | 0.961 | —.2
176 015 | .59 96 4.69. | 9909 | 40571 9647: ') 2
169 015 | .58 82 69 | 1296 | +0.48 | 1.004 | —.3
174 015 | .58 Wh SF 69 | 13.89 | —0.45| 1.015 | —.3

zation figures for nitrogen compare closely with those of Atwater
and Benedict.

Protein-sparing action.— The quantities of alcohol consumed by
our subjects could furnish about 500 calories per day. This is an
equivalent of non-nitrogenous food sufficient to exert a distinct
- protein-sparing effect. As a rule, this favorable nutritive action

6 Lafayette B. Mendel and Warren W. Hilditch.

EXPERIMENT II.
Man. — CoMPOSITION OF

Body weight.
95 per cent.

Day of period
Liquid intake.
Spec. grav

e}| Alcohol taken

°
ic}
°

i — —
ee ee air ge SF Date, 1909.
on
A
se
= GW bw
&2 8} Ammonia N.

bo
=

2
3
4
5
20 6
*f
8
9

a
i=)

has been noted by previous investigators who have administered
similar doses in a comparable way. Rosemann points out that al-
though alcohol may replace fat or carbohydrate or produce an equiv-
alent sparing effect in the later periods of its use, at first this may
not happen. Some observers have recorded higher outputs of uri-
nary nitrogen in the early period of alcohol administration. This has

Influence of Alcohol upon Nitrogenous Metabolism. 7

EXPERIMENT II.
Excreta. (W. W. H.)

es) 2 |2}4|s8|28|/ 38) 42 | 2 | 3
Pe ers) 2 Poe wes | tee | > 8 g
=) a 3) i av

gm. gm gm gm. gm. gm gm. gm. | gm. Vo
.150 015 51 ere 0.88 0.90 12.83 +0.61 1.039 —.2
.150 .014 .50 0.36 0.90 12.67 +0.77 | 1.018 —,.2
.159 .015 1 0.83 0.90 13.05 +0.39 1.029 —.2
73 .013 51 0.68 1.06 12.02 +1.42 | 1.026 —.3
187 .023 48 0.77 1.06 12.04 +1.40 | 1.029

.200 .020 .50 0.17 1.06 11.91 +1.53 | 1.004

.189 .015 52 .009 0.75 1.06 11.86 +1.58 0.950 —.2
184 .021 52 .009 | 0.68 1.06 11.99 +1.45 | 1.004 —.2
181 .019 49 0.60 1.06 12.02 +1.42 | 0.986 —.2
191 .024 51 0.77 1.06 11237, +2.07 0.983 —.3
.204 021 49 0.58 1.06 11272 +1.72 0.976 —.2
.192 .020 5A 0.82 1.06 11.70 +1.74 0.979 —2
.206 .024 252" 0.69 1.06 11.81 +1.63 0.965 —.3
.194 .019 53 0.69 1.06 | sr a7 +1.69 0.993 —.2
.195 .021 oat 1.08 1.06 | 12.35 +1.09 0.983 —.2
.160 .013 50 0.60 0.97 12.20 +1.24 0.986 —.2
170 .014 50 0.78 0.97 12.66 +0.78 1.012 —.1
AW .015 157, 0.75 0.97 13.07 +0.37 1.029 —.3
.156 .012 5 0.86 0.97 12.63 +0.81 1.012 —.3

been ascribed to an increased elimination of nitrogenous katabolites
as the result of an induced diuresis. Rosemann rather inclines to the
view that the toxic (pharmacodynamic) action of alcohol over-
balances its possible nutritive value in the early stages of its use,
leading to cell katabolism. Tolerance may thus noticeably alter
the influence of alcohol on metabolism.

8 Lafayette B. Mendel and Warren W. Hilditch.

CoMPOSITION OF THE Faces — MAN — (AVERAGES, PER Day).

Total Purine Ether
air-dry.} nitrogen. nitrogen. extract.

i gm. gm. p.c. gm. p. c. gm. p.c.
Fore period : 16.8 | 0.71 | 4.25.) 035 | 20 4 2saneeee

Alcohol period . .. . } 17.8 | 0.79 | 4.44 | 041 | .23 | 241 | 13.5
After period 15.6 | 0.68 | 4.36 | 037 | .24 | 240") ise

W. W. H.

|
Fore period 17.2 | 0.90
Alcohol period . . . 19.1 | 1.06

After period 18.9 -| 0.97

In our experiments there was no significant diuretic effect; and
the protein-sparing influence of the alcohol made itself apparent
from the first day. It may be emphasized that the intake of alcohol
was favorably distributed and never induced obvious toxic or un-
toward results. The effect is expressed in the more favorable nitro-
gen balance during the alcohol period in both subjects.

NiItTRoGEN BALANCE: Man — Summary oF Datty AVERAGES.
(THE Datty INTAKE IS ESTIMATED AS 13.4 GM.)

Fore period. Alcohol period. After period.

gm. gm gm.

. in urine 12.4 11.6 122

eer. Daily output a guan 0.7 0.8 07
aoa MOotala oes bee Me a eee oe Sil 12.4 12.9
Nitrogen balance . . .: .. 0.3 1.0 0.5

: in urine 11.9 10.8 a bey

Pays Daily output eye 0.9 rel 1.0
ne ary ‘Totaly en: oh ceo ee ee 12.8 11.9 12a.
Nitrogen balance... . 2. . . 0.6 i ES) 0.7

Partition of nitrogen in the urine: urea, ammonia, creatinine. —No
alteration in the elimination of these constituents was noted be-
yond a reduction in the output of urea associated with the smaller

Influence of Alcohol upon Nitrogenous Metabolism. 9

nitrogen elimination during the protein-sparing alcohol period.
Jackson and Blackfan likewise found the creatinine output in man
to remain unchanged in experiments with alcohol and a nitrogen-
free diet of starch and cream. Since, as will be shown, the purine
metabolism is altered, we may further conclude with Jackson that
the creatinine and purines probably do not have a common origin;
although such evidence as is presented does not eliminate the possi-
bility that the influence of alcohol is exerted on the elimination or
destructive factors rather than the origin of these compounds.
Creatine was not a significant constituent of the urines examined
by us.

Purine metabolism. — Our experiments leave no doubt that
even in the ‘‘moderate’’ doses used alcohol increases the output of
uric acid in man. This is in direct confirmation of the earlier experi-
ments in this laboratory by Beebe.? The minimal changes in the
output of purine bases under the influence of alcohol scarcely de-
serve note further than that they too were in the direction of an in-
crease. Fig. 1 presents the results in graphic form.

Obviously the alcohol may affect either the endogenous or exo-
genous factor in purine metabolism, or both. The previously quoted
studies of Jackson on nitrogen-free diet and newer experiments by
Landau '° on patients with purine-free diets indicate that alcohol
may increase the endogenous output of uric acid in man. We have
obtained direct positive evidence in the case of W. W. H. in two
separate experiments.

In these experiments the diet, etc., were the same as in our previ-
ous trials, with the substitution of two eggs for 100 gm. Hamburg
steak (at noon) and the omission of one egg at supper. An intro-
ductory period of three days on this purine-free dietary preceded
the actual analytical period. The analyses were made in duplicate
on one third of the daily urine.

The average daily augmentation of uric acid-N elimination in
these trials on a purine-free diet was 13.6 mgm. in Experiment III
and 15 mgm. in Experiment IV. The question arises: Is the increase
in uric acid elimination under the influence of alcohol and with an

® BEEBE: This journal, 1904, xii, p. 13.
10 LanpAu: Deutsches Archiv fiir klinische Medizin, 1909, xcv, p. 280.

10 Lafayette B. Mendel and Warren W. Hilditch.

INFLUENCE OF ALCOHOL ON ENDOGENOUS PURINE METABOLISM — Man —
COMPOSITION OF THE URINE.

EXPERIMENT III.

Alcohol, . Specific Uric acid Purine base
95 per cent. Volume. gravity. nitrogen. nitrogen.

c.c. c.c. Pranic. gm. gm.

1300 1.022 112 .012
960 1.026 118 014
920 1.027 135 016
970 1.027 119 017

860 1.027 132 015

EXPERIMENT IV.

1065 1.019
1165 1.020
1085 1.022
1530 1.018
UZhS 1.020
1170 1.020
1500 1.019
1260 1.020

930 1.022
1050 1.021

960 1.022

960 1.023
1235 1.020

1 The urine throughout these experiments was acid to litmus.

ordinary diet entirely accounted for by this increase in the endog-
enous component? Probably not, as a simple calculation shows:

Influence of Alcohol upon Nitrogenous Metabolism. 11

AVERAGE Daity Output oF Uric Acip-N. (W. W. H.)

Endogenous Exogenous-+ Endogenous Exogenous

output. output. output (cal. by
Experiment IV. Experiment II. difference).
mgm. mgm. mgm.
Mumatcohol 6 wl kt 142 191 49
Without alcohol ...... 126 153 27
Increased by alcohol .... 16 38 22
itp 8s W. W. H

210

150 7

Y Uy Y
Vl We Uta V Lo

120

\\\ Ee
\\ Ee oo
WV
WW
\\ ee

WW
WV
WWW
WV EE

a Alcohol i

LALLA

Alcohol

Ficure 1.— Each vertical (ordinate) unit represents 30 mgm. of nitrogen. The hori-
zontal (abscissa) units mark periods of twenty-four hours each. The boxed spaces
represent the daily outputs of uric acid; and the shaded portions represent purine bases.

Landau’s figures for the increase in endogenous uric acid output
in five out of seven cases are of the same order, though somewhat
larger than ours. His normal figures are, however, so widely vari-
-able from day to day that the exact value of his averages must be
questioned. In regard to the excretion of exogenous purines he has

12 Lafayette B. Mendel and Warren W. Hilditch.

concluded that alcohol decreases the output as the result of dimin-
ished permeability of the kidneys towards uric acid. The consider-
ations presented by him are by no means convincing and need not
be reviewed here. It is difficult to understand why the kidney
factors should not be similar, whatever the origin of the uric acid
presented for elimination. ;

At the present stage of investigation a satisfactory hypothesis for
the specific action of alcohol on purine metabolism is not forth-
coming. Diuresis will not account for the augmented elimination
in our experiments. There is no valid reason for assuming any unique
synthesis of uric acid other than through the well-known sequence
of enzymatic processes from purine precursors. If the increased
output were confined to endogenous uric acid alone, one might be
ready to accept Landau’s explanation of a heightened (toxic) disin-
tegration of cell nucleoproteins resulting in a hyperproduction of
purines with consequent increase in uric acid formation and excre-
tion. ‘‘One must conjecture,’ Landau writes, ‘that nucleoproteins
are far more sensitive than the other proteins towards the adminis-
tration of alcohol, and that the increased nuclear disintegration
does not cease when the protoplasm has already become adapted to
alcohol”’ (p. 308). On this view alcohol acts specifically to injure
cells and damage the nucleus. Schittenhelm’s"™ observations on
dogs have also shown an increased endogenous output of purines
during periods of alcohol administration.

The hyperproduction theory here reviewed is interesting in its
relation to possible structural alterations in cells under the influence
of alcohol. Histologists are in no way agreed as to what takes place.
But if the suggestion presented by our calculations and indicating
that exdgenous uric acid or purine output is likewise increased by al-
cohol is substantiated by further evidence, the ‘‘cell destruction”
theory will be inadequate to explain the facts. Some further indi-
cations of the disintegration of cellular constituents might be ex-
pected; for example, an increased output of phosphorus. On this
point the experimental evidence is at present distinctly conflicting.

Another suggestion — that of increased hypoxanthine liberation
in muscular tissues under the influence of alcohol — we are inclined
to dismiss because of the constancy of our findings in respect to the

‘' SCHITTENHELM: Zeitschrift fiir physiologische Chemie, 1909, Ixii, p. 93.

Influence of Alcohol upon Nitrogenous Metabolism. 13

elimination of creatinine. Here again it must be admitted that this
may not be an index of the chemical conditions in the muscular tis-
sues. Plimmer, Dick, and Leib” have lately associated instances
of heightened purine output with the special metabolism of the leu-
cocytes. We have excluded this explanation for reasons presented
elsewhere.”

There remains the more vague hypothesis of some disturbance in
the enzymatic transformations of the purines under the toxic in-
fluence of alcohol — an explanation applicable to both the endog-
enous and exogenous components. Further discussion at the present
moment appears unprofitable.

Other urinary constituents. — The remaining data presented in the
tables call for little comment, since the conclusions to be drawn
from them are obvious on inspection. The constancy of the so-
called undetermined nitrogen derived by calculation is merely another
expression here for the absence of marked alterations in the partition
of the nitrogen. There is a very slight, perhaps insignificant, ten-
dency towards a diminished urinary output of phosphorus (as esti-
mated by uranium solution) during the alcohol period.“ In view of
the marked levorotation which we have observed in the urine of
chronic alcoholism this factor was considered in our subjects. No
abnormal features were detected.

EXPERIMENTS ON Docs.

In the preceding experiments on man the studies of the influence of
alcohol on metabolism were confined to doses which produced no
apparent untoward effects. For the obviously toxic conditions
recourse was had to trials with dogs.

The animals, full-grown bitches, were catherized daily. Their
daily diet consisted of 200 gm. hashed meat, 30 gm. lard, 30 gm.
cracker meal, 200 gm. water; 20 gm. bone ash were added to improve
the consistency of the feces (Gies’s method). The nitrogen intake
was 7.3 gm., in about 600 calories (estimated). The alcohol ad-

2 PrimMeER, Dick, and Leis: Journal of physiology, 1909, xxxix, p. 98.

18 MENDEL and LyMANn: Journal of biological chémistry, 1910, vili, p. 115.

4 Cf. SALANT and HINKEL: Loc. cit., who found a diminished output in dogs
‘and discuss the questions involved.

14 Lafayette B. Mendel and Warren W. Hailditch.

ministered was of 95 per cent volume strength and was added to the
food in doses ranging from 1 to 7 c.c. per kilo body weight, as
indicated. |

EXPERIMENT V.!

Doc —ComposiITION oF URINE. ALCOHOL GIVEN = 1-2 c.c.
PER KGM. Bopy WEIGHT; ONE MEAL PER DAY.

Aha | ae | ee ee § | 34
Dec kgm c.c c.c. 1:0'— gm. gm gm. gm. gm WEB
7 9.6 275 —20 | 6.21 18 5.28 | .09 14) 5
8 9.6 260 =e nae Oe 20 3.195% |" 209 i a
9 9.5 si 275 =), 16.25 22 5.40. | 09°) STS aiees
10 9.4 ee 289. | 18 | 6.75 26 3.16 |, 902 20: sl pee
11 9.4 ae 270 ==18; "|, 6.20 24 5.31 .09 20-1
122 5) 953 18.57} 110 —34 | 4.20 13 3.60 | .09 ea)
13 9.0 see 210 | —26 | 6.99 26 6.00 | .09 28 Ar ine
14 oF sais 270 18 =|, "G08 22 5.63 | .09 19. =e
15 9.0 9.0 260 || 1108 26 4.74 | .09 18) 4)=ae
16 9.1 9.1 280 Sai ass 21 .09 16.) 36
17 9.2 9.2 280 0 gL o.00 22 4.75: .| 09°") ie
18 9.2 13.8 300 iene. 70 23 .08 16 joa
19 9.2 18.4 305 Set 24. 4.71 08 | 16 ia
20 9.2 18.4 315 1D") S79 24 4.745.) 08° |. Gea
21 9.2 18.4 280 —19 | 5.58 224. 4.71 | 08 | s1Gpse
22 9.2 18.4 290 =20 | 5.88 29 5.06 | .08. | -1GD\ee

* Daily N intake = 7.3 gm.; 600 calories. The urine throughout the experiment
was acid to litmus.

* On this day 18.5 c.c. alcohol = 2 c.c. per kgm. (diluted three volumes with
water) were administered with.a stomach sound just before the meal. Within an
hour the dog vomited and refused to eat the vomitus again. To restore equilibrium
two normal days were allowed to elapse before the alcohol feeding (with the meals)
was resumed.

Influence of Alcohol upon Nitrogenous Metabolism. 15

EXPERIMENT VI. — Doc — COMPOSITION OF EXCRETA.

The same animal as in Experiment V was used. The dose of alco-
hol was 4 c.c. per kgm. per day. During the fore period and the
first eight days of the alcohol period, the food and alcohol were
divided into two meals given at 10 A.M. and 4P.M. During the last
eight days of the alcohol period 4 c.c. of alcohol per kgm. were given
with one meal, instead of two separate portions of 2 c.c. per kgm., as
before. The only visible effect of the alcohol was a slight unsteadi-
ness of the hind limbs.

\
EXPERIMENT VII. — Doc — CoMPOSITION OF EXCRETA.

Alcohol given, 2-7 c.c. per kgm. Prolonged administration of
alcohol.

A bitch weighing 12.3 kgm. was brought into nitrogen equilibrium
on the same diet as in the previous experiments with a daily nitro-
gen intake of 7.3 gm. and 600 calories. The food was divided into
two meals to which the alcohol was added in equal amounts begin-
ning with 1 c.c. per kgm. at each meal. When the alcohol intake
was less than 3 c.c. per kgm. twice a day, the food was eaten with
more avidity. The alcohol intake for March 6 and 7 was somewhat
less than is stated in the accompanying table, owing to evaporation,
as the bitch refused to eat all of each meal at once. The morning
meal contained 3 c.c. per kgm., while the evening portion contained
2c.c. per kgm. For the next few days the morning meal was fed
with a spoon, while the food in the afternoon was readily eaten,
although the animal was scarcely able to stand. During the morn-
ing and early afternoon she seemed to be more amiable, while in
the evening she was quiet, with slow respirations, and could be
aroused only with difficulty.

Even the daily doses of 4 c.c. per kgm. caused distinct staggering,
so that the distaste for alcohol which developed was not unexpected.
Only 4 c.c. per kgm. were given on March 11, owing to the depressed
condition of the animal. The daily analyses were now stopped, but
the fixed diet was continued with 4 c.c. per kgm. of alcohol daily
_ except when note is made of a higher intake in the analytical table.

Lafayette B. Mendel and Warren W. Hilditch.”

16
g f
A |g
Jan. Kgm.
10 9.4
11 9.4
12 9.4
13 9.4
14 9.4
15 9.4
16 6.3
17 9.3
18 9.3
19 9.3
20 9.3
21 9.3
22 9.3
23 9.3
24 9.3
25 9.3
26 9.3
oF 9.4
28 9.4
29 9.4
30 9.4
31 9.4
Feb.
1 9.4
2 9.4

EXPERIMENT VI.!

Alcohol taken
95 per cent

2
ere
ss

37.6
37.6
2
37.2
37.2
37.2
34.2
37.2
37.2
STZ
o7.2
37.2
Sii2
37.6
37.6
37.6

=
2 2 :
ey 28 a
te! 8) —
Be rate : 8
3 p's 3 °
p> & &
=
e.c 1.0- gm
1 300 —21 7.22
2 310 ~20 7.03
3 330 —20 6.68
4 320 —20 6.65
1 415 —19 6.32
2 300 28 6.92
3 270 —~24 6.23
4 350 —~20 6.65
5 330 —22 6.79
6 350 ~19 6.90
7 275 98 6.92
8 350 ~20 6.76
9 305 23 6.55
10 315 ~20 6.76
11 305 ~20 6.63
12 320 —~20 6.35
13 285 22 6.81
14 380 ~20 6.77
15 305 ~20 6.78
16 310 4 6.72
1 335 —22 7.06
2 295 —22 6.47
3 315 294 6.47
4 305 21 | 6.66

1 Daily N. intake = 7.3 gm.; 600 calories. The urine

aft (tan, Nar. Ngee

Influence of Alcohol upon Nitrogenous Metabolism. 17

EXPERIMENT VI.

tee | 3. | pio
5 g ie 3 4S 3
: 5 3 E E Z
< 5 s) =

gm. gm. gm. gm. gm.

24 6.41 .09 ae Al

222 6.21 .09

225 5.60 .09°

24 Sif .09

24 525 ‘09

1 5.80 .09

.29 5.14 08

.28 5.68 .08

Al 5.68 .09

By 5.68 .09

.26 5.83 .09

24 5.79 .09

Riis 5.86 .09

227 5.74 -09

.28 5.61 10

24 5.38 .09

roe 5.84 10

24 5.76 .09

30 5.81 .09

226 5.65 .10

ELS 6.13 .09

.24 5.49 .09

25 5.70 .09

GAN 6.11 .09

throughout the experiment was acid to litmus.
EEE —————————————E

oT TS TV's8 ss 770" 97° cL 0s'9 Iv 88 = OSZ OT $09 Beer ji
— oL= oT £9°8 es £70" 87 pa] hg £V'9 (Gia OT'8 R= C97 6 $09 Ear 9
s on cv IVs eg° T¢O° 0c OL 9S°9 OG: 88L Oye — CoP 8 O'T9 GIG ¢
=
ae = Or asks ey’ £70" 07 tS) 60'S BGs to Com ose L O'T9 Ge v
A |) Nae ES: BOL ¢¢ 170° (G6, Cr 86'S Se SOL Goa STe 9 8°8P Gack £
= Ot (Ga 98°L es" £70" ST’ tT oo \We sy Sod ce= OPE ¢ 8°8P COV 6
: 60- $9" LOL es” 070" 07 tT LOS VC FOL i Ore $97 Vv £9¢ TZ ue
a wos Ise is es" cCO" ale aa oes 07 2619 co= Cre iS 9°9¢ (GG)! 38C
> o0= 6Vv he es" 810° (ae Or OFr'9 97 Oc L 0c— O87 c VVC (ma! Lé
5 LO = ov HL es" 170 0c vl ce'9 V2 Beez OG ole T 9'P~E orOL 97
= ce (AS : 88°L 99° £00" ol tT LV9 cw Cas OGee $67 £ dg £° TT S14
3 co— cs” SOL 99° £00° 0c bon F119 87° IG (Ho ore 6 re CT 144
= i= + Wed, 99° 700" 6r ol 7S'9 97° SOL (fb 0672 T vies SG £7
ed a | ree Se (a Ra a me es ea
a
See ea lorie | sete fe | Bl oe | 2 Re eR ae eee
oy | Fe ee a ee Balle Ae: lege | cruel
wn)

‘VIGAOXY FO NOILISOdNOD) — 90d,

TIA LNANTddd xa

Influence of Alcohol upon Nitrogenous Metabolism.

"snyWIOA Aq po}VUTUIL}JUOD OUTTA) ¢ : “poids Ayjeyuaprooe sem ouTIN oY} Jo IV z
“SUI O} Plo’ SBAL JUOUITIOdX9 9Y} JNOYSnNoIY} sur syy, “satoye9QQ9 “WIS oy = ayLJUI N AlIeq

e001 | F726
66 C6L
86 0°99
L6 8S
96 8°cS
OL £°0S
69 00s
09 80S
Os 80s
ee 8°98
ce 9°6P
87 8°87
02 v'S8
tT V8P
eT vss
cl Cel

st— Sv’ e138 es e720" 07° I 86°S OV 09°L 9¢— S67 TT OTS | .oCT 8

20 Lafayette B. Mendel and Warren W. Hilditch.

The food for each day was divided into two portions: one, con-
taining the alcohol, was fed by spoon; the other was eaten volun-
tarily. Within four or five hours after the meal the dog was quiet
and somewhat drowsy. She was easily aroused, and although willing,
she was unable to respond when called, owing to apparent lack of
control of the hind quarters. The drowsiness appeared to diminish
after about the fiftieth day and the lack of control was rarely per-
ceptible after the seventieth day, except at the end of the experi-
ment, when the dose was larger than 4 c.c. per kgm.

During the alcohol period the urine was analyzed for the first
fourteen days and then on the twentieth, twenty-eighth, thirty-
second, thirty-third, fiftieth, sixtieth, sixty-ninth, seventieth, and
finally during the ninety-sixth to one hundredth days. In the in-
tervening days the dog was allowed more freedom in a larger pen,
the diet (including 4 c.c. of alcohol per kgm. of body weight) re-
maining unchanged.

The adaptation of the body to the amount of alcohol given (4 c.c.
per kgm.) is clearly shown by a study of the accompanying table of
analyses. As early as the thirty-second and thirty-third days the
distribution of the nitrogenous constituents of the urine is seen to
be practically normal except for the purine nitrogen, which through-
out the experiment has shown the first and most marked changes.

The approximate nitrogenous equilibrium of the thirty-second
day was accompanied by a slight gain in body weight, with a gradual
increase from 12.4 kgm. on that day to 13.2 kgm. at the end of the
experiment. The total nitrogen output was increased somewhat
when the alcohol intake was increased to 6 c.c. per kgm. The crea-
tinine output remained practically constant. A trace of protein
appeared in the urine beginning with the fiftieth day of the
experiment.

The excretion of glycuronates as measured by the rotation of the
urine gradually decreased from —3.4 V° on the twentieth day to
—.g V° on the ninety-seventh day with a slight increase for each of
the next two days, when the alcohol intake was increased to 5 and.
6 c.c. per kgm.

That the adaptation of the body to 4 c.c. per kgm. was disturbed
when the daily intake of alcohol was,increased was clearly shown
by the reappearance of staggering and drowsiness together with the

Influence of Alcohol upon Nitrogenous Metabolism. 21

changes in the excretion of total nitrogen, purine nitrogen, and
glycuronates, with vomiting on the last day.

A post-mortem examination a few days after the close of the ex-
periment failed to reveal any gross pathological changes in the
organs.

DISCUSSION OF THE EXPERIMENTS ON DOGS.

The three series represent the effects of varying doses of alcohol
under comparable dietary conditions, the effects being pushed in the
last prolonged experiment to the condition of chronic intoxication.
In this case the dog even gained somewhat in weight. The utiliza-
tion of protein was found satisfactory whenever it was determined.
With the smaller doses (1-2 c.c. per kgm.) the protein-sparing
action was apparent in the diminished urinary nitrogen output
(see Experiment V). With the larger doses, however (Experiment
VII), the toxic effect is visible in an increased output of nitrogen in
the urine, so that the nitrogen balance became more unfavorable.
That this condition did not become permanent, even after weeks of
continued alcohol dosage, is shown by the figures recorded for the
last week of Experiment VII. With a reduced intake of alcohol
(4 c.c. per kgm.) the nitrogen output showed the characteristic
tendency to return to its normal level.

The experiments show all stages of the effects of alcohol on pro-
tein metabolism, ranging from its sparing action with ‘‘moderate”’
doses to the toxic katabolism accompanying pronounced intoxica-
tion. The partition of nitrogen, if we except the purine metabolism,
is scarcely disturbed in any case. Thus is furnished another illus-
tration of the capacity of the organism to maintain its katabolic
functions along certain normal channels, despite the interference of
toxic agents. Herein doubtless exists an additional “‘factor of
safety’’ for the body, further exemplified in poisoning with adren-
alin,’ hepatotoxic sera,!® hydrazine,!’ cyanide,!* and other instances.

A slight rise in ammonia-N output during the stages of severer
intoxication may be associated with the production of organic acids

15 Cf. UNDERHILL and Cosson: This journal, 1906, xvii, p. 42.

16 Cf. JACKSON and PEARCE: Journal of experimental medicine, 1907, ix, p. 552.

17 Cf. UNDERHILL and KLEINER: Journal of biological chemistry, 1908, iv, p. 165
18 Cf. RicHARDS and WALLACE: Ibid., p. 179.

22 Lafayette B. Mendel and Warren W. Hilditch.

which is suggested by the distinct increase in the levorotation of the
urine. There is some evidence that glycuronates are excreted under
these circumstances.!® We have repeatedly observed the same
phenomenon in the urine of intoxicated animals and patients in the
acute stages of delirium tremens.?? The glycuronates disappear
with the cessation of the intake of alcohol.

With regard to the influence of alcohol on the elimination of purine
derivatives in the dog (Experiment VII) the same features as those
recorded on man are brought to light. It must be borne in mind that
the end product of purine metabolism in the dog is represented by
allantoin, the production of which we have not studied. The recent
experiments by Schittenhelm *! supplement our own in respect to
allantoin, the excretion of which he has found to be increased by
alcohol in both endogenous and exogenous relations. In our experi-
ments extending over weeks the intermediary purines (including
uric acid) continued to be eliminated in quantities much larger than
normal. What has been said from a theoretical standpoint in the
earlier discussion on page g applies here also.

Is alcohol a food? — The argument commonly advanced in favor of
the nutritive value of alcohol asserts ‘that it can replace iso-
dynamic quantities of carbohydrate and fat, and, like them, spare
proteins. For ‘‘moderate’’ quantities (500 calories per day) this is
doubtless true. But we know of no nutrient which will, in compar-
able amounts, increase the katabolic output of purines. Indeed, the
tendency of foodstuffs is, if anything, in the reverse direction.” The
contrast between what is doubtless a ‘‘toxic”’ (or pharmacodynamic)
effect of alcohol and its real nutritive features always deserves to be
kept in mind.

SUMMARY.

A study of protein metabolism and utilization, and especially the
partition of nitrogen in the urine, under the influence of alcohol has

’ Cf. NEUBAUER: Archiv fiir experimentelle Pathologie und Pharmakologie,
1901, xlvi, p. 133.

0 A review of these data will be published later.

*! SCHITTENHELM: Zeitschrift fiir physiologische Chemie, 1909, Ixii, p. 93.

* Cf. KaurMANN and Mour: Deutsches Archiv fiir klinische Medizin, 1902,
Ixxiv, p. 141; RocKwoop: This journal, 1904, xii, p. 38.

Influence of Alcohol upon Nitrogenous Metabolism. 23

been carried out on man and dogs under fixed and comparable con-
ditions of diet. In man the doses used were moderate, 7. ¢€., 500
calories daily in the form of alcohol distributed in six portions. With
the animals a range of dosage leading to distinct intoxication was
employed.

The findings in general were as follows: There is no pronounced
disturbance in the alimentary utilization of the foodstuffs. Moder-
ate doses exert a protein-sparing action, which is succeeded by loss
of nitrogen when larger quantities of alcohol are administered. The
partition of urinary nitrogen remains remarkably unaltered with
the exception of an increased elimination of ammonia-nitrogen (ac-
companying other évidences of perverted metabolism as indicated
by the appearance of optically active (levorotatory) compounds in
the urine) following ‘‘toxic’”’ doses, and a higher output of purines.
The theoretical significance of the latter, which affects both the
endogenous and exogenous fractions, is discussed at some length;
and its bearing on the assumed nutrient properties of alcohol is
indicated.

The most significant impression, perhaps, which the analytical
data afford, is the absence of pronounced alterations indicative of
markedly disturbed protein metabolism, even when comparatively
large doses are continued for days and weeks. This has been inter-
preted as another evidence of the ‘‘factor of safety” in metabolism.

A STUDY OF THE ISOLATED KIDNEY. — THE INFLUENCE
OF PULSE PRESSURE UPON RENAL FUNCTION.

By D. R, HOOKER.
[From the Physiological Laboratory of the Johns Hopkins University.]

i fa 1904 Erlanger and Hooker! published a protracted study of

the blood pressure in two men, one of whom had at the time or-
thostatic albuminuria. In the present connection two points are of
especial interest: first, the amount of urine excreted varied directly
as the magnitude of the pulse pressure; and, second, the amount of
protein excreted varied inversely as the magnitude of the pulse pres-
sure. Blood-pressure changes were induced by exercise, hot and cold
baths, compression of the legs and abdomen, altering the position of
the body with respect to the horizontal, etc. No constant relation-
ship was observed between the functional activities of the kidneys
and either the systolic or diastolic blood pressure. Finally, the tenta-
tive explanation of the facts observed was advanced that the nutri-
tive state of the renal epithelium is dependent upon the pulsatile
variations of the blood-pressure; that these variations act by affect-
ing the amount of blood passing through the organ, and consequently
the oxygen supply to the cells.

In both subjects the blood-pressure changes were of the same kind,
but different in degree, as seen in Table I. The blood pressures
were uniformly lower in the albuminuric, and showed a greater
tendency to instability. This difference was further emphasized by
greater fatigue upon standing still and liability to syncopal attacks.

A study of a second case of orthostatic albuminuria has resulted
in similar findings.?

* ERLANGER and Hooker: Johns Hopkins Hospital reports, 1904, xii, wa,
145.
* Hooker, HEGeMAN, and ZarTMAN: This journal, 1909, xxiii, Dp: x
24

A Study of the Isolated Kidney. 25

These papers are the only ones known to the author® in which
special emphasis is laid upon the pulse pressure as a factor influencing
functional activity, although attention has frequently been called to
the value of a pulsatile pressure, especially in maintaining a normal
condition of tissues when studied under artificial perfusion.*| Whether
the beneficent effect is the result of an increased blood flow® or of

TABLE I.

_ Normal individual.! Albuminuric.?

Dias- Pulse Dias- Pulse
Systolic. tolic. | pressure. | Systolic. tolic. | pressure.

126.8 87.2 39.6 115.8 85.5 30.3
Sitting 127.5 91.0 36.5 113.6 85.7 27.9
Standing Pay By | 99.2 28.5 113.3 92.5 20.8

1 Averages from five experiments under constant conditions.
2 Averages from nine experiments under constant conditions.

the “‘shock”’ consequent, to the pulse, remains an open question,
although the latter has the stronger experimental support.®

The study of the relation of renal function to circulatory changes
in man or in the intact animal is obviously complicated by many
uncontrolled factors. Hence conclusive evidence of such a relation-
ship must naturally depend upon the successful perfusion of the iso-

3 For a résumé of the literature on orthostatic albuminuria see HOOKER:
Archives of internal medicine, 1910, v, p. 491.

4 Jacopi: Archiv fiir experimentelle Pathologie und Pharmakologie, 1892,
xxix, p. 25; Bropie: Journal of physiology, 1903, xxix, p. 266; HOFFMANN:
Archiv fiir die gesammte Physiologie, 1903, c, p. 242; HAmet: Zeitschrift fiir
Biologie, 1889, xxv, p. 474.

5 ERLANGER and HooKeEr believed they were able to show that where the
pulse pressure was the only factor changed, the velocity of blood flow varied
pari passu.

6\MettzER: Johns Hopkins Hospital reports, 1900, ix, p. 135; MALL and
WE cH, quoted by Wetcu: Thrombosis and embolism, ALBuTT’s system of
medicine, 1899, p. 254. See also FLEISCHL VON MaArxow: Eine neue Theorie
der Respiration, Stuttgart, 1887.

26 D. R. Hooker.

Ficure 1 A.

FicurE 1.— Perfusion apparatus. The following letters are used: a, reservoir from
which perfusion fluid is drawn to be injected into circulatory system and into which
fluid returned from perfused organ is emptied. Oxygen is continuously bubbled
through the fluid. 6,c, cylinders from which fluid is alternately discharged into
circulation. d, e, piston rods. f, g, h, 7, valve boxes. Flow is permitted in the di-
rection of the arrows. 7, about 1o cm. rubber tubing in an otherwise rigid system.
Serves to absorb secondary pulsatile waves caused by unavoidable mechanical im-
perfections in action of pump. &, crank axle to which power is applied and on
which cam is fixed. The axle travels with sliding carriage (w). 1, cam. Fixed on the
axle (k) and bearing on point (x), it causes the sliding carriage (w) to move forward
and back when the axle (k) revolves. This alters relative position of k with re-
spect to cylinders (b and c), thus controlling the velocity of movement of pistons and
so shape of pulse curve produced. m, point of union of tubes from cylinders. 1,
large air trap. 0, vent through which air bubbles are allowed to escape. #, screw
clamp. Serves to control peripheral resistance and so mean systemic pressure. g‘

lated organ. It is the purpose of this paper to give some of the results
obtained by the latter method of study.

THE PERFUSION APPARATUS EMPLOYED.

The essential parts are seen in the drawings shown in Fig. T.
Fluid is drawn out of the reservoir (a) and into the cylinder (6 and c)
by the back stroke of the piston (d and e), during which movement

A Study of the Isolated Kidney. 27

place where side tube leading to organ originates. s, screw clamp. Serves to con-
trol magnitude of pulse pressure in tube leading to organ. Beyond s tube divides,
one arm going to organ and containing thermometer, the other arm going to maxi-
mum, minimum, and mean manometers. ¢, screw clamp. Substitutes for resistance
of organ in preliminary adjustment of perfusion pressures. v, warm bath. w, slid-
ing carriage, movements of which in response to changing cam radius alter) piston
velocity by altering relative position of k with respect to cylinders (b and c).
x, bearing point of cam. yy, crank axle.

the valve (f and g) is open. The forward stroke of the piston closes
the valve (f and g) and opens the valve (x and 2), giving a systolic dis-
charge of the perfusion fluid into the circulatory system. During the
forward stroke of one piston the other is returning, so that immedi-
ately one systolic discharge is completed another may be begun. The
apparatus thus differs from the cardio-vascular system in that there
is practically a continuous discharge into the circulatory tubes. The

28 D. R. Hooker.

latter are rigid,’ so that each piston stroke must displace an amount
of fluid equal to that forced into the tubular system. The waves of
pressure within such an inelastic system must therefore have a shape
which is the resultant of the forward velocity of the piston plus the
resistance to the emptying of the sys-
tem: Consequently the crest of the
pulse wave so produced will occur at
the time of greatest piston velocity,
and all other points on such a wave
FicurE 2.—To show the type of will bear a direct relation to the piston
pulse curve produced by the ap- : me ;
rarattis when sthetcouneasle (2) velocity at corresponding times.
revolves about a fixed point. The pulse curve. — The apparatus will
yield a standard pulse curve when the
crank axle (k) revolves about a fixed point. An example of
this curve is given in Fig. 2. Obviously such a curve differs
entirely from a sphygmogram. Since, however, it is essentially de-
pendent upon the piston velocity, we can, by altering the latter, alter
the type of pressure curve produced. This is accomplished by means
of the cam (/) and sliding carriage (w). The sliding carriage (w) holds
the axle (k) to which the power is applied and on which the cam (J)
is fixed. A reduction in the radius of the cam at its bearing point (x)
will cause the carriage and consequently the axle to move forward,
thus adding velocity to the movement of the piston. Conversely an

FIGURE 3-— Pulse tracing obtained from the apparatus with the Hiirthle manometer-

increase in the radius of the cam will cause the carriage and axle to
move backward, thus reducing the velocity forward of the piston.
Hence, with a properly constructed cam, we can control the relative
velocity of the piston at any time of its stroke, and so control the
shape of the wave of pressure produced by the discharge of fluid into
the circulatory system. The modification of the standard curve as
employed in the present research is reproduced in Fig. a

” Except for a short section (j). The slight elasticity thus provided serves to

absorb secondary pulsatile waves caused by unavoidable mechanical imperfec-
tions in the action of the pump.

A Study of the Isolated Kidney. 29

I wish here to express my indebtedness to my friend R. G. Van
Name of Yale University, for the method of cam construction used,
’ and for much valuable advice without which the development of the
present method would have been impossible.®

The circulatory path.— This may be seen in the drawing. Both
cylinders join a common tube (m), which empties into the upper part of
the glass chamber (7). From the lower part of this chamber, which
serves as an air trap and from which air bubbles may be permitted
to escape through the vent (0), the tube continues to end beneath the
surface of the fluid in the reservoir (a). At p a screw clamp is pro-
vided, by means of which the peripheral resistance and so the sys-
temic pressure may be controlled. At the point gq a branch tube
leads to the organ to be perfused. This branch tube is provided with
a thermometer and with a connection to the manometers for record-
ing the perfusion pressures. The greater part of the system containing
the perfusion fluid is submerged in the warm bath (v).

The venous flow from the organ is returned to the reservoir (a),
after being measured by the outflow recorder described by Williams ®
and used in the United States Weather Bureau for recording the
rainfall (Marvin’s tipping bucket rain-gauge)."” During an experiment
oxygen or carbon dioxide, or a mixture of the two, is continuously
bubbled through the fluid in the reservoir (a).

The control of the perfusion pressures. — The screw clamp (p) acts
as peripheral resistance and so controls the mean systemic pressure.
In the body alterations in the magnitude of the pulse pressure de-
pend primarily upon alterations in the cardiac output." In the ap-
paratus this might be accomplished by a change in the length of the
piston stroke (change in the length of the crank axle y).- The neces-
sary complication of construction makes such a method impracticable.
The simple device employed serves the desired purpose without ap-
parently any distortion of the.pulse curve (see Fig. 4).

At the point s on the branch tube leading to the organ, another

8 It seems unnecessary to occupy the space required to give the method of
cam construction. I should be glad to furnish it upon request.

® WILLIAMs: Journal of pharmacology and experimental therapeutics, 1910,
i, p. 457.

10 ABBE: Report on the meteorology of Maryland, Maryland Weather Service,
special publication, 1899, i, part III, p. 327.

11 ERLANGER and Hooker: Loc. cit., p. 153.

30 D. R. Hooker.

screw clamp is placed. Constriction of this tube will alter the pres-
sures distally by (1) lowering the mean pressure, (2) lowering the
maximum pressure, and (3) raising the minimum pressure. If there-
fore it is desired to reduce the pulse pressure only, it is necessary to
tighten both clamp s and clamp p. The proper adjustment of these

FicurE 4.— Pulse tracings obtained with the Hiirthle manometer during experiment of
April 5, 1910. To show that the method employed to alter the magnitude of the
pulse pressure does not seriously distort the pulse curve.

clamps is then obtained, when without alteration of the mean pres-
sure the pulse pressure is reduced the desired amount. An increase
of pulse pressure only is obtained by a similar manipulation, the
clamps being opened instead of closed.

Preparation for an experiment.— The apparatus is set up as
shown in the drawing. The temperature of the large water bath (v)
is raised to about 40° C. The reservoir (a) and the glass chamber (m)
are filled with the perfusion fluid. The arterial and venous tubes are
temporarily connected and a screw clamp is applied at ¢ with such a
pressure that the amount of fluid passing this point is approximately
equal to the normal outflow from the kidneys. The pump is now
started and allowed to run until all air is removed from the system
by collection and expulsion from the air traps. The pressures are then

A Study of the Isolated Kidney. an

approximately adjusted. The temperature is noted, and the gas to
be used is bubbled through the reservoir (a).

The animal under morphia and ether is eviscerated, and the su-
prarenal glands are carefully tied off. The aorta and inferior vena
cava are dissected out, and all branches except the renal are ligated
for a distance of about 5 cm. below, and 2 cm. above the renal vessels.
Cannulas are tied into the aorta and vena cava distally and looking
towards the kidneys with the usual technique. Strong ligatures are
laid, but not tied, about the great vessels above the kidneys. A can-
nula is tied into the urinary bladder, its distal opening being well be-
low the hody level, so that it acts as a siphon to keep the bladder
empty. During this procedure the blood supply to the kidneys is at
no time interrupted, and the organs have suffered very little handling.

Connection is now made between the arterial tube and the can-
nula in the aorta, all air being carefully excluded. The clamp which
guarded the aortic cannula is removed, and for an instant the kid-
neys receive a mixture of perfusion fluid and normal blood. The
clamp guarding the venous cannula is removed, and the ligatures
previously placed on the aorta and vein proximally to the kidneys are
quickly tied. The organs are thus isolated on the artificial perfusion
without interruption of their circulation. After the perfusion fluid
has had time to wash out all the blood previously contained by the
kidneys, the venous cannula is connected with the tube leading back
to the reservoir. During the time occupied in washing out the kid-
neys and while the perfusion fluid is thus wasting, opportunity is
offered, if need be, to adjust the pressure conditions more accurately.

DISCUSSION OF EXPERIMENTS.

The protocols of all the experiments here considered are collected
at the end of the paper. The experiments reported were conducted
to study the influence of the pulse pressure upon the amount of
urinary filtrate formed and upon the presence of protein in the uri-
nary filtrate. The technical difficulties of such a study are great,
and many experiments failed completely. Successful experiments
yielded, however, positive results in support of the hypothesis that
_ the magnitude of the pulse pressure varies directly with the amount

32 D. R. Hooker.

of urinary filtrate and inversely with the amount of protein in the
urinary filtrate.

The duration of the perfusions varied from forty minutes to two
hours and a half. In general the longer experiments were those in
which defibrinated blood in a dilution of one to two parts salt solution
was used as perfusion fluid.

The perfusion fluid. —In the earlier experiments Locke’s solution
modified in various ways was employed. (£dema of the kidneys and
of the surrounding tissues was very marked. In the other and later
experiments defibrinated dog’s blood diluted with either Locke’s
solution or with o.g per cent sodium chloride was used. (Edema,
while not entirely absent, was rare and not so extensive. Both per-
fusion fluids yielded data. The defibrinated blood, however, naturally
conserved normal conditions better. The addition of sodium nitrate
to the Locke’s solution was of no help in maintaining the blood flow.

Evidence of normal function.— The justification of conclusions
drawn must depend largely upon the evidence that the organs ap-
proached the normal in function. The distinctly venous color of the
outflow blood evidenced active oxidation. The urinary filtrate was
neutral or very faintly acid to litmus, although the perfusion fluid
was always alkaline. In the early stages of the experiments with de-
fibrinated blood, the urinary filtrate contained very little protein.
In one (April 21) the first urinary filtrate yielded a flocculent precipi-
tate. The next seven samples contained much less protein (not
enough to give a precipitate on boiling with dilute acetic acid). The
remaining samples again yielded flocculent precipitates which pro-
gressively increased in amount. In these experiments also the uri-
nary filtratewas free from blood coloration for a considerable time.
Indeed in some of them it did not occur at all.

After varying lengths of time the formation of the urinary filtrate
began to decrease. This usually occurred roughly coincident to a
decrease in the venous outflow from the organs. It was possible to
maintain the previous conditions by an increase of perfusion pressure.
However, this change was taken as an indication that the organs were
no longer normal, and served to delimit the experiments.

The progressive deterioration which invariably expressed itself by
a gradual cessation of the blood flow might be due to cedema, the en-
trance of air bubbles producing embolism in the finer vessels, or to a

A Study o

f the Isolated Kidney.

b uv ' I C= sq) el ne ae Gal ty Ate oa Ui es (URS), Yeoman Stet 0}
Cons poet eee the eek ese ae eS By ee ek BN
Due dete tee ees de et ee eG cee oe ee ee ei ae
See ames te Sec AE Rc ne ae Be cea a ree Oe ag
1S} IS} Pog Og gf i= oe =e eee =| @oog Ca aa Se ra mee y=
a 2 ee eee or ee ia te eee oro meee aoe Ps = v Sai ee oe as 6

Fs) -

Bee Soe Oe eR Boel e.g oma 5 fo ho eS kD oe eee o
‘S Cary eS Hey os wa a oe Ba ere Se SO rel mg) Stee Gags 3.85 8
tae Sere foe 8 eo We rete) a3 8s Bose te co ve) gece ya ca es eee
= id Nn — x o Us| C ; NS
Bae sete FeO re. oe 8 Ser ST Goo 2-2 © vs aOR ee USE "Se Set uaa egy ce nee te
Riese ee ne OPO eel ri ee ee eg get OT 2 Yon me me ati ee

(‘S oe we peor) LES eal’ ut eB) c. ~
Se Gee Bee ee lg eB Bo BR ee 2 oe Oe el ot fae
SOMES ee OBE Re CRE GBR HOM SBE Say es
) oO . ~ al (b) Oo sf O Om Sle SO) vay jer =f BL Ae) io} ga
BeONO Sere oes oe eS ee oO Ree OSes ey et a eS ow gt OG Ve akh i eae
SSsHhbateonr one Sag te anes hu 8.2 O'S | oe See ce et ae Oa) MON at Mist iat lot iS)
ee ee) oe [OY AIG) lh | sy) TS sh tb) le) BE toy A el Wyn by AO Gen) fetal gD ech TS) Bat ea fob es) Ge

‘oanssoid osjnd jo sosuryo 07 oyeqyy Arvurn oy} Jo
asuodsar yoInb ayy sMoys plod OY, “sutUNJoo AanogoUl OY} JO SUOISINXA 9Y} UsduUYp ATo}a[dUIOD ysow]e 07 oULISISOI JUDLOIFNS Sursod
“AOU! PUL SAATLA dy} Sulpnpoxa Aq pourrzqo sva oanssoid uvout oy ‘omy Aq poydyypnur oq o1ojo.toy} ysnur Sutovsy oy} Jo sonyea oanssoad
MYL “Sojowourut AMo1ow OF Supra] soayea wNUTUIUE pur wuNUIxvU Aq papdooor odoM soanssord uoisnjiod oy, “Anoquie, v uodn Surry
ayeayy Aavutn jo sdorp oy} spsooer sBurovsy oanssoad oy} yyrouoq oul] OY, “Yova spuooes AjATY] JO spowod ur out] oy} saats wWoyo0q oY}
MOF OUI] PUOIOS IY J, “SITJOWTPUI OIqnd UaAds JO MOYINO. snoudd v syuasordar pur yoxonq ayy jo dip oy} Aq paonpoid sem oury styy Jo uoy
-dnatoyul poe ‘oanssaid o19z oy} SI oul] ySOMo] ay, *AMoAOUT JO SoTOUITT[TUE UL Satnssoid uoisnjiod wntuuruW pur winuwExeut oy. JUoserdos

soul] ysouttoddn omy oy, “o16r ‘tz pudy jo WUoULIAdxXO UL paure}qo plodsar orydeas oy) Jo Wey ‘OZIS [RULALIO OY} pry} ouUO ynoqy —°S ayo
FY PET CY ST EI TY CS STS PSP ay VE I Ye YY PY a a Me MOO ON AW ew EW ee we ae We WY L
Ps a a APCS ALI PA RLTLs SUE. OLS SL Oe RC OG I ae aS 2 OC Sem mI aay 0 0
Rronich: DTA LET LTT A TT ATT TI ATT TT SS AS SSE a] a Entered £ We Waa Ql =

34 ; D.R. Hooker:

points. With each increase of pulse pressure there is an increase in
the number of drops of urinary filtrate. The mean pressure is practi-
cally constant throughout (94-97 mm. Hg). The protocols of all the
experiments show clearly that this small difference is insufficient to
account for the urinary changes. Undoubtedly considerable changes
in mean pressure do affect the amount of urinary filtrate,” but in this
paper we are dealing with practically constant mean pressures.

The minimum pressure is lower during the periods of greater ex-
cretory activity. Probably the decrease of this factor as such plays
no part in renal activity.

The maximum pressure is higher during the periods of greater
excretory activity. Change in the value of this pressure is parallel
with the pulse pressure change, and is a necessary concomitant of the
latter. In the method of study employed the maximum pressure
could not be changed independently. It is unlikely that such a
change can occur in the normal circulation. Therefore it is probable
that this factor alone could not influence renal function.

The conclusion that the amount of urinary filtrate formed is de-
pendent upon the magnitude of the pulse pressure is based upon six
experiments in which the value of the pulse pressure was changed
over thirty times.

Blood flow through the organs. —— The venous outflow responds to
perfusion pressure changes, as does the urinary filtrate. In the graphic
record reproduced in Fig. 5 each break in the lowermost line repre-
sents an outflow of 7 c.c. In this, as in all of the experiments, the
rate of outflow shows a steady decrease. The influence of the pres-
sure changes upon the rate of outflow, the figures of which are given
in the protocols, is graphically shown in the accompanying plotted
curves (Fig. 6). The curve of outflow tends to fall in a step-like
descent. The steps are, however, not of the same depth. A consid-
erable fall occurring in a period of low pulse pressure is much lessened,
entirely stopped, or converted into an increase in the following period
of high pulse pressure. This would indicate that the higher pulse
pressure has a beneficial effect upon the blood flow. The influence of
the magnitude of the pulse pressure upon the rate of blood flow
through the kidneys may be reproduced upon an artificial capillary
bed. Thus, when the organ was replaced in the perfusion system by

12 SOLLMANN: This journal, 1905, xiii, p. 253.

A Study of the Isolated Kidney. 35

a screw clamp on the arterial tube, the rate of flow through the con-
stricted area was with a large pulse pressure 48 c.c., with a small pulse
pressure 36 c.c., and with an intermediate pulse pressure 44 c.c. per
minute.

1 2 3 + 5 6 7 Periods.

Ficure 6.— The results obtained in experiment of April 5, 1910, plotted to show the rela-
tion of the venous outflow from the organs to the pulse pressure. The horizontal
sections of the lower (solid) line represent the values for the pulse pressure in milli-
metres of mercury during the different periods. The corresponding horizontal sec-
tions of the upper (broken) line represent the rate of venous outflow in cubic
centimetres; a continuous decrease in the rate of outflow, except in the last period,
will be noted. This decrease is not, however, uniform, but is more marked in the
periods of low pulse pressure. An exception is seen in the second period when an

_ increased pulse pressure is accompanied by an unusually large decrease in outflow.
This may perhaps be accounted for by the incomplete adaptation of the organs to
the experimental conditions.

The beneficial effect of the pulse pressure upon the rate of blood
flow through the organ is evident in each of the three experiments
on this point.

The effect of the pulse pressure upon the protein content of the urinary
filtrate. — Two experiments were performed in this connection.
They both showed clearly that the amount of protein in the uri-
nary filtrate bears an inverse relationship to the magnitude of the
pulse pressure. In several of the experiments with defibrinated
blood, primarily directed to a study of the amount of urinary filtrate,

30 D. R. Hooker.

the urine was examined for protein. The protocols state that there
was no apparent difference in the protein content to correspond with
the differences in the magnitude of the pulse pressure. This observa-
tion was undoubtedly due to the brevity of the periods in which the
urine was collected. When these periods were increased in length,
the differences in protein content came out strikingly enough to leave
no doubt.

PROTOCOLS OF EXPERIMENTS.

The protocols of the experiments above discussed follow:

May 19, 1909. About 1200 c.c. blood obtained from a large dog.
Defibrinated by whipping. Filtered through paper and diluted to
approximately 5000 c.c. with o.g per cent NaCl.

Small bitch eviscerated and the kidneys prepared for perfusion in
the usual manner. Peptone (0.3 gm. per kilo) injected into left
femoral vein.

Pump started at 1 p.m. Dog connected and proximal ligatures
tied. At 3 Pp. M. it was discovered that both ureters had been tied off.
They were distended with urine, which began to flow as soon as the
ligatures were removed.

The rate of blood flow from the organs was not recorded; it was
distinctly venous in color. The blood was “‘arterialized”’ simply by
exposure to the air. The urinary filtrate was slightly blood-tinged.

It was subsequently found that the proximal ligature on the vena
cava occluded the left renal vein. Consequently the experiment
represents the results from the perfusion of the right kidney only.

Perfusion pressures in ee
mm. of Hg. | Urinary
filtrate |Temper-

: ime. a
Sat ars aan ACL LOp StL ature. Time Procedure

Mean. | Max. | Min. | Pulse. > min.

115 150 85 65 10 39.5° C. | 3.07-3.12
4 Bll: Pressure changed

115 135 95 40 7 39 3.16-3.21
5 A Bele Pressure changed

115 150 85 65 1014 | 38.5 3.25-3.30
3.31 Pressure changed

118 135 95 40 6 JS 3.34-3.39
on Pressure changed

A Study of the Isolated Kidney. a7

The data obtained would indicate that the amount of the urinary
filtrate formed is independent of (a) the temperature, (b) the mean
perfusion pressure, (c) the maximum perfusion pressure, and (d) the
minimum perfusion pressure. The only factor which varies consist-
ently with the urinary filtrate is the magnitude of the pulse pressure.

Jan. 17, toto. Perfusion with Locke’s solution, made up as
follows:

36.0 gm. NaCl
1.22 gm. CaCl,
met tieS sahewes eee eeny Oo, on KCl
lee gm. NaHCO;
\ 4.0. gm. Glucose

Fox terrier bitch eviscerated. Ligatures and cannulas as usual.
No peptone used. Animal in excellent condition.

Pure carbon dioxide was fed to the perfusion fluid throughout the
experiment.

Outflow Urinary
from organs. filtrate.
No. of sec. No. of sec. | Tem-

: : | =f
==) eG iiiel xe required | perature. Period.

Perfusion pressures in
mm. of Hg.

fill given for 10 drops

Mean. | Max. | Min. | Pulse. Sa fall

| -

1 Note that a decrease in the values of the figures indicates an increased rate of
flow, and vice versa.

The experiment was interrupted at the end of one hour and thirteen
minutes by the entrance of visible air bubbles into the arterial can-
nula. The duration of the periods of observation indicate the time
which elapsed from the change of the perfusion pressures until the
' readings were made.

38 D, R.- Hooker.

The figures show that change in the mean perfusion pressure has
no conspicuous effect upon either the blood flow through the organs
or the amount of urinary filtrate formed. Alterations in the magni-
tude of the pulse pressure cause both the amount of the urinary fil-
trate and the blood flow to vary in the same sense. There is quite a
steady decrease in the rate of blood flow through the organs through-
out the experiment, which is more prominent in the periods of low
pulse pressure.

The urinary filtrate was neutral to litmus.

March 24, 1910. Perfusion with Locke’s solution made up as
follows: |

(49.0 gm. NaCl

1.22 gm. CaCl,

In 4 litrcs 1.68 gm. KCl
0.8 gm. NaHCO;
4.0 gm. Glucose
9.3 gm. NaNOs.

The sodium nitrate was added in this, and in the experiments of
March 31, April 5, April 9, and April 11 with the hope that it would
maintain vascular relaxation. The present data are insufficient for
a positive answer as to its value.

Oxygen gas used. The temperature throughout was 35°-36° C.
No peptone was injected. The accompanying figures were taken
from a graphic record:

Perfusion pressures in mm. of Hg. D fur
rops of urine

in three
minutes.

Mean. Max. Min.

The amount of urinary filtrate formed varies directly with the
magnitude of the pulse pressure. The difference is not marked,
due probably to the fact that the periods of observation were short

A Study of the Isolated Kidney. 39

(three to five minutes), so that it was necessary to count the drops
of urine almost from the instant the pressure conditions were
changed. In those experiments in which time was allowed for the
kidneys to become adjusted to the new conditions this difference
is very much more striking. No record was obtained of the rate of
venous outflow.

March 31, tg10. Perfusion with Locke’s solution as used in ex-
periment of March 24, except that 45 gm. instead of 40 gm. NaCl
were used.

Oxygen gas was used. The temperature throughout was 38° C.
Considerable cedema occurred. The figures were obtained from a
graphic record.

Perfusion pressures in mm. of Hg.

Drops of urine
in three
minutes.

Mean. Max. Min. Pulse.

118 96

The urinary filtrate follows the pulse pressure. The amount
formed obviously decreased as the experiment progressed. This
decrease may have been due to the progressive oedema. No record
was obtained of the rate of venous outflow.

April 5, toro. Perfusion with Locke’s solution made up as
follows:

NaCl 45.0 gm.
KCl 1.68 gm.
Jam ed SSS es eee CaCle 1.22 gm.
Glucose 4.0 gm.
NaHCO; 0.8 gm.
NaNoz 0.39 gm.

Oxygen gas was used. The temperature throughout was 38° C.
No urinary filtrate, due probably to a severe hemorrhage during the

operation. The experiment served for a study of the influence of the
pulse pressure upon the rate of blood flow through the organs. The

D. R. Hooker.

figures were obtained from a graphic record.

Perfusion pressures in mm. of Hg.

Venous

21
63

90 14

90 152 48 a] 104 | 105 +7

The venous outflow decreased throughout the experimental
period here recorded, except at the end, when an increased outflow
occurred. The figures show, however, that the amount which the
outflow decreases is very distinctly less during the periods of large

pulse pressure than during those of small pulse pressure. This fact
is more obvious in the plotted curves shown in Fig. 6.

Apparently in this, as in all of the experiments here reported, the
conditions are such that the rate of blood flow through the organs
is abnormal, and this abnormality increases as the experiment pro-
ceeds. It is much more accentuated during the periods of low pulse
pressure. In other words, the larger pulse pressure tends to main-
tain the normal condition of blood flow.

There are reproduced in Fig. 4 the sphygmograms obtained dur-
ing the different periods of this experiment with the Hiirthle ma-
nometer. The numbers on the tracings correspond serially to the
blood pressures given in the table, except the last (tracing 8). It
will be noted that the character of the pulse wave remains practi-
cally constant in the different periods in spite of the variation of
the pressure of the writing lever against the drum. There is only
the change in the vertical elongation of the curve which might be

anticipated to occur with altered pulse pressure values.

A Study of the Isolated Kidney. AI

April 14, 1910. Perfusion with defibrinated blood diluted with
an equal part of Locke’s solution (formula of March 31).

Oxygen gas was used and the temperature was maintained at
37° C. The figures were obtained from a graphic record. The rate
of venous outflow was not satisfactorily recorded and is here omitted.

Perfusion pressures in mm. of Hg. }
P = Urinary filtrate,

a. 4 —o drops in five

| :
Mean. Max. Min. Bilce. minutes

92 114 66 48 | 26
‘

98 102 94 8 | Did

The only factor conspicuously associated with the change in the
amount of urinary filtrate is the pulse pressure, the amount of uri-
nary filtrate varying directly with the magnitude of the pulse

pressure.

The urinary filtrate was neutral to litmus.

April 21, rg10. Perfusion with defibrinated blood diluted with
an equal part of o.9 per cent NaCl.

Oxygen gas was used. The temperature was 337-35 C. The
figures were obtained from a graphic record, part of which is repro-
duced in Fig. 5.

This experiment is perhaps the most satisfactory of the series.
It lasted for about two hours and a half. The urinary filtrate varies
conspicuously with the magnitude of the pulse pressure, as does
also the venous outflow from the organs. There is a tendency for
the outflow to decrease in amount throughout the experiment. In
the second and last periods, however, there is an actual increase
coincident to the change to a larger pulse pressure, and in two of
the other periods in which the large pulse pressure replaced the
small, the decrease of outflow is interrupted.

42 DR. Hooker.

Perfusion pressures in mm. of Hg. Urinary

Alceace Venous Decrease
SF = — : Z outflow in | of venous

| : | drops in five min. outflow.
Mean. | Max. : Pulse. five min.

95
95
O2

The urinary filtrate collected during the first five periods was
entirely free from blood coloration. Subsequently the “‘bloody”’
color steadily increased. All the samples when boiled and treated
with dilute acetic acid were found to contain protein. The first
sample yielded a flocculent precipitate. The following seven sam-
ples showed opacity, without evident precipitate. The remaining
samples yielded flocculent precipitates which progressively increase
in amount. It is perhaps fair to assume that the large amount of
protein present in the first sample was due largely to the operation,
and other manipulation of the kidneys, which by disturbing their
circulation or otherwise affected the organs injuriously. With the
return to more normal conditions there was a slight recovery, which
lasted for some time.

There was no apparent association between the amount of pro-
tein in the urinary filtrate and the magnitude of the pulse pressure.
This was doubtless due to the shortness of the observational periods
(five minutes in the first eight), which did not permit of a complete
adjustment to the new conditions before a change again occurred.
This explanation is supported by the experiments especially directed

A Study of the Isolated Kidney. 43

to this point, and by the graphic records which show that the forma-
tion of the urinary filtrate lags behind the change of perfusion
pressure.

The urinary filtrate was neutral to litmus.

June 8, 1910. Perfusion with defibrinated blood diluted with
two and one half parts of o.9g per cent NaCl.

Oxygen gas was used. The temperature was maintained at 39° C.

Two periods only were observed in order to obtain enough urinary
filtrate for accurate analysis and to insure that the organs were
adapted to the pressure conditions under which the urinary filtrate
was formed. The perfusion pressures were:

Mean. Max. Min. Pulse.
IU ee Nee eee mae 122 150 98 52
JL 2 SS eee ee 116 174 74 100

The urine taken from the bladder before the perfusion was begun
contained a very slight trace of coagulable protein. The two speci-
mens of urinary filtrate obtained during the perfusion contained
sufficient protein to yield a flocculent precipitate on boiling with the
addition of dilute acetic acid. The specimen obtained in the second
period with the large pulse pressure contained very distinctly less
protein than that obtained in the first period with the small pulse
pressure. Assuming that the prolongation of the experiment in-
creased the deviation of the functional activity from the normal, we
might expect an increased protein content, such as occurred in the
experiment of April 21, 1910, rather than the contrary. The results
obtained, therefore, indicate that the magnitude of the pulse pres-
sure bears an inverse relation to the protein content of the urinary
filtrate.

Both specimens obtained reacted neutral to litmus, and were
entirely without tinge of blood.

June 30, 1910. Perfusion with defibrinated blood diluted with
an equal part of o.9 per cent NaCl.

Oxygen gas was used. The temperature was maintained at 38° C.

The pressures were maintained until the urinary filtrate was
sufficient in amount for satisfactory tests of protein content and to
insure adaptation of the organ function to the new conditions.

The control sample of urine taken from the bladder was acid to
litmus, and contained only a very faint trace of protein.

44 D, Rk. Hooker.

Perfusion pressures in mm. of Fig.

Period. |-—— - - — Protein in urinary sample.

| Mean. | Max. Min. Puls:.

140 | | 90 Faint trace; more than in con-
trol. :

140 | 5 30 Flocculent precipitate; | much
| | more than in first period.

140 | Faint trace; about the same
amount as in first period.

There was no blood tinge to any of the samples thus collected.
The last specimen was doubtfully alkaline to litmus.

This experiment shows clearly that the magnitude of the pulse
pressure definitely influences the amount of protein in the urinary
filtrate.

SUMMARY.

1. A perfusion apparatus is described which yields a pulsatile
wave of pressure similar to the normal pulse wave and which allows
of alteration in the magnitude of the pulse pressure.

2. The use of the apparatus in a study of the isolated (dog’s)
kidneys yielded the following results:

a. With a constant mean perfusion pressure the amount of uri-
nary filtrate formed varied directly as the magnitude of the pulse
pressure.

6. With a constant mean perfusion pressure the amount of pro-
tein in the urinary filtrate varied inversely as the magnitude of the
pulse pressure.

c. With a constant mean perfusion pressure, the rate of blood
flow through the organs varied directly as the magnitude of the pulse
pressure.

SENSORY CHANGES IN THE SKIN FOLLOWING THE AP-
PLICATION OF LOCAL ANESTHETICS "AND OTHER
AGENTS: —J. HTAYL. CHLORIDE.

By SHEPHERD IVORY FRANZ ann WILLIAM C. RUEDIGER.

[From the Psychological Laboratories of the Government Hospital for the Insane and the
George Washington University.]

ECENT physiological studies of skin sensations and anatomical

investigations of nerve endings in the skin have shown a much
greater complexity of nerve structure and an association of the sen-
sory functions of the nerves which had previously been overlooked.
Following the division of nerves, Head and others have been able to
find a dissociation of skin sensations different from that previously
reported, and this work has led to a new hypothesis. regarding the
functions of the peripheral afferent nerves.!. Examinations of individ-
uals in whom the peripheral nerves have been cut or injured have
shown a marked and hitherto undescribed variation in the sensibility
to light touch, pressure, pain, and temperature stimull.

The conclusions of Head regarding the combination of sensations
following destruction of or injury to nerves are of special interest to
us in the present work. Head finds that the sensory mechanism of
the peripheral nerves consists of three systems:

“A. Deep sensibility, capable of answering to pressure and to move-
ment of parts, and even capable of producing pain under the influence of
excessive pressure or when the joint is injured. The fibres subserving
this form of sensation run mainly with the motor nerves and are not
destroyed by division of all the sensory nerves to the skin.

“B. Protopathic sensibility, capable of responding to painful cutaneous
stimuli and to extremes of heat and cold. This is the great reflex system
producing a rapid, widely diffused response, unaccompanied by any
definite appreciation of the locality of the spot stimulated.

“C. Epicritic sensibility, by which we gain the power of cutaneous

1 HEAD, Rivers, and SHERREN: Brain, 1905, xxviii, p. 111.
45

46 Shepherd Ivory Frans and William C. Ruediger.

localization, of the discrimination of two points and of the finer degrees
of temperature, cold and warm.”’

The following is Head’s account of the sensation changes coincident
with lesions of nerves, although, it must be understood that these
conclusions give only the grosser results of his extensive work:

‘Loss of epicritic sensibility abolishes: recognition of light touch over
hairless parts or parts that have been shaved; cutaneous localization; dis-
crimination of compass points; appreciation of difference in size, including
the accurate discrimination of the head from the point of a pin apart from
the pain of the prick (acuesthesia); discrimination of intermediate de-
grees of temperature, from about 25° C. to about 40° C.

“Loss of protopathic sensibility abolishes: cutaneous pain, especially
that produced by pricking, burning, freezing, together with that of stimu-
lation with the painful interrupted current; over hair-clad parts plucking
the hairs ceases to be painful; sensations of heat from temperatures over
45° C.; sensations of cold from temperature below 20° C.

“After destruction of all cutaneous afferent fibres the part is still en-
dowed with deep sensibility; pressure can be recognized and its gradual
increases appreciated; pain is produced by excessive pressure (measured by
the algometer); movements of the muscles can be recognized; the point of
application of pressure can be localized; the patient can recognize the ex-
tent and direction of movement passively produced in all joints within the
affected area.”’ ?

These conclusions of Head have been accepted in a general way,
but it is more than likely that they will have to be modified slightly
in accordance with further studies and examinations of patients.
For details of the analyses the reader is referred to the articles by
Head and his collaborators and by others.’

The conclusions from these results so greatly alter the views of the
functions of the sensory nerves that we considered it advisable to in-
vestigate the sensory changes following the application of local anes-
thetics to the skin. The present series of papers has a twofold func-

* Heap and THompson: Brain, 1906, xxix, p. 551.

* Heap, Rivers, and SHERREN: Op. cit.; HEAD and THompson: Op. cit.; HEAD
and SHERREN: Brain, 1905, xxviii, p. 241; Rivers and Heap: Brain, 1908, xxxi,
p. 323; Trorrer and Davies: Journal of physiology, 1909, xxxvii, 134; FRANZ:
Journal of comparative neurology and psychology, 1900, xix, pp. 107, 215.

ae Gn SPIO mAs

Ae at ny NS Clg tg is

Sensory Changes in the Skin. A7

tion: to contribute to the analysis of the sensations from the skin,
and at the same time to fill a gap in our pharmacological knowledge
regarding the actions of the so-called local anesthetics and analgesics.

GENERAL METHODS.

Throughout the present series of experiments certain areas on the
hand and arm were selected for careful examination. These areas
were carefully examined before any application of the agent which
was to be investigated and the normal sensitiveness of the part was
determined. For mbst of the work an area on the radial side of the
forearm, about 4 cm. from the bend of the elbow, was selected. This
area was 5 cm. square. It was accurately divided into twenty-five
squares of one square centimetre each. The skin was marked with a
ro per cent solution of silver nitrate, so that at the time of the ex-
periments the tests could be made at definite places, and so that the
areas were accurately defined and kept constant throughout the series
of experiments. After the solution of silver nitrate had been applied
to and had dried on the skin, the lines were developed by the applica-
tion of a photographic developer solution (hydroquinone). This pro-
duced a relatively permanent stain which made the areas distinct and
which persisted for three or four weeks before it became too faint and
had to be renewed. The application of the silver nitrate was always
made at least a full day before experimental determinations on the
part were made.

The area selected for the experiments was partly endowed with
long hairs and partly free from them. In most of the experiments
this part was carefully shaved to prevent disturbances of the light
touch experiments, and the shaving was always done six hours, and
sometimes a full day preceding the experiments, since there was the
possibility (although we had no definite indication that this was the
case) that the application of hot water, lathering, and shaving might
produce changes in the sensitiveness of the skin. In certain experi-
ments it was deemed advisable to leave the part unshaved, and this
was always done when we wished to test the sensibility of the hairs.

In all the work the two authors acted as subjects. Both were in good
health at the time of the experiments, and the tests were made at ap-

proximately the same time of the day. Each had considerable previous

A8 Shepherd Ivory Franz and Wilham C. Ruediger.

experience in acting as a subject in similar experiments and may prop-
erly be considered a skilled subject. At the time of the experiments
general introspective accounts were made by the subject in addi-
tion to the measurements, both of which were recorded by the
experimenter.

In some preliminary experiments on the effects of different agents
on touch sensations we tested the sensibility of the skin to light touch
by means of a camel’s hair brush. In these early tests we found cer-
tain deviations from the normal which could not be expressed in any
but general terms, and it was therefore deemed advisable to use a
more accurate instrument. For effects that pass away rapidly it is
impossible to utilize the von Frey hairs for the determination of
threshold touch values, for a number of touch hairs must be used and
only a few points can be examined with any degree of thoroughness.
The touch weights which have been utilized in this work have the
same fault, for they require a considerable amount of time and are
not suitable for the determination of conditions that are fugitive.
The use of the single von Frey hair with the possibility of adjustment
to different lengths saves some time, but it also requires more time than
is given for these experiments. For these reasons, because it can be
used with little expenditure of time and because it gives relatively ac-
curate results, and, moreover, is at all times comparative in its reading,
we selected the Block esthesiometer, which had previously been em-
ployed by one of us in other work on the skin sensations.* This in-
strument enabled us to determine the amount of pressure necessary
to produce a sensation of light touch if the sensation could be pro-
duced by a pressure under 2000 mg. Since we were concerned with
relatively increased or decreased sensibilities, as measured by the
amount of stimulus necessary to produce sensations, we have given
in the tables which follow the figures on the instrument rather than
the estimations of unit-area-pressure such as have been deemed essen-
tial by von Frey.°

Pain sensations were tested with another instrument similar to the
well-known Cattell algometer, but with a finely pointed needle as the

* FRANZ: Op. cit.; FRANz: This journal, 1907, xix, p. 23.

* von Frey has worked out a formula for the threshold of pressure in relation
to the cross section of the stimulating surface. This is of no special value in the
present work because the area of stimulation was kept constant.

art ee SASS ae Rag ey eR tein pontine

Sensory Changes in the Skin. AQ

stimulating surface, and, consequently, with a much weaker spring.
By the use of both instruments very small areas can be investigated,
and the measurement of threshold values be made for the touch and
pain points.

The ethyl chloride was spite to the skin from a tube which was
held from 30 to 4o cm. from the part which was to be tested. We
attempted to have the central part of the square on the arm affected
by the spray and to have the surrounding area kept normal. At the
time of applying the ethyl chloride the outer area of 16 sq. cm.
was protected by pieces of a heavy rubber bandage and of gauze to
prevent the action of the ethyl chloride upon this area, but in practice
it was found that the spray affected not only the inner square but
that some escaped and acted upon part of the outer area. The amount
of area affected by the spray was always determinable, and in each ex-
periment we noted the amount of area affected and compared the
sensations from this area with those from the areas known to have
been unaffected by the ethyl chloride. By the application of the ethyl
chloride the skin was frozen, but little or no freezing of the underlying
tissues took place. Not more than one application of the ethyl chlo-
ride was made to the same area on one day. In the application of other
agents which were not to be made by means of spray the exact location
of the application could be previously determined and kept constant.
In each case careful notes of the location of the application were made,
and, as in the case of the ethyl chloride, not only this area but the sur-
rounding supposedly normal regions were carefully examined.

After the application of any agent, tests were begun immediately,
z. €., within ten seconds, or as soon as the ethyl chloride tube could
be laid aside and the testing instruments taken up. For obvious
reasons the different forms of sensibility were not tested at each sit-
ting in the same order. At times the sensibility to light touch was first
investigated; at times temperature sensations; at times the hairs
were stimulated by a brush or by pulling; at times the pain thresholds
were first investigated; and at other times the general feelings of the
subject when his or the experimenter’s finger were brushed lightly over
the area were described and noted. At a sitting we were usually able
to make tests of only two or three kinds of sensibility.

_ Itis necessary to mention that it was impossible to have the appli-
cation of the ethyl chloride and other agents exactly alike in all ex-

50 6Shepherd Ivory Franz and William C. Ruediger.

periments. It is well known that the amount of the freezing and the
extent of the frozen area depend upon a number of factors: the con-
dition of dryness or moistness of the skin, the amount of the ethyl
chloride employed, the height of the tube from the skin and the con-
sequent spreading of the spray, the temperature of the skin and of the
room and the consequent rapidity of evaporation. A sufficient num-
ber of experiments were performed, however, so that we are satisfied
the anesthetic or analgesic results, 7. e., from a sensation standpoint,
are due to distinct temporary alterations in the irritability or conduc-
tivity of the end organs in the skin.

RESULTS.

When the ethyl chloride is sprayed on the skin, there is at first a
sensation like that from a light touch, probably produced by the rain-
like dropping of the spray; with this there is a sensation of coolness.
The sensation of touch from the application of the spray soon dis-
appears, and the coolness sensation passes over into one of distinct cold,
which in a very short time is felt only at the edges of the area affected
by the spray. On the skin frost is formed, which disappears in a few
seconds, leaving the skin blanched, and at this time it is relatively or
absolutely anesthetic and analgesic. After the passing of the frozen
stage the skin is reddened and continues to be so for a length of time
varying with the amount of ethyl chloride used (or the amount of the
freezing of the subcutaneous tissues). During the stage of anesthesia
the sensibility of the skin to the different forms of specific stimuli is
altered, the amount of alteration, 7. e., the degree of anesthesia, and
the duration of the change varying for the different sensations.

Temperature sensations. — Immediately after the disappearance of the
frost, temperature stimuli are not felt on the sprayed area except as
pressures. Soon the area becomes sensitive to cold and cooled objects,
but these have not the same sensation effect as they have on normal
areas. At the time the sensations of coolness can be aroused the ap-
plication of warm or even of hot objects produces no corresponding
sensation and they are felt only as pressures. The correspondence in
sensation between the normal and the sprayed areas does not occur for
a relatively long period, the difference being noticeable for a longer
time than the disturbance in touch sensation. Following are some
records of the results of experiments in this field:

Sensory Changes in the Skin. 51

April 24, 1909. — Subject F. Anesthesia slight. Fifteen minutes after
application of ethyl chloride, a cold glass rod was appreciated as cold in
the normal area, but only as cool within the area acted upon by the
ethyl chloride. Warm objects did not produce the sensation of warmth
when applied to the anesthetic area.

June 10, 1909. — Subject R. Anesthesia marked. Two and a half min-
utes after the application: a warm metal rod (45° C.) produced the
sensation of warmth, but it was much less distinct than when applied
to the normal area, a cold stimulus (8° C.) was felt to be decidedly
cold in the normal area but only cool in the anesthetic area. Forty-five
minutes after application: cold (9° C.) was felt as cold in both normal
and anesthetic areas, and warm (40° C.) as warm. Im all the experi-
ments the sensations appeared to be more distinct when the stimuli
were applied to the normal area.

June 10, 1g09. — Subject F. Anesthesia marked. Two minutes after ap-
plication: hot (60° C.) was felt to be hot in the normal area, but only
warm in the anesthetic area; cold (6° C.) cool on the sprayed area, cold
on the normal area. Fifteen minutes after application: warm (40° C.)
was felt as warm on the normal area, indifferent on the sprayed area;
cold (8° C.), cold on the normal and only cool on the anesthetic area.
Thirty minutes after application: warm (37° C.) and-cold (8° C.) gave
approximately similar sensations from both areas.

June 12, 1909. —- Subject F. Anesthesia marked. One and a half minutes
after application: 10° C. felt cold in normal, cool in anesthetic area;
40° C. warm in normal and indifferent in anesthetic area. Thirty
minutes after application: 16° C., cold in both areas, but with a slight
dulling in the area to which the ethyl chloride had been applied; 57° C.
felt warm and apparently the same in both areas. Sixty minutes after
application: the temperature sensations were alike in both areas.

June 13, 1909. — Subject R. Anesthesia marked. One minute after ap-
plication: 10° C. felt cool in both areas, but colder in the normal zone;
50° C., warm in the normal but no temperature sensation in the anes-
thetic area. Thirty minutes after application: no difference in tem-
perature sensations were discovered.

From these notes it will be seen that the loss of ability to properly
sense warm and cold objects continued for only a short time after the
application of the ethyl chloride and that partial recovery took place
within about two minutes after the freezing of the part. One pecu-
_ liarity was observed, v7z., that soon after the beginning of recovery the
hot and cold stimuli did not produce corresponding sensations of hot-

52 Shepherd Ivory Franz and William C. Ruediger.

ness and of coldness in the anesthetic part, but only the sensations of
warmth and coolness. This difference in sensation was further brought
out when, instead of applying the point of the stimulating areas to the
skin, the length of the stimulant was applied so that it partly stimu-
lated the normal and partly the anesthetic areas. The description of
the resultant sensation given by the subject was that the stimulus felt
hot at either end and only warm in the middle (the rod being laid across
the anesthetic area so as to stimulate it and the normal areas on either
side), or that it felt warm at the ends and indifferent (or only a pres-
sure) in the middle. These latterly mentioned results are like those
noted above, and are in line with results obtained by one of us in the
examination of temperature sensibility after the lesion of nerves.®

Sensibility of the hairs. — When a hair on a normal part of the skin
is lightly brushed, there results a sensation similar to that of light
touch; when it is pulled, the sensation becomes clearer and more inten-
sive, and when sufficient traction is exerted, there ensues a distinct feel-
ing or sensation of pain. This pain appears to differ in character from
that produced by extremes of pressure, as for example that produced
by an algometer, for it is rather burning in character. When traction
is further increased, the hair may be plucked out with its bulb, resulting
in a stinging pain almost like that of a burn. Both the light and heavy
tractions causing pain produce sensations which are accurately localiz-
able, while the brushing of the hair gives a rather poorly localizable sen-
sation. If the hairs be very long, the sensation resulting from light
brushing may be localized on the average closer to the part of the skin
over which the end of the hair lies, but the sensation from plucking is
always localized, as has been said, accurately, 7. e., at the point of in-
sertion of the hair in the skin.

In the experiments variations of both of these forms of sensibility
were found. The following are protocols of a few experiments:

May 8, 1909. —- Subject R. Slight amount of ethyl chloride, not sufficient
to freeze the skin. Within twenty seconds after the application cf the
ethyl chloride there was found no variation in the sensibility of the
hairs; the sensations like those of light touch (from brushing) and those
of pain were apparently normal and were obtained from all the hairs
stimulated.

® Franz: Journal of comparative neurology and psychology, 1909, xix. 19.
See especially pp. 223-233.

Sensory Changes in the Skin. 53

May 8, 1909 (second series). — Subject R. Skin frozen to a medium
degree. Immediately after freezing the skin, it was found that the sen-
sations from traction were present, although at first there were no
sensations from lightly brushing the hairs. The latter sensations
returned rapidly, within three minutes, although for five minutes a
small area of the arm, about 5 cm. in diameter, was still dull (to the
stimulation of light touches on the skin, not on the hairs).

May 29, t909. — Subject R. Marked freezing. Twenty-one seconds after
the application: the sensations from lightly brushing the hairs were ob-
tained, but they appeared to be dull; the sensations from traction were
absent, and no pain resulted even when the hair was plucked out. In
fifty-five seconds: a dull pain from traction was first felt, and this
state continued for about a minute and a half. Two minutes after the
application of the ethyl chloride, the sensations from brushing the hairs
were normal, but in five minutes the pain resulting from the pulling of
the hairs had not become normal in quality (or quantity).

May 29, 1909 (second series). — Subject R. Anesthesia of medium degree.
Ten seconds after application: touches felt on the skin, but the move-
ment of the hairs was not accompanied by a sensation. In twenty-three
seconds: there was no sensation, no pain, on pulling the hairs, but the
sensations similar to touch appeared to be normal. Sixty seconds after
application: pain on pulling returned, but the sensation was not so
keen as in normal parts. For two minutes the sensations accompany-
ing traction remained different from the normal.

May 209, 1909 (third series). — Subject R. Medium degree of anesthesia.
No change from the normal was noticed when the hairs-were brushed
or slightly moved, but on traction pain was first felt in one hundred
seconds, and this kind of sensation continued to be duller than that on
normal parts for five or six minutes.

May 29, 1909. — Subject F. Slight degree of anesthesia. The sensations
from stroking the hairs was normal when first investigated, but no pain
on traction appeared until fifteen seconds after the application of the
anesthetic. The pain from pulling was acute at thirty seconds, and at
sixty was diffuse and not localized as is the pain from normal parts.

June 10, 1909.—- Subject R. Anesthesia marked. The brushing of the
hairs was first felt in about thirty seconds, but at that time no sensa-
tions were obtained from the pulling of the hairs. The pain sensations
from traction returned in two and one half minutes.

June 12, 1909.— Subject F. Anesthesia medium. Immediately after
laying down the ethyl chloride tube the brushing of the hairs was
begun and found to give a normal sensation. There was no pain trom

54 Shepherd Ivory Franz and IWilliam C. Ruediger.

traction until fifty seconds, and the sensations continued to be duller
than normal for a comparatively long period.

June 12, 1909. — Subject R. Anesthesia medium. After one second brush-
ing was felt; after ten seconds traction was felt, but it was not painful;
the pain sensation from traction: appeared at twenty-one seconds, but
was much less acute than on normal parts; thirty-four seconds after
application, the sensations from traction were quite acute, but ap-
parently duller than normal.

From the accounts of the experiments it will be seen that after the
anesthesia usually the sensations from lightly brushing or disturbing
the position of the hairs returned sooner than the sensations from
traction on the hairs. At times there appeared to be a marked dissocia-
tion of these two forms of sensation, and a result similar to that obtained
on a case of nerve division was obtained.’ It appears that these
results indicate the normal dissociation of these two forms of
sensibility. That the sensations from brushing returned sooner than
the pain and pressure-like sensations from traction, and that the brush-
ing sensations were sometimes present even from the first indicates
that the traction sensations are not exaggerations of the brushing sen-
sations, or, in other words, that the pain and pressure-like sensations
resulting from traction are not due to the increased stimulus alone,
but are separate kinds of sensation or are mediated by separate nerve
endings.

Light touch. — The sensations of light touch differed in intensity or
clearness in the different experiments, and appeared to be affected by
the anesthesia a much longer time than the sensations of temperature
and those from the stimulation of the hairs. Like the two forms of
sensibility already described, the touch sensations were not much
affected by slight degrees of anesthesia, but they were abolished if the
anesthesia was marked.

The results of the experiments in this field are given in Table I. In
this table are given the results of seven distinct experiments on the
two subjects. The individual determinations have been averaged,
given in the first lines of each test, and the average variations deter-
mined, given in the second lines of the test results. The number of

7 FRANZ: Of. cit., pp. 215-223.

* The results of the experiments on Subject R, April 24, were grouped in sets
of threes after the completion of the series, and the calculation of the average was

iN.

he Sk

Changes 1

Sensory

109
| |
nua | aun ene SV ehete | iced enedietid Oc. | ICT (OT) ee \ poyreyy J Os wal oun
a ODN Se) OL) os Sen Oe) (Ol) ge oe
1&¢, I 'F re {09 SC (6 +0'0F SLT pouluuly |
. (pe. On) Oe = len lies le GO Gio tie | |
LY oT] O'; | POI | OE Zp ounf || 80'T 19¢ Ore O€
we | DERG: (O1) a ‘powury | ees oulen a (or) ate (01) ie |} poyaeyy |} SI lf OT cunt
| op | «) | | oD (Ol) OT
‘i ODN 8 eos (G)es 2: sal =n O} 0 é
89°C a a | QE 9] sls) tas -+0'0F 1G ap pourwuy |
ro, | OV ce (s) oe (fPaeW {| st [poreunt|| oo | ($0) ee ae
| | | > UNIPITY | 4 |} @ ARTA
a (OL) Ol) ¢ itaers (Sz) oS | (OS) Ge eae
£e'97 +66 cy | || peurey, | | 2ot r81 6 pouty |

coy | 642) ay (8) | umpayy st | ye-dy || 06 (21) sh (¢1) ee cath || Z +7 -ady

‘oIsoysodAPy] "[RULION | ‘oIsoy sod APy) *[BULION
*]VULION “RISOU} | | *[ BULLION ; ; “vIsay}
o1sayy | -souv jo | ‘soup | “sae Iso} -Souv Jo “SOUL, *soyeq,
-sodApy “sonra | uoTTpuol) -sod Ay “son[eAa _ UOTIPUd,)
ployseiy] ssvr9Ay | pljoyseiy] ssv19AVy |

“A pelqns- "WL pelqns

“HCINOTH) TAHLIW HIIM ONIAVAdG AALAY GNV TVNAXON ‘HONOT, LHOIT 10 SATOHSANH T,

56 Shepherd Ivory Franz and William C. Ruediger.

experiments we made in each case and from which the average was ob-
tained is given in parentheses. The averages in the series of tests on
the normal areas differed from day to day, partly owing to a difference
in sensitiveness and partly to a difference in the estimation of the sub-
ject. At times the subject would wait for a very clear sensation before
announcing its appearance, and at times a less distinct sensation was
taken as the threshold. On any one day, however, the results are com-
parable for the normal and anesthetic areas. The average variation in
all the experiments was quite large, and this was due to the actual
sensory condition and partly to the inaccuracies in reading the instru-
ment. Where there is a constantly increasing stimulus, as in the use
of this kind of instrument, it is impossible for the experimenter to
catch the exact point on the instrument at which the sensation ap-
peared, and at the same time it must be remembered that the subject
takes some time to announce the fact that he feels the stimulus. The
error of reading the instrument may be considered in many of the ex-
periments to be about one division on the scale, and the error of the
subject, the reaction time we may say, was approximately two divi-
sion, when the threshold value was above ten. In some of the experi-
ments we were unable to determine the threshold value after the anes-
thesia on account of its amount. The instrument when pressed down
to its greatest extent did not produce a sensation, and in these experl-
ments, therefore, the threshold average is given as 40+. From some
early experiments (April and May) it was found that the examination
of a number of points showed a quite rapid return to the normal con-
dition, and that the later determinations were on the average lower
(nearer the normal) than the first determinations. For this reason the
number of points examined in the later series was reduced to ten, which
number could be made in a comparatively short time and have the
sensational condition approximately constant during the whole time.

It will be seen that with slight degrees of anesthesia the touch thresh-
old was not affected (Subject R, April 24). With the more marked
degrees of anesthesia the threshold values were higher for a period of
at least fifteen minutes. No variation of less than 10 per cent, perhaps
it would be better to take 20, can be considered to be distinctive of a

made. The actual determinations have been mislaid (although those on pain
sensations on the same day were preserved) and it was impossible to calculate the
average variations.

Sensory Changes in the Skin. 57

sensory change. However, it should be noted that although the thresh-
old measurements gave a normal or nearly normal average from the
anesthetic area fifteen minutes after the application, the general feel-
ings of the subjects were that the touch sensations were not at that
time exactly like those from normal parts. The difference could not be
described in any other terms than that the skin did not appear to give
sensations as clear as those from the surrounding (normal) regions.
The sensation difference just noted disappeared within an hour, and
after that time there were neither subjective nor objective evidences of
functional alterations.

Pain sensations. —The results of the measurements of the sensa-
tions of pain from pressure are given in Table II. The average, the
average variation, and the number of experiments in each test are
given in the same way asin Table I. It will be seen that in general the
pain thresholds were higher in the areas sprayed by ethyl chloride,
that the effect of the spraying persisted for about thirty minutes, al-
though the period of complete analgesia was of only very short dura-
tion. In some of the experiments the algometer did not measure the
amount needed to produce pain, and these results are indicated in the
table by the plus sign. On June to, subject R was so completely anal-
gesic in the sprayed area that the limit of the instrument was reached
in all of the first series and no pain occurred. On that day after fifteen,
thirty, and forty-five minutes, there were some points that did not react
to the stimulus by a pain sensation, although most of the area had re-
covered partly the sensibility to the extreme pressures. In the last-
mentioned experiments the quotient of the division of the average
threshold for the normal area into that of the anesthetic area is not
high, and would have more nearly approached unity had it not been
for the one or two abnormally insensitive points. The difference be-
tween the proportions for F on June ro, after fifteen and thirty min-
utes, is due to the fact that at the later time one of the ten determina-
tions was abnormally large. This is also indicated by the average
variation.

The effect of the ethyl chloride on the pain sensations may be
summed up as follows: With a slight amount of freezing there was no
pain threshold difference after eight minutes; with a medium amount
_of freezing there was a very slight or no effect in one subject after two
minutes, and only a slight effect in the other subject after fifteen min-

C. Ruediger.

1avit

ory Franz and Will

srl

eO'e

"BULION

soy
-sodA py

)
BS
78

(OD) Lb 7 (OD)
OL
oe

(91)

‘OIsayysodAfy) *;VULIOoN

‘Son[RA
ployseiy] osvioay

uoHTpuED

9C'1

ec I

OL'I

LVI

S0°¢

LOT

*/BULION

~oIsoyy
sadApy

Shepherd I

“A pelqns

Le £01 ;
(Ol) @9S | (9) Shr
01) O9 | (S) -9'F
“Ss SP
OW fee |) (01) f'6-.
69S SP
(OW) (8) 9'TT
+0'08 ULV
cL (O)e S's
LSP COL
CDresis O17
ai onau OLE
CCL ee NED) 07
OG 1 6£
‘osaysodAFy] “[RULION
‘son[R uolIpuor
ployse1yy ssvioAy

"WX pefqns

“HdTSOTH) TAHLY HLIM ONIAVAdS WAIAV AGNV TVNAON ‘NIV LO S@IOHSHYH T,

VW ATavs

poe yy

WINtpa JA

WYSYS

“RISO
-souv Jo

“pouluy

aST

“SOUL,

QL cunt

8 AVI

rz ady

Sayed

Sensory Changes in the Skin. 59

utes; with a marked degree of anesthesia the effect persisted for at
least thirty minutes, and even though the general average threshold
does not greatly differ, after forty-five minutes there was sufficient
indication to show that some of the analgesia remained. To judge
from the averages one could conclude that the hypalgesia remained for
a long time at a fairly constant level after the return of the pain sen-
sations, and it was noticed that the area did not appear to be normal
even after the day’s experiments were finished.

It is worthy of note that the pain on traction of the hairs returned
more rapidly than that from pressure of the algometer. This would
indicate that there were separate end organs for the initiation of the
pain impulses from the two kinds of stimuli, and that the hair end organ
regained its sensibility sooner than the end organ in the skin.

SUMMARY AND CONCLUSIONS.

Ethyl chloride is not only an analgesic, but also an anesthetic.

The anesthetic effect is of short duration, while that of analgesia is
relatively persistent.

The sensibility of the hairs is affected in much the same way as the
sensibility to touch and to pain. The sensations of lightly brushing
them disappears for only a short time, while those for pain and for
pressure on traction do not become normal for a much longer period.

The differences in the reappearance of the touch-like and the
pressure-like and pain sensations from the stimulation of the hairs
indicate that the hairs have two distinct sensory end organs for the
appreciation of stimuli, and they contradict the assumption that the
traction sensations are only exaggerations of the touch-like sensations
obtained by lightly brushing the hairs.

The variation in temperature sensations is similar to that obtained
on section of a nerve and its subsequent regeneration; vz., on the
part acted upon by the ethyl chloride the hot and cold stimuli, when
they are first sensed, are appreciated as warm and cool rather than
hot and cold. The results of the experiments do not indicate any
difference in the nerve supply for these kinds of sensations.

CONCERNING THE SECRETION OF THE INFUNDIBULAR
LOBE OF THE PITUITARY BODY AND ITS PRESENCE
IN THE CEREBROSPINAL FLUID?

By HARVEY CUSHENG ann EMIL GOETSCH.

[From the Hunterian Laboratory of Experimental Medicine, the Johns Hopkins University.]

HE suggestion was first advanced by P. T. Herring’ in 1908

that the faintly staining, granular, hyaline, or colloid masses
seen in the posterior lobe of the hypophysis cerebri represent products
of secretory activity of the epithelial investment which find their
way through the tissue interstices of the pars nervosa and ultimately
discharge between the ependymal cells into the cavity of the third
ventricle.’ This conjecture, primarily based on Herring’s studies
of the cat’s hypophysis, was strengthened by the appearance taken
on by the gland after experimental thyroidectomy,* in which state
he observed a marked increase of the hyaline bodies, which under
normal conditions are seen in relatively scant numbers. As re-
corded heretofore,’ in a report from this laboratory covering the
main results of our earlier observations on the function of the canine

1 Presented before the American Association of Pathologists and Bacteriolo-
gists, Washington, June, 1910.

* HERRING: Quarterly journal of physiology, 1908, i, p. 151.

* It will be recalled that HALLER (Morphologisches Jahrbuch, 1896, xxv, p.
101) believed that the secretion from the epithelial portions of the gland was col-
lected in tubules which emptied directly into lymph spaces contained in the dural
envelope of the gland —a view not supported by EpINGER, SALZER, STERZI, Or
HERRING. PEREMESCHKO, furthermore (Archiv fiir pathologische Anatomie,
1867, xxxvili, p. 329), described a direct communication between the cleft-like
relic of Rathke’s pouch and the ventricle, and in a single specimen from a kitten
HERRING has met with such a condition, though he believes that it is rare. In
our long series of observations on nearly two hundred canine glands serial sections
have never given evidence of any such communication.

* HERRING: Quarterly journal of physiology, 1908, i, p. 281.

* CRowk, CusHinc, and Homans: Johns Hopkins Hospital Bulletin, roro,
xxi, Prersr.

The Secretion of the Infundibular Lobe. 61

gland as modified by various operative procedures, histological
studies of the tissues so far corroborated Herring’s description that
his interpretation of the significance of the hyaline bodies was fully
accepted.

Further confirmation of these views has come from the more
detailed investigation of the tissues of a new series of animals oper-
ated upon during the present year (1909-1910). These findings,
coupled with the ultimate demonstration of a substance in the cere-
brospinal fluid of man which gives physiological reactions similar
to those produced by extracts of the posterior lobe itself, seem to
establish beyond peradventure not only that the hyaline bodies are
the product of posterior lobe secretion, but also that their primary
destination is the ventricular cavity, where they enter into solution
in the cerebrospinal fluid.

In normal states of activity the pars nervosa of the posterior lobe
is sharply demarcated from its narrow investment or epithelial cells
(Markschicht of Peremeschko; Epithelsaum of Lothringer, pars in-
termedia of Herring) by a layer of capillary vessels which do not
penetrate between the cells of the epithelial covering. Above the
interlobular cleft, however, there is a massing of cells of the pars
intermedia type. Here they fuse more or less intimately with the
anterior lobe cells and send off a tonguelike process which closely
envelops the anterior surface of the infundibular stalk. This portion
of the pars intermedia, unlike the investment proper, is highly vas-
cularized. Indeed it is the only subdivision of the entire gland which
receives a collateral supply, being nourished partly by the vessels
from the stalk destined for the anterior lobe, and partly by the more
remote and distant posterior lobe vessels. At this point, further-
more, there is no such clear demarcation between the anatomical
subdivisions of the gland as exists elsewhere.

It is in this neighborhood that one sees the most striking examples
of invasion by histologically unaltered pars intermedia cells into the
very centre, at times, of the nervous lobe (Fig. 1). Instances of this
are not at all uncommon in the supposedly “normal” glands of
man,’ and the appearance simulates the cellular invasion which one

® Danpy and Goerscu: American journal of anatomy, toro (to appear).
7 For example, in a series of seventy-five pituitary bodies which we have had
the opportunity of examining through the courtesy of Dr. ApotpH MEYER, a con-

62 - Harvey Cushing and Enul Goetsch.

—

sees in certain of the infiltrating ectodermal epitheliomas. These
pictures doubtless represent extreme deviations from the physio-
logical normal. However, in glands which have been modified by
experimentation or disease one may see conditions suggesting the

Ficure 1.— Showing invasion of pars nervosa (PN) by unaltered cells of pars inter-
media (PI). From a case of acromegaly, acute symptoms of glandular deficiency
having occurred three days after an accidental surgical division of the stalk, result-
ing in extensive necrosis of anterior lobe.

same process, rare though it may be for completely unaltered pars
intermedia cells to penetrate for any distance into the pars nervosa.
Thus one may always discern at certain areas of the investment an

siderable number of the glands showed this condition in more or less marked de-
gree. The material came from the routine autopsies of asylum patients and there
were no histories of the individual cases.

The Secretion of the Infundibular Lobe. 63

apparent breaking through of cells into the pars nervosa proper. It
is the rule for these cells, however, to undergo a prompt granular or
hyaline transformation, and the resultant masses are found to dis-
tribute themselves in such a way through the posterior lobe as to
give an appearance of streaming upward toward the infundibular
cavity. Hence the suggestion that this is their destination was a
natural one.

Herring’s description of these hyaline bodies, observed chiefly in
the cat, answers very well for their appearance in the canine gland.
At the various points where cellular invasion occurs, the pars inter-
media cells — at times forming islets or acini or tubules in the centre
of which faintly staining hyaline material may be seen, at times
seeming to wander in as individual cells — soon begin to lose their
normal staining reactions and assume the characteristics of hyalin.
Oftentimes when near their points of origin ghosts of nuclei may
still be discernible in the faintly eosinophilic masses, but as they
approach the infundibulum it is less common to find any traces of
the original cell, the masses having become more or less irregular in
shape and quite structureless. As stated, their general tendency
seems to be in a direction toward the neck of the infundibulum, and
the amount of hyalin in the tissue interspaces becomes obviously
greater as the neighborhood of the ventricular cavity is reached.
The substance seems to lie in distinct interneuroglial spaces which
are radially disposed toward the infundibulum. Immediately under-
lying the ependyma, collections of the material may be seen (Figs. 3
and 6), and at times globular masses appear to be extruding them-
selves between the ependymal cells directly into the ventricular
cavity, where not uncommonly an amorphous accumulation of the
substance, not as yet entirely in solution, may be made out.

In the course of our experimental work we have met with far
better histological examples of posterior lobe secretion than are
furnished by the study of normal glands alone. Hyaline bodies in
excess or showing an unusual distribution have been observed under
a number of conditions, of which six types may be selected as
illustrations: §

* Needless to say, it is requisite to the perfection of these histological studies
' that the tissues be removed not only with the greatest delicacy but within a few
minutes after death, and that they be immediately fixed in a preservative which

64 Harvey Cushing and Emil Goetsch.

I. After extirpation of other ductless glands. — In accord with Herring,
the hyaline bodies seem to be more numerous subsequent to a
thyroidectomy, but the most striking examples of their increase
in our present series have been furnished by the glands of animals
after nearly total extirpation of the pancreas. Though our studies
on the functional interrelation of the pancreas and hypophysis will
be reserved for another communication, the excess of hyalin in the

FicGuRE 2. — Mesial sagittal section of the hypophysis of a dog sacrificed eight days after
a total pancreatectomy. V, ventricle; AL, anterior lobe; PN, pars nervosa; PI, pars
intermedia investing PN; C, cleft. (Squared area magnified in Fig. 3.)

posterior lobe after a total pancreatectomy is appropriate to our |
present theme, for there have been few better illustrations of abun-
dant hyalin without associated tendency to colloid formation than

will cause the least distortion. Though fixation in ZENKER’S fluid gives satisfac-
tory results, we have found that BENSLEy’s fluid is by far the best preservative.
This fixative consists of equal parts of (1) a 2.5 per cent aqueous solution of
bichromate of potash and (2) a saturated HgCl. solution in 95 per cent alcohol.
The tissues have all been embedded in paraffin and the sections cut 4 or 5 microns
thick. Unless great care is taken with these thin sections the hyalin will be dis-
solved out, leaving merely the wide-meshed spaces which contained it.

The sections here reproduced were stained in ExRiicn’s hematoxylin and
alcoholic eosin.

The Secretion of the Infundibular Lobe. 65

are furnished by these glands. The normal-appearing gland (Fig. 2)
under magnification shows a definite concentration of the widely
distributed hyaline bodies near the tip of the infundibular cavity
(Fig. 3), where they crowd the tissue spaces (Fig. 4) and break
through the ependyma.

_ Ficure 3.— Magnification of area squared in Fig. 2, showing tissue spaces filled with
hyaline bodies, which are massed at the lower and anterior surface of the ventricle.
(For squared area see Fig. 4.)

2. After mechanical injuries. —It is our impression that almost
any direct mechanical injury will serve to increase the hyaline
masses observed in the posterior lobe, and in view of the apparent
close relation of the posterior lobe secretion to the storage of carbo-

66 Harvey Cushing and Emul Goetsch.

hydrates, it is not improbable that the transient glycosurias which
often follow traumatic injuries of the head are due to an increased
discharge of hyalin in consequence of a posterior lobe lesion; but this,
too, is another story. Ina single case a particularly striking increase
in hyalin was observed after an injection of India ink had been made

Figure 4.— Magnification of area squared in Fig. 3. showing interneuroglial spaces
crowded with hyalin. Dotted lines indicate a few of the masses in the neighborhood
of H.

in the substance of the posterior lobe and the animal sacrificed a few
hours later. Here, in contradistinction to the condition shown in
Fig. 2, one can see the accumulation of colloid ° in a multitude of
new-formed vesicles (Figs. 5, 6, and 7) at the junction of pars inter-

* Lest there be some confusion from an indiscriminate use of the terms colloid
and hyalin it may be advisable to make clear at the outset that we believe that
colloid accumulation in these new-formed vesicles described above is merely a
precursor of free hyaline globules. These, however, may apparently originate
also as a direct transformation of individual wandering cells; hence the nuclear
remains occasionally seen in the hyalin bodies originating in this way rather than
as a product of primary secretion into vesicles.

The Secretion of the Infundibular Lobe. 67

media and pars nervosa. Indeed, the injection mass itself and the
resultant corpuscular extravasation appear to be tending, just as do
the multitude of hyaline globules (Fig. 6), toward the infundibular
cavity, where they may be seen (Fig. 6) massed under the epen-
dymal lining of the ventricle.

3. After partial hypophysectomies. — Particularly interesting exam-
ples have been given by the hyalin secreted into the infundibular
wall by tags of pars intermedia left after “‘total’’ hypophysectomies.

" ae 4

Ficure 5. — Mesial sagittal section of gland after injection into pars nervosa of India ink.
Central streak in PN shows injection mass with some extravasation. Note layer of
colloid globules arising from epithelial investment of PN.

Owing to the projection of the tonguelike process of the pars
intermedia along the anterior part of the infundibular stalk, it is
almost inevitable for a larger or smaller fragment of this portion of
the gland to be left adherent to the stalk, even in what (from the
anterior lobe standpoint) '° we regard as a “total”? hypophysectomy.
These pars intermedia fragments undergo a definite hyperplasia
(indicating a compensatory activation) (Fig. 13), even during the
few days of life possible for the animal deprived of the bulk of the

” For a remaining, even a microscopical, fragment of anterior lobe may suffice
to preserve life, whereas a “total” anterior lobe removal is invariably fatal,
though a considerable portion of viable pars intermedia may remain.

68 Harvey Cushing and Emnul Goetsch.

gland proper, including the entire pars anterior; and in these case°
one may find evidences in the tissues, serially sectioned, of a stream

FicurE 6. — Enlargement of squared area in Fig. 5. Showing upper edge of injection
mass and many hyaline bodies streaming toward and crowded below ventricle.

of hyaline globules passing from the active epithelial fragment
toward the cavity.
A protocol of such an experiment will serve in illustration.

Dog 19 (Series 1909-1910). — Total hypophysectomy; sacrificed on twelfth
day, owing to onset of acute cachexia hy pophyseo priva.
Healthy, mongrel, male puppy; weight, 3.8 kilo. Twenty-four-hour
urine estimation, 180 c.c.

The Secretion of the Infundibular Lobe. 69

November 8, 1909. Operation. — Hypophysectomy without complications:
presumably a total removal (confirmed by subsequent sections of
removed tissue). One testis removed. Glycosuria demonstrated
three hours later.

Ficure 7. — Further enlargement of area squared in Fig. 6, showing formation of col-
loid by cluster of pars intermedia cells, making temporary vesicles (C). Wall becomes
thinned and contents discharge into tissue spaces of pars nervosa. H, hyaline bodies.

November g. — Eating solid food; responsive; seems well. Normal tem-
perature (38.8° C.). Glycosuria persists.

November to. — Good condition. Slight suspicion of reduction of Fehling’s
solution.

November 11 to 18. — Apparently normal in all respects. Wounds healed
per primam; good appetite; lively and friendly. No glycosuria; no
polyuria.

November 19. — First evidence of cachexia hypophyseopriva shown by
sensitiveness to cold; shivering. Temperature, 38.1° C.

70 Harvey Cushing and Emil Goetsch.

November 20, 11 A.M. — Pulse, 72; respiration, 14; temperature, 32.1° C.,
a drop of 6°. Loss of appetite. Definite ataxia; hypesthesia; over-
active reflexes; muscular twitching, etc.

5 P.M. — Pulse, 68; respiration, 12; temperature, 30.3° C. Typical and
very A ronouaced symptoms of aces no improvement on raising
body temperature with external heat (electric pad). Animal sacrificed.

FiGuRE 8.— One of a series of coronal sections of the infundibular block of Dog 19,
Showing base of i-fundibulum capped by a small organized clot (C), which encloses
a minute fringe of pars intermedia, from which hyaline bodies pass toward the ven-
tricular cavity (cf. Figs. 9 and 10).

Autopsy immediately made. In gross, organs showed no abnormalities
beyond hepatic mottling from fat deposit. Apparently total removal
of hypophysis.

Microscopical. — Sections (serial) of base of brain. At tip of infundibulum
is an organizing clot (Fig. 8) and, adjoining the stalk, a fragment of
pars intermedia showing hyperplasia. From this fragment a proces-
sion of hyaline bodies can be seen streaming toward the infundibulum
(Fig. 9). Many of the masses show what appear to be transformed
nuclear remains (Fig. 10).

The Secretion of the Infundibular Lobe. 7k

4. After experimental obstruction. — Not only excessive accumu-
lations of colloid and hyalin, but also a great increase in the
cellularity of the entire pars nervosa, was observed by Crowe, Cush-
ing, and Homans in cases of experimental stalk division and was
pictured in their report." It occurred to us that the condition was

FicurE 9. — Squared area from Fig. 8. Showing cluster of hyaline globules (near H)
spreading from pars intermedia fragment in the direction of the ventricle (V), which
they approach at a different level, as shown in the series of sections.

presumably due to a stasis of the products of secretion, coupled
possibly with a compensatory hyperactivation of the posterior lobe,
whose circulation under these conditions remains intact.
Experiments have been undertaken in our 1909-1910 series to
throw further light on these changes, and in a number of animals
after the usual exposure of the gland a silver ‘“‘clip”’ has been placed

Wt Loc. cit., Figs. 28 and 30, p. 151.

72 Harvey Cushing and Emil Goetsch. :
1

on the infundibular stalk, the procedure being comparable to a simple
ligation of the stalk with an avoidance of the trauma, infiltration,
and scar formation incidental to the earlier operative divisions.
The tissues from these animals have shown the same appearances
which were observed by our predecessors. The posterior lobe be-

4
'

Ficure 10.— Oil immersion enlargement of area adjacent to H in Fig. 9. Showing
granular hyaline bodies with nuclear remains (H, H).

comes extraordinarily cellular; there is a great increase of hyalin in
the tissue spaces; the former sharp demarcation between pars nervosa
and the investment is largely obscured.”

* It would have been interesting to compare the physiological reactions of an
emulsion of one of these obstructed posterior lobes with the posterior lobe of a
normal animal for control, but the histological examination of the tissues
seemed of more immediate interest. It may be noted that two out of the three

animals subjected to a clean-cut placement of the clip on the stalk showed post-
operative glycosuria.

The Secretion of the Infundibular Lobe. 73

A typical protocol of one of these cases may be abstracted as
follows:
Dog 54 (Series 1909-1910). — Experimental “clip” obstruction of stalk;
polyuria; animal sacrificed after fourteen days.
Healthy, white, mongrel, 4.6 kilo (1014 Ib.), female puppy, about
seven months of age. Average twenty-four-hour urine, 100 c.c.

Ficure 11. — Mesial sagittal section of gland after ‘“‘clip” experiment (Dog 54). Arrows
overlie scar where stalk was crushed through during placement of “‘clip,” and circular
clear spaces show the imprint of the “‘clip’’ removed after fixation. Note enlarge-
ment of anterior lobe (AL) from vascular stasis: very cellular condition of pars ner-
vosa (PN), hypertrophy of pars intermedia (P/); mass of colloid and cellular debris
in cleft. PJ, pars intermedia; PN, pars nervosa; AL, anterior lobe.

April 26, 1910. Operation. — Usual approach; gland well exposed and
U-shaped silver clip compressed on hypophyseal stalk, which was not
broken off. No surgical misadventure; good recovery from the
operation.

A pril 28. — No post-operative complications; no glycosuria. Playful and
active. Normal temperature. Polyuria, 320 c.c.¥

18 Tn another of the clip experiments polyuria persisted for a month, on the day

after the operation reaching 1750 c.c., whereas the normal for twenty-four hours
had been 125 c.c.

74 Harvey Cushing and Emil Goetsch.

May 3. — Slight elevation of temperature (39.1° C.) with accompanying
diarrhoea; nevertheless a gain in weight of one pound since operation.

May 9. — Excellent condition; active and playful. Diarrhoea has im-
proved. Polyuria is disappearing (135 c.c.). Animal sacrificed.

Ficure 12.— Squared area from Fig. 11. Including tip of ventricle (V) which had been
pinched off by clip. C, a mass of colloid and cellular debris in upper edge of cleft.
Note hypertrophy of pars intermedia with tortuosity of processes of cleft: very cel-
lular character of pars intermedia (section encloses the remaining small relatively
non-cellular central area, cf. Fig. 11).

Autopsy immediately after death. In gross, organs show no change, though
there is some suggestion of hyperplasia of the adrenal medulla com-
monly seen after hypophysectomies. Block cut as usual from base of
brain, including the infundibular region with adherent gland; silver

The Secretion of the Infundibular Lobe. 75
clip seen embedded in tissues of stalk. Clip cut at the bend and ex-
tracted after fixation of tissue.

Microscopical. Anterior lobe shows no especial change other than that of
engorgement. It seems larger than usual, but there are no cellular
changes; no necroses.

Posterior lobe. — Holes made by clip apparent to naked eye (cf. Fig. rr).
A fragment of tongue of pars intermedia has been crushed off above

Ficure 13. — Enlargement of area squared in Fig. 12. Showing cellular hyperplasia of
pars intermedia without tendency to accumulation of colloid, though there are numer-
ous minute vesicles in the process of formation.

site of placement of clip. Below clip pars intermedia and posterior
lobe show the following extreme alterations from the normal: The
entire pars nervosa is exceedingly cellular: the epithelial cells of the
investment show, in addition to a marked hyperplasia (Fig. 13), a
tendency to invade the pars nervosa (Fig. 12). Hyalin in large amounts
is distributed thoughout the entire pars nervosa as far up as the seat of
obstruction. Above this, though hyaline bodies are present, they are
not more abundant than normal.

76 Harvey Cushing and Emil Goetsch.

6. In transplantations. — Interesting too have been the evidences
of hyaline secretion which has formed in posterior lobes after auto-
plastic transplantations into the cerebral subcortex. As mentioned
in a former paper from this laboratory, a number of experiments
had been undertaken in the attempt to prolong the life of a totally
hypophysectomized animal by immediately engrafting the extir-
pated gland. The more satisfactory ‘‘takes’’ occurred when the
tissue was planted in the cerebrum, necessarily exposed during the
course of the operation. Unaware at the time how minute a frag-
ment of anterior lobe would suffice to maintain life, the tissue grafts,
when studied histologically, were not looked upon as very success-
ful, though in all of them clusters of viable cells were seen. Of in-
terest, however, to our present subject is the fact that in a number
of instances typical posterior lobe hyaline bodies were present in
the brain substance adjoining the partly organized graft, showing
the same appearance and possessing the same staining reactions as
the hyalin to be seen in the channels of pars nervosa or infundibular
stalk under normal conditions.

The hyaline bodies of Herring, which we have chiefly considered,
differ in many respects from the colloid masses often seen encysted
in the pars intermedia, especially where it adjoins the anterior lobe.
Here these accumulations may at times assume large proportions
and occupy large cysts visible to the naked eye, which may even
break into and fill the cleft itself. This substance has a different
staining reaction from hyalin proper and resembles more the ap-
pearance of the colloid in chronic goitre — the residual stage of
earlier periods of hyperthyroidism, according to Marine.

Hypophyseal colloid, furthermore, does not seem to contain the ac-
tive principle, whatever it may be, so easily demonstrable in the
posterior lobe and presumably present in the hyaline bodies. Our
experiences coincide with those of others in this respect. For ex-
ample, we have variously injected an emulsion of colloid taken from
cysts of the pars intermedia and from the cleft of the fresh gland
of the pig and ox without eliciting any reactions worthy of note.
Whether colloid represents the same basic substance as hyalin but

‘’ Crowe, CusHING, and Homans: Quarterly journal of experimental physiol-
Ogy, 1900, ll, p. 380.

The Secretion of the Infundibular Lobe. Te

which has become desiccated and stored in spaces once lined by
actively secreting cells, is purely a matter for conjecture. The hya-
line bodies have been spoken of as thin colloid, but it would seem
almost better to designate colloid as thickened and encysted hyalin.
However this may be, colloid does not give the reactions of the
posterior lobe which we are inclined, with Herring, to ascribe
to the hyaline bodies. Such fresh colloid, if one wishes to call it
colloid, as may be secreted by the newly formed acini of the pars
intermedia, (as pictured in Fig. 7), may be a precursor of hyalin
and may possibly be active.

These new-formed vesicles, as has been described, may be found
in certain experimental conditions distributed in great number about
the circumference of the pars nervosa. The pars intermedia cells
apparently group themselves in a cluster and become separated by
the secreted material, until an acinus lined by a single layer of cells
is formed. The contained material takes a feeble hematoxylin
stain, whereas the substance immediately on its discharge into the
channels of the pars nervosa upon rupture of the acinus takes on
eosinophilic properties. It would be idle to speculate at present
upon the reason for the activity of hyalin and the inactivity of col-
loid, both substances obviously being products of secretion of pars
intermedia cells. The pars nervosa presumably may play a part in
activating the secreted masses during their passage towards the in-
fundibulum and thus serve as something more than a mere channel
of exit. It is not impossible too that the neuroglial and also the
ependymal cells possess some additional secretory function quite
apart from that of the pars intermedia.

Now, largely as a result of the early investigations of Howell, and
of Schafer and his co-workers, it has long been known that the so-
called active principles of the pituitary body — that is, such prin-
ciples as are demonstrable by the usual injection of extracts — are
found only in the posterior lobe. An emulsion of this part of the
gland, when introduced intravenously or subcutaneously, has a
marked effect on arterial tension, a pressor response usually pre-
dominating,” produces diuresis through distention of the renal

1s There appear to be two substances, one having a pressor, the other a de-
pressor effect.

78 Harvey Cushing and Emul Goetsch.

vessels, stimulates the smooth muscle fibres of bladder, intestine,
and uterus, and dilates the frog’s pupil. These responses are suffi-
ciently characteristic and delicate to identify the substance and dis-
tinguish it from adrenalin—a substance the reactions of which
simulate in many respects those of the hypophyseal extracts.

This posterior lobe substance, which may conveniently be called
_“pituitin”’ to distinguish it from extracts of the anterior lobe, “‘hy-
- pophysin,”’ is soluble in water, glycerine, or alcohol, and does not
lose its activity on boiling. It apparently does not exist, at least in
an active state, in more than scant amounts in the pars intermedia,
according to the observations of Schafer and Franchini, with which
the observations made in this laboratory and to which we have re-
ferred are in accord. Furthermore, as has been stated, pituitin, at

least in an active form, is not present in the cystic accumulations of
' colloid which often are found in this part of the gland.”

On a physiological basis, therefore, we are in possession of definite
facts in regard to certain reactions brought about by injections of
extracts of pars nervosa. On the other hand, purely on a histologi-
cal basis, Herring ventured to interpret the nature of the hyaline
masses seen in this same portion of the gland as products of secre-
tion whose destination was the ventricular cavity. This view seems
to be fully corroborated by the appearances furnished by the glands
which we have observed under variously modified states of activity.
Obviously one step remains to be taken before the full significance
of these hyaline bodies of Herring can be appreciated — the exami-
nation of the cerebrospinal fluid for the presence of a substance
capable of eliciting reactions identical with those of an extract of
the pars nervosa itself.

The first opportunity of making such a physiological test of the
cerebrospinal fluid, after the idea had occurred to us, was afforded,
in June, rgto, by a case of congenital internal hydrocephalus. Dur-
ing the progress of the clinical studies preliminary to a decision as to
the proper method of drainage appropriate to the particular form
of ventricular hydrops shown by this infant, a number of punctures
were made, ventricular and lumbar, and large amounts of fluid
(300 to 600 c.c.) were removed from time to time. The ventricles

*© Tt is obvious from the distribution of hyalin that the infundibular wall may
be as active as the pars nervosa, for it often contains a large amount of hyalin.

The Secretion of the Infundibular Lobe. 79

could be drained as readily from a lumbar as from a ventricular
puncture, and the fluid from each source was clear and on the cus-
tomary tests showed nothing more than the characteristics of
normal cerebrospinal fluid.

The first observation was made on June 15 with some of the fluid
which had been evaporated over a water bath to about one sixth its
volume. Five cubic centimetres of the ventricular fluid thus con-
centrated were injected into the external jugular of an anesthetized
dog and of a quiet and unanesthetized rabbit, causing a preliminary
fall followed by a long-enduring rise in pressure, with slowing of
pulse, amplification of beats, and an occasional dropped beat such
as one often sees after injections of posterior lobe extracts. Subse-
quent injections of an equal dose of the concentrated lumbar fluid
gave precisely similar reactions.

On the following day the experiment was repeated on another
animal, with an even more marked pressor response, and a few drops
of the concentrated fluid added to the normal salt solution, bathing
an isolated frog’s eye, caused a prompt and wide dilatation of the
pupil. Positive results were obtained with a third dog on July 18
(cf. Fig. 14), when positive diuresis was observed. The pupillary
reaction also was found to be more prompt and wider to the fluid
preparation than to a freshly prepared emulsion of the posterior
lobe of a dog or to an aqueous emulsion of the dried posterior lobe
prepared by Armour and Company.

Again on July 20 a series of injections in a 6-kilo dog of 5 c.c. of
ventricular fluid in a 30 to 1 concentration gave: (I) with ventricu-
lar fluid, a rise from 132 to 140, following a preliminary fall to 122,
and also respiratory acceleration; (II) after an interval, with /umbar
fluid, a particularly long enduring though low pressor response from
120 to 130; (III) after an interval, with lumbar fluid again, a rise
similar to II, but with higher response, from 120 to 146.

Judging that the rabbit might show still more definite responses,
further observations were made with other specimens of fluid from
this same infant, in various concentrations; and in view of possible
sources of error, not only through bulk of the fluid injected (for 5 c.c.
of normal salt solution may cause a pressor response in a rabbit),
. but through too great concentration of its inorganic salts, we finally
came to adopt a certain standard amount and concentration, namely,

‘9g] 0} asuodsad s0ssaid zYysI|s & Aq pamorjoy Suorqoafut BUIMOT]OF AJaVVIpOWUIUT QZT OF PHT wor [ry AreurAg “UOL}BAYWIIUOD TO} O¢ B UL pINy reno
-UdA JO ‘9°9 ¢ Jo ‘Bop Of g ‘pezoyiswur uv jo rvjnsnf yeudayxo oy} OFU! uotoafur snousaArizur 0} asuodsay *9ZIS jeuIsu0 ay} J[vy AUG — "FT AAAI

=
~~
a PHARAASAAARAIAIILINAIINUUINIRARAAORDIARIOOURMALURAIIRIAIGRRORRL
sib *‘sD9S OUI],
S
Ch = ee ea
YY [=e aaunssoudo
= ‘oler'et Ainr
= ‘bop Oly Q “Uolpeupus0U0d
i) LOPE UE PINJ xejMONIP UDA
Q ‘9'9G Jo uotjafut snouaaeuyuy

EEOC!

i i MIP Hi
a Hau

Harvey Cushing and

The Secretion of the Infundibular Lobe. SI

50 c.c. concentrated to 2.5 c.c. Fifty cubic centimetres of fluid is
an amount which with care can usually be obtained by a lumbar
puncture from man, and 2.5 c.c. of normal salt solution introduced
into the external jugular of the rabbit shows practically no blood-
pressure response through the amount of fluid introduced into the
circulation. In quiet and unanesthetized rabbits partly eviscerated
we observed in most cases marked peristaltic effects with vigorous
contractions of bladder, uterus, and intestines. In one case the
animal was eviscerated, and direct injection into a branch of the
mesenteric artery produced vigorous peristaltic contraction of the
corresponding intestinal loop.

We have made corresponding observations with the cerebrospinal
fluid obtained from eight other individuals as follows: the ven-
tricular fluid from two patients with brain tumor causing obstructive
hydrops; the lumbar subarachnoid fluid from six patients — (1) epi-
lepsy, (1) recent trauma, (2) acromegaly, (2) hypopituitarism with
adiposity. The direction taken by our clinical investigation be-
comes apparent from the selection of these cases. For, unless we
have misinterpreted the significance of these reactions, it is an-
ticipated that it may prove possible to determine by tests of the
cerebrospinal fluid the functional activity of the physiologically im-
portant posterior lobe in conditions associated with primary or
secondary glandular involvement. This naturally will require
many more observations and a certain standardization of methods
with more controls than we have had as yet.

All of the fluids which we have so far examined and tested on the
rabbit have shown blood-pressure responses, often with the primary
fall characteristic of posterior lobe extract, followed by a long-en-
during rise, a slowing of pulse and amplification of the pulse wave.
They have invariably dilated the frog’s pupil; and diuresis with
constriction of the musculature of bladder, intestine, and uterus
has been commonly seen. The reactions have apparently been more
pronounced in the case of the fluids from obstructive hydrocephalus
than in the other conditions from which specimens have been ob-
tained. This might be expected, in view of a continued posterior
lobe secretion into the hydrocephalic ventricles with little possi-
bility of its escape into the general circulation; and it is notable that
patients thus afflicted are apt to be well nourished, as'is true of, ani-

82 Harvey Cushing and Emil Goetsch.

mals with experimentally produced glandular deficiency. We have
had occasion, too, to demonstrate post-mortem in the posterior lobes
of such cases that retention (?) cysts of colloid are apt to be present
and may reach an enormous size,

The fluids, on the other hand, which have given the least response
have been those from the two cases of trauma and epilepsy. Fifteen
cubic centimetres of normal fluid obtained by lumbar puncture in a
dog and concentrated to 2 c.c. gave inconclusive reactions. It will
be necessary to use a larger animal in order to obtain normal fluid
in sufficient amount for further investigations.

A brief summary of two experiences may suffice:

I. Reactions of the fluid from the congenital hydrocephalic. — A well-
nourished child, aged twelve months, with typical internal hydro-
cephalus producing an enormous head, 60 cm. in circumference. A
number of previous tappings, both lumbar and ventricular, had been
made with withdrawal of fluid, which had been found to give positive
reactions (120 to 146 mm. of Hg) in strong concentrations both in a
dog and a rabbit.

Inasmuch as animals of the same weight and species differ greatly
in their reactions to a measured dosage of carefully prepared posterior
lobe extract, the primary depressor response in some individuals being
pronounced, in others inconspicuous, an injection of the concentrated
cerebrospinal fluid was followed after an. interval by an injection of
1/30 gm. of posterior lobe extract (Armour and Company) dissolved in
an equal bulk of fluid.

Experiment, July 22, 1910. — A few drops of a 25 to 1 concentration of ven-
tricular fluid (250 c.c. reduced to 10 c.c.) promptly and widely dilated
an isolated frog’s pupil at rest in dim light in normal salt solution.

A rabbit prepared for the injection was eviscerated, the urine ex-
amined, and when at rest the blood pressure registered 62 mm. (cf.
Fig. 15). Two cubic centimetres of the fluid in its 25 to 1 concentra-
tion were injected into the external jugular. There was a momentary
struggle, with slight irregularity of the blood pressure, followed by a
slow, progressive rise (100 mm. at its highest point), which was long
maintained. Active peristalsis was started up by the injection, even
visible in the usually quiet large gut, and this was followed by a copious
evacuation. The previously emptied bladder rapidly filled. The
urine, previously negative, showed slight reducing properties.

This injection after an interval was followed by 1/30 gm. of dried

‘yurod jsoysry ye “UIU QOQT 01 ZO woJj asuodsar sossaid Buoy ‘[pey Areurtad WSYS Jy ‘wqqri
pezHoyseuvUN pozeIoOsIAd UB JO Ie[NSnf [eusd}xX9 Ul UOI}eI}UIIUOD T 0} CZ B UF PING ABNIVUGA JO I'd Z JO uOTaLUT ‘azis [RUISIIO ay} JTe [ 9:UD —‘¢] aanoLT

obe A

‘So8s OUI]

TTT TTT TT TTT TT TTT TTT TT TTT

aunssaidO

bulai

;

90 Z + 2efuy

‘ bt eal

zgdg
Lnet

und

‘Ol6l'22 “Ir
“UOTpe4}yUSNUOD 1 O} G2
ul Pm} 4e[NoitpUsA eo) A
Jo (q) uorpoaly TX yqqey

~
~
“SY
ics)
~
ay
~S

2Ccretion oO

if.

Harvey Cushing and Emil Goetsch.

posterior lobe extract dissolved in 2 c.c. of normal salt solution. This
solution dilated the isolated frog’s pupil, and when injected gave almost
exactly the same response, raising the pressure without preliminary
fall from 66 mm. to 106 mm. with a very slow return after about eight
minutes to its previous level.

Reactions from the ventricular fluid obstructed by tumor.— A _ patient
with a cerebellar tumor and obstructive hydrocephalus producing the
usual outspoken general pressure phenomena characterizing these cases.
A ventricular puncture was made August 10, 1910, at Kocher’s point of
election, and 60 c.c. of clear fluid primarily under great tension were
removed. This fluid was evaporated over a water bath to 3 c.c. —a
20 to I concentration.

Experiment, August 11, rgro (Fig. 16). — Pupillary dilating prop-
erty not tested. Rabbit with bladder alone eviscerated. Urine nega-
tive for reducing substances. Injection 1. — Into the external jugular
2.5 c.c. of the (20 to 1) fluid was injected, producing the usual momen-
tary struggle and prompt rise from the preceding level at 84, to 114
mm., with slowing of pulse and amplification of beats; respiration
shallow and rapid. Very slow return of arterial pressure to previous
level. Effect on peristalsis unobserved as intestine was not. exposed,
but the bladder first contracted strongly under the injection and then
subsequently filled; the urine showed no reducing substance.

Injection 2. — For comparison with this response after an interval
a control injection was made (and recorded on the same drum raised
9 mm.) with an equal bulk (2.5 c.c.) of normal salt solution, with no
obvious results beyond a momentary rise of a few millimetres. Sub-
sequent injections in this same animal with other 2.5 c.c. emulsions of
pars nervosa, etc., gave positive reactions similar to the one shown.

If our interpretation of the responses which we have obtained
proves to be correct, and they are actually due to the presence of
posterior lobe secretion in the cerebrospinal fluid, an excellent op-
portunity will be offered for the chemical isolation of the active
principle of pituitin. Some preliminary observations have already
been made in this direction. From the fluid of the case of idiopathic
hydrocephalus the ash was separated by Dr. John King, and we
found that a solution of the inorganic salts thus isolated dilated the
pupil, but produced a marked depressor response without a subse-
quent rise in pressure — an effect possibly due to potassium salts in

libular Lobe.

Pun

=
=

mw
©
=

ecretion of t

iz es

‘UOTNOs
yes Teutou Jo yjnq jenba ue jo uoysefut J9qe] B UINAp dues UO [O1}UOS IO “WU FT] 0} FZ Woy asuodsas Jossord Sutmpus-SuoT ‘z0urn4 Aq uoroniysqo

JO esv> WOAF PING AL[NSIUIA Jo uoLyVIyUIDUOD [ 0} QZ B JO'D'D CZ Jo NIQquA e Ul Uo}efur Ue Oj asuodsay ‘azIs ;eUISI0 9y} JTey auO noqy — ‘oT TANT

‘sdas UI SUI]
rrr EET ETE TET TOPE TET TTOTET-TPE TEPPER T-TREE FICO TT TEETER TTT TTT TT PTT
- - -_ —— 7 —-+-~

daund 4OMOT auns ssoud¢

dAund uaddn aunssaidQ,

; Rs JRO} dolu
a AA Ne ee eGR armen ea eet IRS PHN ID S207

Ps,
ee once

a
pA sash Pron, RP EN

u MUD) Terof S 2 patty

nae ae

—

A UNE HUHNE padi anes Ra ANY
a aa yet VW

alata aatatetawnpt yy ye aT MY

pila

‘nero ARAMA ALAAIAOADAAA YAM AAALAAAARAAA A pring VEIT oer arene eo: py

RARE v
ARAN AAAI ee AI AAAS NONLIN SOON ES Ge DATE OP RAILS NI ELL ELI SO ae WWII nanny ANANVAAAKAANAANA N
Vv i ¥

| O16 ®"¥-#q9°Y
90 € F p°4¥OD'D 09
(Aouopey) pin] ] aejno1yus,

86 Harvey Cushing and Enul Goetsch.

concentration. It is conceivable, however, that we may be able to
separate the pressor and depressor substances which are evidently
present in posterior lobe extracts and which exist in widely different
proportions in the extracts variously prepared by the commercial
laboratories. Some specimens of pituitin which we have tested
contain the depressor substance alone.

It is obvious that there are many controls which must be instituted, . ..

particularly in regard to the presence in the fluid of such substances
as choline. This substance, however, is said to constrict the frog’s
pupil and, as Halliburton has shown, it apparently exerts a de-
pressor effect —responses unlike those of pituitin. It is to be
remembered too that cerebrospinal fluid normally contains a trace
of proteid (globulin) and also a substance which reduces Fehling’s
solution and which Nawratski identified as glucose, though it rarely
gives a fermentation test.

SUMMARY.

The object of this communication is to call attention to the pres-
ence of a substance in the cerebrospinal fluid which gives the same
reactions as extracts of the pars nervosa itself, indicating in all
probability that the active principle long recognized as being con-
fined to this anatomical subdivision of the gland is actually secreted
into the ventricular cavity.

This would seem to establish the theory that the hyaline bodies
of the pars nervosa, regarded by Herring as products of secretion
of the posterior lobe —a view supported on experimental grounds
by ourselves — actually discharge, as their histological appearance
suggests, into the third ventricle and represent the source of the
active substance resembling pituitin in the cerebrospinal fluid.

OBSERVATIONS ON AURICULAR STRIPS OF THE CAT'S
HEART.

By JOSEPH ERLANGER.

[From the Physiological Laboratory of the University of Wisconsin.]
!
T has long been known that strips of the sinus venosus and, in

some hands, of the auricle, as well as the entire excised cold-blooded
heart, continue to beat long after preparation, the only necessary
condition being the prevention of desiccation. It has also long been
known that strips of the cold-blooded ventricle can be made to beat
for hours or even for days merely by immersing them in certain simple
salt solutions. On the other hand it would appear that the continued
or prolonged activity of the excised warm-blooded heart, or of a (ven-
tricular) strip thereof, has heretofore been attainable only through
perfusion of its blood vessels.

While the author was engaged in a problem with which we are here
not concerned, he had occasion to perfuse mammalian hearts in the
usual manner and then to cut into them in various ways. In the course
of one of these experiments it was noticed that a bit of the right auricle
that remained in connection with the atrio-ventricular junction by
only a narrow pedicle continued to beat vigorously, although it was
clear that it could not have had any functional or vascular connection
with the rest of the heart. It then occurred to him that it would be in-
teresting to determine whether the strip cut altogether free of the
heart, but immersed in the Locke’s solution that was pouring from
the open heart, would continue to beat. To his surprise it did; indeed
it continued to beat long after it had been completely severed from the
heart and even after it had been connected with a recording lever and
immersed in a bath of Locke’s solution. Eventually, however, it
ceased beating. Then it was found that prolonged series of beautiful

1 A preliminary report of the earlier experiments of this research appeared in
the Proceedings of the American Physiological Society, This journal, 1909; xxiii,
pi XXXiii,

8S Joseph Erlanger.

beats could be obtained repeatedly by stimulating the strip tetanically
from time to time. In this way the beat was maintained, excepting
certain interruptions, for almost seven hours after preparation.
Methods in general. — The method that has been employed more
or less generally throughout this research was briefly as follows: Under
ether anesthesia the animal was rapidly bled to death through both
carotids. The heart was excised, perfused with Locke’s modification of
Ringer’s solution, and, after it had begun to beat, strips were prepared
in various ways and as they were required for study. In some ex-
periments, before excising the heart, a thread was passed through the
auricle by way of the two cave so as to make these vessels clear in
the collapsed heart. In other experiments the veins were marked be-
fore excision by ‘‘S”’ hooks which were passed through each vein
close to its opening into the auricle. The whole of the intrapericardial
venous region was always removed with the heart and usually a part
of the extrapericardial portions of the cave also. The strips in most:
of the experiments were made successively; that is, one was made,
mounted in the Locke’s solution and studied often as long as two or
more hours, when another strip was prepared from the heart which
in the meanwhile had been constantly perfused. Some of the more
common positions of the ‘“‘S”’ hooks as well as the positions of the
strips most commonly employed are shown in Fig. 1. In other
experiments several strips, as many as four, were prepared all at the
same time and mounted in one and the same vessel. The strips were
fixed below; the upper ends were connected with counterpoised levers.
The beaker containing the Locke’s solution was arranged so that it
could be quickly slid up and down a rod and so cover the strip with
Locke’s solution or uncover it. The volume of solution was so large
that its temperature changed but slowly. In some experiments the
temperature was kept practically constant throughout by means of a
burner properly adjusted under the beaker. In most of the experi-
ments the strip was so connected with the lever that the contraction
of the whole length was recorded; in some, so that the contractions
of either or both ends could be recorded, a selected central point being
fixed below, the free ends extending upward to the recording levers.
As arule, the stimulating current was led through the full length of the
strip; not infrequently, however, the electrical connections were so
made through switches that either one half or the other, or the whole

Observations on Auricular Strips of the Cat's Heart. 8g

strip could be stimulated. In a few experiments the strip was stimu-
lated through a pair of platinum electrodes applied by the hand of the
operator to selected spots or regions. Where it was desired to study
the effect of simultaneous stimulation of several strips under the same
conditions they were con-
nected in series in one circuit.
As a rule, the strips were stim-
ulated through and through
with a tetanizing current de- ; g ae
veloped in a Harvard induc- nN Y vee ale
tion coil driven by one Edison- sux” ee a
Lelande cell, type S. The
strength of the stimulus em-
ployed is expressed by the
distance in centimetres of the
secondary from the primary Fem ______ racine
coil. Some idea as to the

_— SUP. CAVA STRIP
+ —-SUPCAVA _-—

\sinus——J——* —-x

: : FiGcureE 1. — Outline of the right auricle distended
strength of stimuli employ g
8 ‘ ge ed with blood, showing the three regions of the

may be formed with the knowl- outer wall, the commonest positions of attach-

edge that-the threshold stimu- ment, and the strips (areas included in dotted

lus with the electrodes applied lines) most carefully studied, namely, the hori-
: zontal superior cava strip and the horizontal

touthe tongue was obtained ... =

2 . inferior cava strip.

when the coil was at thirteen

and rotated through 20°. Unless it is otherwise stated, the strips

were stimulated in air usually as soon as possible after they were

raised out of the solution; they were reimmersed usually immediately

after cessation of stimulation. Cats alone were used. The results

herein recorded were obtained from a study, often extending over

many hours, of thirty-nine hearts.

RESULTS.

Parts that can be revived by stimulation. It was found early in
the course of this research that certain parts of the auricles can be
made to beat when treated as has been outlined above, whereas others
cannot. One of the first problems therefore was to determine as ex-
actly as possible the limits of the responsive area or areas. This proved
' to be a rather difficult matter, for the reason that in the empty heart,
and especially in suspended strips, the anatomically and development-

rele) Joseph Erlanger.

ally different parts cannot be clearly recognized. Furthermore, the
descriptions of this part of the heart in the books at our disposal have
not been clear enough for our purposes, nor do they agree in their
terminology. The recent article by Keith and Flack ? has not materi-
ally helped us in this regard. Therefore, rather than to employ terms
and limits that may not withstand the test of time, we have attempted
to plot our results on the supraventricular parts divided into regions
that can be easily distinguished by means of gross differences in
appearance.

In the outer wall of the cat’s heart distended with blood (Fig. r) the
auricular appendix is recognized without difficulty. The left or pos-
terior edge (heart suspended from aorta and viewed from behind) of
the right appendix is usually perfectly clear. To the left of the lower
half of this edge is seen a more or less triangular area which in some
ways resembles in appearance the appendicular region and also that
part of the supraventricular region lying between the appendix above
and the atrio-ventricular junction below and with which it is con-
tinuous along the lower part of its anterior edge. This triangular
area we have, for the sake of convenience, termed atrium. It is prob-
ably coextensive with the trabeculated parts of the region usually
termed atrium. To the left of the right appendix above and of the
atrial area below is found what we have termed the sinus region. In
the distended heart it resembles in appearance the walls of the cave,
and includes all of the outer wall of the right auricle, including the
intrapericardial parts of the cave, not accounted for above. The
sulcus terminalis is clearly marked only on the upper edge of the
auricle and has therefore been of little assistance to us as a landmark.
We have attempted to determine the position of the sino-auricular
junction by means of the arterial circle described by Keith and Flack,
but have not succeeded in satisfying ourselves in many cases of its
existence as an easily recognizable entity even in injected hearts. There
is, however, one very constant artery which has served us as a land-
mark, namely, one which ascends from the atrio-ventricular groove
somewhat posterior to the appendix over about the middle of our atrial
area (shown in Fig. r). After ascending a short distance, the main
branch curves to the left in the direction of the inferior cava, smaller
branches going in various directions.

* KeiTH and FLack: Journal of anatomy and physiology, 1907, xli, p. 172.

Observations on Auricular Strips of the Cat’s Heart. 91

By passing ligatures through the distended auricle at various points
along the boundary of the sinus and atrium, as above described, the
limits of the sinus as viewed from within can be accurately determined.
It is thus seen that the sinus includes practically all of the non-tra-
beculated parts of the outer wall of the auricle. The so-called crista
terminalis (corresponding with the sulcus terminalis) is apparently a
very broad structure in the cat’s heart. Running over it toward the
superior cava from the right outer edge of the mouth of the inferior
cava is a grayish fold of tissue which is probably a remnant of the
venous valve. This fold lies considerably to the caval side of the right
edge of the sinus as we have described it.

Other regions we have arbitrarily distinguished require no especial
delimiting, namely, in the left auricle, the pulmonary region and the
appendix; the right and left vaults (anterior or aortic surface of the
auricles); and, common to both auricles, the septum and the coronary
region (including the region of the node of Tawara and the atrio-ven-
tricular bundle). In how far these regions will be found to correspond
with those now distinguished by embryologists and comparative anat-
omists we have not been able to exactly determine. We shall, however,
have something more to say in regard to this subject later.

The attempts to revive strips that have visibly ceased beating have
met with success only when they have contained the whole or parts of
the sinus, atrium, coronary, or septal regions, and only when these
parts were stimulated. There has been only one exception to this
rule: upon one occasion a strip composed presumably of only the
right appendix gave two spontaneous beats when immersed after
stimulation in air. With this single exception strips composed only of
right appendix, vault (exclusive of septal attachment), pulmonary, and
left appendicular regions have never been revived.

Furthermore, in the case of strips that are beating spontaneously
or have been made to beat by previous stimulation, an increased rate
of beat subsequent to stimulation is obtained only when they contain
the whole or parts of the sinus, atrium (probably), coronary, or septal
regions and only when these parts are stimulated. Nevertheless, as
will be made clear later, stimulation of the right appendix, vault, pul-
monary, and left appendicular regions is not without effect.

The method that has been employed to determine more exactly the
anatomical limits of these two areas has been to excise practically the
whole outer wall of the right auricle and to suspend it from three

g2 Joseph Erlanger.

points, namely, (1) from the atrio-ventricular groove at about the
place where arises the artery that has already been described, (2) from
near the mouth of the superior
cava or actually from that
Gf ak vein, and (3) from the tip of
the appendix. The first point
was fixed immovably; from
it the sinus region upon one
side, the appendicular upon
the other, extended obliquely
outward and upward to the
other two points of fixation
Figure 2.— Outline of the strip used in Exp. which were connected with
29 (Table I) showing some of the points counterpoised lewene: By
stimulated. ‘i
means of this arrangement
the strip was spread wide open so that a pair of platinum electrodes
could be applied to any desired point while at the same time the
contractions of the two ends of the strip could be recorded sepa-
rately. In every case a sketch was made of the strip as it appeared
when viewed through the sides of the beaker, that is, magnified in
the horizontal axis, and the points stimulated were drawn into the
sketch.

The results, selected at random from a typical experiment, are pre-
sented in tabular form (Table I, columns 1, 2, 3, and Fig. 2).

These, and many other experiments in which the same as well as
other spots were stimulated, show that the accelerating after-effect is
almost invariably obtained when the region immediately about the
mouths of the great veins is stimulated. Possibly it is not so frequently
obtained when the stimuli are applied midway between the veins, and
even less frequently further to the right. It can, however, be obtained,
sometimes in marked degree, as far to the right as the trabeculated part
of the auricular wall. Indeed, on one occasion, and one only, a slight
acceleration resulted when the appendix was stimulated after it had
been separated from the sinus end by means of a cut parallel to,
and a slight distance to the right of, the thick trabecula that lies
just to the left of the left border of the appendix. The accelerating

after-effect of stimulation certainly does not diminish below the
inferior cava.

LOOKING INTO

TRABECULATED
REGION = Noes tee x

Observations on Auricular Strips of the Cat’s Heart. 93

Other experiments show that to the left of the veins the accelerating
after-effect extends into the auricular septum. It does not, however,
extend beyond the junction of the right with the left auricle.’

The veins themselves do not seem to beat, and stimulation, even
close to the intrapericardial parts, has, if we except certain very doubt-
ful cases, seen only when the superior cava is stimulated, been entirely
without any after-effects.

The area, therefore, that responds with increased rate of beat sub-
sequently to stimulation and that can be made to beat spontaneously,
may be described as including the whole of the area we have termed
sinus, the interauricular septum, and to a less degree the atrium.
This area will probably be found to be coextensive with the area,
as determined by Adam,‘ that responds to local warming with an in-
creased rate of beat. It seems to be somewhat larger than His’ sinus
reuniens (including here the coronary sinus),’ but on the right edge
only. Whether or not it is coextensive with the region Keith and
Flack believe contains tissue derived from the sinus venosus we have
been unable to definitely determine. It would, however, appear that it
extends considerably further to the right than their sinus, to include
a part of the region Keith and Flack term the auricular canal.‘
It certainly is more extensive than the sino-auricular ‘‘nodal tis-
sue,” since, according to Keith and Mackenzie,’ “in the mammalian
heart the sino-auricular nodal tissue is concentrated on the right
auricle, chiefly along the sino-auricular junction and in front and on
each side of the superior vena cava.”” The statement of these authors
to the effect that ‘the excitable regions of the heart are those where
the nerves come into a specially direct connection with the special-
ized tissue which we have termed nodal”’ is not, therefore, borne out
by experiment.

Relative rate of beat of the different parts of the spontaneously rhythmical
region. — The relative rate of beat of strips each made up in part
of a portion of the spontaneously rhythmical region of the auricle,

* It is to be regretted that the great length of the specimen tracings illustrating
most of the points made in this paper precludes their reproduction here.

4 Apam: Archiv fiir die gesammte Physiologie, 1906, cxi, p. 618.

* His: Beitrage zur Anatomie des menschlichen Herzens, Leipzig, 1886.

6 Keity and Frack: Loc. cit.

7 Kerru and MAckENzIE: Lancet, Jan. 8, 1910.

Joseph Erlanger.

04.

OF a OIF | 90 TOT oe IG
“pulpoooid se ures noqy Cattery) eeee wear | Ce ete wees Perec) ‘dy JIAO IV
c8 oT 80 | LO CP + ¢£
Ute Fal oe i! ay | ia! 06 Cv I!
iG vl cn m0) OL Qs (¢
oT OT 80 TO OL Qs (6
‘ySa1 Buoy] “-asuryo ON hae ite & Clos ae te etic co Sari wrest | g
‘asuRYyo ON : sore : | re
“ole 0} Bul 4 [0 OT | [yeus Ara A, OLT a9 | C
“UTEM 19}J8 PUP JSAt SUC] 1oI;V Ol [Jews AoA 90 | [jews Aza, OWal OL | I
See cyl O07 OT COT oS | C
aoa ee 9SBIII UL ON O8BI19 UL ON 09 09 | e
o9 OT 90) 0) O'S CS | ¢
"ys V7 yews A109 A oe SZ Gy 09 I
Buoy INVFR 24 OE 0} poster -duwio 7, ésulyraq JON |ésurjyroq JON oe —¢o O'R oe aid
‘asvaroUL ON ea i Stee eee 2»: bine are f
Ca m0) cal SO OTT , | C
= oe oT Gay LO TOL OF |!
aber OF jyetus AIO A LO } 90 CP se €
O07 AY) cab | LO COT CV I
oP z arg OT (0) TOL OS i
09 SuIyvaq JON $0 LO O'S LY 2
gt say Surjeoq JON BULyLAq JON OT, (0) CPU O's C
“WU UT “WU UT “WU UT | “WU «UT ‘90S OT Jod ‘N98 OL Jod | . aVVMUITS
*SYAVULDYT ‘toqye “jdure “AIO JOC “Toye “jdue “d1OJOq Syewag “Joye |syragq ‘arojoq I oe aie
"Y ‘XPT ‘\dwy -y tye || (allude) He d}EI “SC "XPT 9}PI'°S yulod

(@ “OI AAS) 6% “AX ‘SENIOdd GALOATAS ONILVTAWILG AO SLOW] AHL JO ANOG ONIMOHS

lL Ghali

Observations on Auricular Strips of the Cat’s Heart.

"yMsor jens

“IV[NSoI AOA osuodsay
“sures §

“dw J, ) i

~ Le OF SULWIRM Io], Y

SUIMOIS PoyAP]Y

“‘Suyyey “dura y, ) Ivpnso1I1]
of LE
0} poweM Uo, “;BAIOyUT Buo’T

‘asuvYyD ON

“sulyyey “duro 7,

SOF >
“AP Udy Us RUT
poatyoq drys usym ‘279¢

“dure *s

Ul ASKAIDUT JYUSYS st asuryo A[UO

0°
0'8
c°9

“IOUT O

osur

yjeus

jjeuts A193 A
4
90
SL
9'0
L8
OC

sO
yo ON

L0
SO
OP
“Toul ON
07
“Toul “ON
Lie
£0
2
“TOUT 1S
“TOUL 1S

[jews Aro \,
|!
OT

= Sit0)

]jews Aro \
m0)
[jews Aro A
]jeuls Ar A,
jews Aro A,
|Jeus Aro A
LO
[wus Aro \
]euls Aro A.
[jews Ara A,

A. AA. A. A.

ee ee ee

—

=~
=
ee

TMNANOM”W AW!
“AmomMmrwnr
5 =

£01
9
8 0F

—
Oo

—
10) COrm

oD sh

~
i

H~MOtN
“NADOrr- Ar

1 Hun

C
[ jnoqe pur 3y
C
‘dy 19A0 lv
p punory

f punory

CoN At

f purory
‘dy I9A0 TV

6
6

gO Joseph Erlanger.

that is, the region that can be made to beat spontaneously through
electrical stimulation, when subjected as nearly as possible to the same
conditions, is shown in Table II. The maximum rate of beat developing
subsequent to stimulation has been used in compiling these statistics.
The strips were stimulated through and through.

There can be no doubt but that to a certain extent the great varia-
bility of results shown in the table is due to the fact that the stimuli
cannot be applied to the different strips in exactly the same way, man-
ner, and place. Andsince the effect subsequent to stimulation depends
upon the density of the stimulus and the place to which it is applied, it
follows that the results obtained with different strips in the same ex-
periment and with the same strips in different experiments are not
exactly comparable. By way of illustrating this point the following
experience may be cited. With a strip consisting of the whole outer
wall of the right auricle, it was found that under conditions that were
otherwise exactly the same, stimulation ‘‘through” resulted in a beat
at the rate of 54 per minute, whereas applied directly to the sinus with
platinum electrodes, the stimulation resulted in beats at the rate of
go per minute.

Viewing in the light of this difficulty the results tabulated above,
the only warrantable conclusion seems to be that the three regions of
the auricle most carefully studied, the superior cava, the inferior cava,
and the coronary sinus regions, possess approximately the same grade
of rhythmicity.

This conclusion is borne out also by the results obtained from local
stimulation of the outer wall of the right auricle, which have been re-
ferred to above. Whenever acceleration is obtained the resulting rate
of beat under a given set of conditions is approximately the same (see
Table I, columns 1, 2, and 3).

Discussion of foregoing results. The results detailed above are of
considerable interest in view of the physiological importance that has
been attributed, mainly by histologists, to the so-called sinus and
auricular nodes. These two structures have such a peculiar and char-
acteristic appearance histologically that their discoverers have been
tempted to designate them the motor centres of the heart. This sug-
gestion has been accepted by some physiologists and clinicians almost
as a demonstrated fact.

The most important experimental evidence favorable to this conten-

Observations on Auricular Strips of the Cat's Heart. 97

tion is furnished by Hering.* (1) In the first place this investigator,
finding that a cut carried through the auricle so as to presumably inter-

TABLE II.
SHOWING RELATIVE RATE OF BEAT OF STRIPS UNDER APPROXIMATELY THE SAME
CoNDITIONS.
Exp. No. Nature of strip. Temp. ett Remarks.
4 (a) Sup. and Inf. cave Under exactly the a
and intervening tissue ? 138 same conditions.
(b( Septum including cor.
sinus and A-V bundle ic 1 A al Pete a Pann, eae
'
5 (a) Sup. cava ? 96 Under exactly the
same conditions.
(b) Inf. cava ? fg Aa lar nase
(c) Coronary sinus ? 102
6 (a) Sup. and Inf. cave and
intervening tissue .. . | 30— 84 Under exactly the
(b) Vertical strip just to same conditions
SESE OF (GB) Cle) eee a te 30— 114
7 (a) Sup. cava—horizontal 33 144 weeiata
(b) Inf. cava of 33 132
9 (a) Sup. cava s ? 72
(b) Inf. cava « ? 84 teletatale
10 (a) Sup. cava - 33+ 74 g
(6) Inf. cava ss 33+ 96
11 (a) Sup. cava 30 84 Peewee
(b) Inf. cava sf 30 78
12 (a) Sup. cava fe 34? 102
(b) Inf. cava 6 34? ho eee cc Sere
13 (a) Inf. cava 33 (iP
(b) Sup. cava s 33% 114
14 (a) Inf. cava es 34 138 eleielelalon
(b) Sup. cava ie 34? 108

sect the sinus node, and that a cut 1 cm. long in the “furrow which
runs from the apex of the angle formed where the superior cava joins the

8 Herinc: Miinchener medizinische Wochenschrift, April 27, 1909. A review
of Hering’s earlier work on this subject will be found here.

ys Joseph Erlanger.

right auricle” suffices in many cases to bring the supraventricular
parts of the heart to a standstill, “has no doubt but that normally the
impulses of the mammalian heart (inclusive of man) arise in the sinus
node described by Keith and Flack.”’

While there is every reason for believing that a cut that stops the
beat of the heart acts upon the motor centre, this action need not be
directly upon the centre. The effects of stimulation, particularly of
the heart, are not limited to the point of their application, — witness
general fibrillation from local stimulation. Furthermore, in Hering’s
hands a cut made as described by him has not always stopped the
heart, while cuts made in other parts of the sinus region may stop the
heart.®

(2) A second method that has been used to locate the place of
origin of the normal cardiac excitation wave has been to determine in
the dying heart the part last to beat. Many investigators have shown
this to be the region of the great veins. More recently Hering !° located
the uliimum moriens (a) in the mouth of the superior cava in the vicin-
ity of the sinus node of Keith and Flack and (0) in the coronary region
in the vicinity of the auricular node of Tawara. The logic of this
method of determining the normal cardiomotor centre is not, however,
clear. It serves only to locate the most viable parts of the heart, not
those possessing the highest rate of rhythm, which, after all, determines
the seat of the pacemaker of the heart.

This argument suggests another and perhaps the oldest way of de-
termining the region of the heart whence originates the cardiac im-
pulse, namely, by locating the part of the heart endowed with the
highest rate ofrhythm. In so far as the mammalian heart is concerned,
this was attempted by Erlanger and Blackman." By dividing in vari-
ous ways the auricles of the perfused heart by means of cuts and
crushes they found that the region of the great veins possesses the
highest grade of rhythmicity but that the rhythmicity of this part ex-
ceeds but little that of the coronary sinus region. The method used,
as was pointed out by these authors, is open to the objection that the
disturbance of the supply of the perfused fluid to the several parts of
the auricle may have altered their normalrhythmicity. This objec-

* ERLANGER and BLackMAN: This journal, 1907, xix, p. 125.
Herinc: Loc. cit.
11 ERLANGER and BLackMAN: Loc. cit.

Observations on Auricular Strips of the Cat’s Heart. 99

tion does not hold in the case of the present experiments. None of the
specimens was perfused; they were merely suspended in Locke’s solu-
tion. It is true, however, that this method has its objections also, the
main one being the different permeability of parts of unequal thick-
ness; the thicker parts, it might be assumed, are less under the influ-
ence of the surrounding medium than the thinner parts. In view of the
’ fact, however, that the heavy strips made from adult animals survive
just as long and manifest just as high a grade of rhythmicity as the
thinner strips from young animals, it would seem that this objection is
not a valid one.”

In conclusion we may repeat that present physiological conceptions
permit us to assume that any region possessing the function of deter-
mining the heart beat must possess the highest rate of rhythm. If
either the sinus node or the auricular node is to be considered the
cardiomotor centre, it must be shown that it beats decidedly more
rapidly than other parts of the heart. Experiments herein recorded
show that this is not the case. They show that the region of the inferior
cava and the region to the right of the mouths of the veins, neither of
which contains any of the tissue of the nodes as now delimited, possess
a degree of rhythmicity that is distinctly exceeded neither by the part
of the heart containing the sinus node nor the part containing the
auricular node.

Behavior of strips during stimulation.— The behavior of a strip while
it is being stimulated through and through, owing to circumstances
that are not altogether clear, is rather variable. The response to
tetanic stimulation of the beating strip is usually a fairly high initial
extra contraction, sometimes two or three, followed by small, rapid,
and irregular beats. Not infrequently the initial contraction is fol-
lowed by complete relaxation; it may, however, be that this quiescence
is only apparent and is actually due to responses so fine and rapid as to
be imperceptible. It would appear therefore that there is almost
every gradation of response to stimulation between large irregular
contractions and complete quiescence of the strip. Therefore there
seems to be no good reason for believing that the quiescence is the

2 Very recently Lewis (Heart, 1909-1910, i, p. 262) has called attention toa
slight difference in the time of appearance of the action current in different parts
of the auricle as determined with the string galvanometer. He is inclined to believe
that this difference is such as to indicate that the heart beat normally arises in
the vicinity of the superior cava.

See NY

A at 6

ETTTTTTTTIVSTETO Peder a VPP TET PTTTTPTP PTT TOOSVNOTENISTY OTT ANVATTTITTOTITIPVTOVONTC STOTT? TT

al size). — Showing the characteristic behavior of the sinus end (S) and the appendicular end (A) of a strip

when one or the other end is stimulated through and through. From Exp. 27.%

FicurE 3 (one half the origin

Joseph Erlanger.

result of stimulation of an inhibitory
mechanism.

In the case of strips so mounted that the
movements of the appendicular and sinus
ends may be recorded simultaneously,
stimulation of the sinus end through and
through while the whole strip is beating
results, as a rule, in complete cessation
of the sinus end, so far as visible beats are
concerned, the appendicular end usually
stopping also or showing a few small beats;
while stimulation of the appendicular end
through and through results in apparent
cessation of its beats, or in rapid irregular
beats which rarely are of considerable am-
plitude, more commonly exceedingly fine;
whereas the sinus almost invariably beats
rapidly and very irregularly (Fig. 3).

These rather variable results can be
brought into more or less harmonious ac-
cord if we assume, and apparently justifi-
ably, (1) that the apparent cessation of
beat during stimulation is not an inhibition
properly so called, but rather what might
be termed a complete tetanus in relaxation,
the intermediate stages, showing irregular
contractions, then being incomplete tetan1;
(2) that the more irritable (or better, per-
haps, the more rhythmical) tissue of the
sinus responds to such stimulation at a
more rapid rate and consequently falls into
a more complete tetanus than the appen-

13 GENERAL. —All records read from left to right.
Time in seconds. Where stimulations are not indi-
cated by the signal (uppermost line) they are in-
cated by irregularities of record. The lowering and
raising of the beaker to expose and cover the strips

are indicated by the elevation and depression of
the record respectively.

2 el

ee eee ee ee

Observations on Auricular Strips of the Cat’s Heart. to1

dix, and (3) that the sinus, being more irritable than the ap-
pendix, can respond more accurately to the slower impulses emerging
from the tetanized appendix than the more sluggish appendix can
respond to the rapidly recurring impulses sent out from the tetanized
sinus. Then it becomes clear why, during stimulation of the sinus,
while this part may or may not be quiescent, the appendix almost in-
variably is quiescent; whereas during stimulation of the appendix
the sinus almost invariably beats.

Sup.
syn! OSA
Ficure 4. — Showing that a stimulus (coil at 9) too weak to call forth irregularities de-

termines no subsequent improvement of beat. Two strips (superior cava and in-
ferior cava) mounted in series.. From Exp. 10.

The strength of stimules required to produce these effects jis rela-
tively great. As a rule, the rhythm of the beating strip is not
disturbed, nor are beats induced in the quiescent strip until the sec-
ondary coil has been carried to a point within 10 cm. of the pri-
mary. Frequently no effect is obtained until this distance is re-
duced to 7 or even 6 cm. The manner in which the terminals are
connected with the strip, as well as their location, affects the thres-
hold somewhat.

After effects of stimulation on amplitude. — The character of the beat
subsequent to stimulation is altered or a series of beats is started
only when the stimulus has been sufficiently strong to produce one or
the other of the disturbances in rhythm mentioned above (Fig. 4). In
that event the behavior of the strip depends, to repeat, upon its nature
and upon the point of application of the stimulus.

Strips composed of the left auricle only, or of any part thereof (ex-
cepting the interauricular septum), show no detectable after-effects.
They never beat excepting during stimulation.

Strips of the right auricular appendix only, with but a single excep-

ee ee EDL et TSA LST Teh Tats Tat

stimulation of the sinus region (1 in Fig. 2) and the appendicular end (3)
le arranged to record with separate levers the movements of the sinus end

yey a ee Fe

of a strip composed of the entire outer wall of the right auric

Ficurr 5 (about one half the original size). — Showing the effect of
(S) and of the appendicular end (A).

Joseph Erlanger.

tion with which there is some doubt
connected, have contracted only during
stimulation.

Strips composed of the vault only, either
right or left, likewise do not beat subse-
quent to stimulation. This statement, in
so far as it concerns the right vault, is
made with some reserve, since only two
experiments have been made on this part
of the auricles.

Stimulation of strips composed only of
septum or coronary sinus or atrium may
start a series of beats in the case of the
quiescent strip or may increase the rate and
to some extent the amplitude of an existing
beat.

Usually, however, in this investigation
strips have been so made as to include two
or more of the supraventricular parts; as
a rule, they include some part that would
upon stimulation yield a series of beats and
extend out from this to include parts not
spontaneously rhythmical, —to the right
or left appendices or to the vault, for
instance.

When such strips are stimulated through
and through, a series of beats usually re-
sults, or if the strip was previously beat-
ing, both the rate and amplitude of beat
are markedly increased (see Fig. 4). The
same result is obtained in certain cases
when only the spontaneously rhythmical
part is stimulated. Stimulation of the non-
rhythmical, e. g., appendicular, end only
at a time when no part of the strip is
beating, is without effect; in case, however,
the whole strip is beating the amplitude of
beat of the non-rhythmical part may be
markedly increased (Fig. 5); and if only
the sinus is beating stimulation of the

From Exp. 29.

Observations on Auricular Strips of the Cat’s Heart. 103

non-rhythmical end may result in the extension of the beat into the
latter.

It would seem therefore that under the conditions of our experiments
there exists in the rhythmical regions of the supraventricular parts a.
mechanism which, when stimulated, brings about an increase in rate
mainly, to a certain extent, however, an increase in amplitude of beat
of that part also; whereas stimulation of the non-rhythmical parts,
either directly or through the medium of the rhythmical parts, causes
only an increase in amplitude of beat or puts them in a position to re-
spond with beats to impulses previously impotent. In other words,
the rhythmical regions are provided with a chronotropic mechanism
mainly, the non-rhythmical with an inotropic almost exclusively.
There is obviously a division of Jabor and, it would appear, a division
very well adapted to the different work of the two parts.

The results obtained in this connection through local stimulation of
the outer wall of the right auricle, excised and suspended as described
on a previous page, confirm in every respect those just mentioned, and
demonstrate in addition that all parts of the rhythmical region upon
the one hand and all parts of the non-rhythmical region upon the
other respond to stimulation in practically the same way. The supra-
ventricular parts which contain the nodes behave the same in this re-
spect as the other rhythmical parts of the auricles.

The carrying power of sinus impulsés. — Attention was called above to
the fact that in the case of the compound strip stimulation of the
sinus, when it alone is beating, may cause the whole strip to beat, or
when the whole strip is beating, may greatly increase the amplitude
of the appendicular beat. By some, such results would be explained
by assuming that the stimulus has altered the conducting power of the
non-stimulated regions. We, however, believe that this apparently
dromotropic influence is due entirely to an altered carrying power of
the sinus impulse, not to any alteration in conductivity. The feebler
impulses, we assume, do not reach all of the muscle fibres, while the
stronger ones have a wider influence; hence the varying amplitude
of contraction. There is no need of assuming here any shortcoming of
the ‘‘all or none”’ law.

At times the impression is gained in experiments with local stimula-
tion that impulses emanating from certain parts of the sinus are par-

104 Joseph Erlanger.

ticularly effective in increasing the amplitude of the beat of the non-
rhythmical parts. This is strikingly seen in the case of impulses started
by local stimuli applied just below the mouth of the superior cava.
_ These seem to spread most readily into the appendix and to elicit there
the highest contractions. Such a result might be explained as being
due either to the relative strength of impulses emanating from, or to
the proximity of, this particular part of the sinus to the appendix.
That the latter interpretation is probably the correct one is indicated
by the fact that in an experiment in which stimulation of the superior
cava region was regularly inducing higher contractions of the appendix
than stimulation of the inferior cava region, a cut made so as to in-
crease the distance between the superior cava and the appendix, but
without injuring the superior cava region, at once did away with the
unequal influence (see Fig. 2).

At this place attention should be called to the fact that the two
points selécted for stimulation in this experiment were on the thick
tissue of the sinus region and were probably close to the gray line of
the venous valves. In our experience these places above all others have
responded to stimulation very constantly. In one experiment the only
part of a strip consisting of the entire outer wall of the right auricle
that could be seen to beat was the gray band described on page 91,
and which probably represents the remnant of the venous valves. The
beat in this case started in the vicinity of the superior cava and trav-
ersed the gray band to the inferior cava.

Description of typical series. — This is scarcely necessary after the
discussions and figures of the foregoing sections. Briefly, however, a
rhythmical strip may or may not begin to beat while in air after stim-
ulation. Should it beat then, it is usually with decreasing amplitude.
When the strip is immersed immediately after stimulation, the beats, as
a rule, are small at first and infrequent, but, presumably as the strip
takes on the temperature of the bath, they rapidly increase in ampli-
tude and rate, reaching the maximum usually within ten to twenty
seconds. As a rule, this is not maintained very long, but almost at
once amplitude and rate diminish, much more slowly, however, than
they increased, and reach the original height and frequency usually
within thirty to sixty seconds (see Figs. 6, 7). Sometimes the beat
remains improved over a very long period of time. This long-
continued improvement is seen most impressively in the case of

105

‘ul ‘7 {no ‘9 +p ‘dx Worq « ‘sUOTRINUINS AIssadINS YI JUoWaAoIduT ay Sumoys —‘(aZIS [VUISO oY} PITY] aUO ynoqr) 4 AANOTY
Kod Oe, Ca A Henur ! i!

I O I (0) I O I O
114, a SS ALAIAAAAADAAARAAAAAAAADAAROAAANDAAAAAAABAADDARNANANT GAN ATDNGONININIIUIII\UVINTII N\A

‘TT ‘dxq wor ‘Ajoyeredos poyunow sdiajs vaAvo JOMojur pur qouodng ‘ynuiys (¢ 7 [109)
Joyvoa UvYI (F Iv [109) Jo8u0.s Jo opnzydure pue o}e1 Uo Jago yuanbosqns J0}v915 oY} SuIMOYS — *(9ZIS [VUISIIO dy} J[Vy IU ynoqr) 9 ANNI a

N\A

‘dns

Observations on Auricular Strips of the Cat's Heart.

100 Joseph Erlanger.

strips which, after excision and suspension, fail to beat at all until
they have been stimulated, and then beat without further treatment,
it may be, as long as a half hour or more. Departures from these,
perhaps the normal, responses will be considered later.

Some conditions influencing the response of strips. — (1) The influence
of the strength of stimulation has been studied quantitatively only
for the purpose of determining threshold values. In this connection
it will be said merely that with the apparatus we have employed the
first improving effects, asa rule, were obtained when the secondary coil
was from to to g cm. from zero. This is clearly shown by the speci-
men record reproduced in Fig. 4. The strength of a sufficient stim-
ulus is not without effect, the stronger currents increasing to a certain
extent the amplitude as well as rate of beat (Fig. 6).

(2) Influence of duration of stimulation. —On account of the
cooling of strips exposed for stimulation quantitative results on the
subsequent influence of tetanic stimulations of varying durations have
not been obtained. It is, however, obvious that the length of the
subsequent series, but probably not the rate of subsequent beat, varies
in the same direction as the duration of stimulation. This is certainly
true for stimuli lasting as long as twenty-four seconds. Brief stimuli
of different durations, one to three seconds, also have an appreciable
effect.

(3) Influence of repeated stimulation. — Often, when a single stimu-
lation followed by immersion is without effect, repeated stimulation
under exactly the same conditions will eventually start a series. Then,
or at any time while the strip is not beating with maximum ampli-
tude, repeated stimulation up to the number of three, four, or even five
may successively increase the response (Fig. 7).

(4) Fatigue and recovery with rest. — Although it is possible in most
experiments to obtain series after series by means of stimulation, it is
found in some cases that sooner or later the strip fails. Then a rest
varying in duration from some minutes to many hours may so change
the condition of the strip that responses to stimulation equal to those
first obtained may again be elicited. It would seem, therefore, that un-
der the influence of stimulation the strip eventually fatigues and that
a long rest serves to remove these fatigue effects.

The longest time it has been found possible to keep a strip beating by
our method, allowing rests for recovery from fatigue effects, has been
nineteen hours.

Observations on Auricular Strips of the Cat’s Heart. 107

(5) Temperature.— The best responses to stimulation are ob-
tained at temperatures lying between 29° and 36° C. The maximum
temperature has not been determined. An active strip when gradually
cooled usually stops beating at 25.5° to 24.5° C. Upon warming, it
may begin to beat spontaneously at from 27.5° to 28.5°C. Strips
cooled to 26.5° C. have given beautiful series subsequent to stimula-
tion. A study of the exact relation between temperature and rate of
beat has not been included in this research.

(6) Is perfusion essential ?— As a preliminary to practically all of
our experiments the heart was perfused with Locke’s solution. This
was done for two reasons: first, because it was discovered while work-
ing with the perfused heart that strips could be made to beat; and,
secondly, by washing the heart tissue free of blood the fluid bathing
the strip remained clear, so that the behavior of the strip could be noted
with the eye as well as with recording instruments. It should be added,
however, that this preliminary perfusion does not seem to be essential
to the obtaining of series of beats, although the unperfused strips do not
seem to behave quite so well as the perfused.

(7) Influence of certain substances upon the strips. — A few experi-
ments were made for the purpose of studying the influence of certain
substances in the bath upon the behavior of the strips.

Oxygen, it was found, is not essential to the success of an experi-
ment, at least in amounts over those ordinarily held in solution by
freshly distilled water exposed to atmospheric air. In larger amounts,
however, such as can be added to a solution by bubbling the pure gas
through it, oxygen improves decidedly the beat of active strips.

A non-rhythmical strip, such, for example, as one made of the right
appendix, and which could not be made rhythmical by stimulation,
could not be made rhythmical either by substituting CO, for the O» or
by treating it with a 0.9 per cent solution of sodium chloride even when
stimulated tetanically. In pure sodium chloride solution such strips
slowly lose their irritability, in CO, saturated solution rapidly, but they
can usually be revived by reimmersion in pure Locke’s solution, pro-
vided they have not been exposed too long to the unfavorable condi-
tions. The irritability of rhythmical strips is affected by CO, and
sodium chloride in the same way as is that of the non-rhythmical, and
if they are beating the amplitude diminishes gradually to complete
disappearance. The rate of beat is not certainly altered by such
treatment.

108 Joseph Erlanger.

It is interesting to call attention here to the fact that although the
non-rhythmical terrapin’s ventricle “ and limulus heart muscle” can
be made to beat spontaneously by treatment with sodium chloride,
non-rhythmical mammalian auricular tissue cannot be made rhythmi-
cal in the same way.

(8) Effect upon strips of sudden elevation of temperature. — In the
vast majority of our experiments we have had to deal with two im-
portant factors influencing the response of strips. It will be recalled
that the strip is taken out of the bath to be stimulated: it is, in other
words, taken out of the warm bath and temporarily exposed to room
temperature. The question therefore suggests itself, how much, if
any, of the effect subsequent to stimulation is to be attributed to the
sudden elevation of temperature upon immersion and how much to
tetanic stimulation? The conditions of the experiment have been
varied in order to determine the relative values of these factors.

As a result, it has beenshown that immersion alone may or may not
exert some influence. Strips quickly prepared from a perfused heart
when immersed in the bath sometimes beat, usually feebly, very rarely
with considerable vigor, and, as a rule, such beats soon fail. When the
strip has stopped beating, or if it has not contracted at all, exposure to
room temperature and reimmersion has rarely if ever inaugurated a
series of beats (see Fig. 7). If the strip is beating spontaneously or has
been made to beat through tetanic stimulation, exposure to room
temperature followed by reimmersion but without stimulation may
improve the beat somewhat, never however as much as tetanic stimula-
tion followed by immersion (Fig. 8, a and 8).

It should here be noted that the beneficial influence of stimulation
is preserved a very long while in the unimmersed strip. If, for in-
stance, a feebly beating strip be stimulated tetanically immediately
after removal from the solution and then be kept exposed to the air,
it may or it may not suffer some improvement immediately after
stimulation, but when it is immersed a long while, it may be many
minutes, subsequent to stimulation, the amplitude and rate may be
increased by much more than from simple reimmersion.

Stimulation out of the bath without subsequent immersion may, as
has been noted above, increase the amplitude of subsequent beats, even

44 For literature, see MarTIN: This journal, 1904, xi, p. 103.
* Carson: This journal, 1908, xxi, p. 11.

Observations on Auricular Strips of the Cat’s Heart. 109

Ficure 8 (one half the original size). — Showing (a) slight improvement of beat consequent upon temporary exposure to room temperature

and (b) the much greater improvement consequent upon exposure plus stimulation. From Exp. 12.

when the strip is kept at room temperature,
but such conditions cannot be continued
any great length of time owing to the exces-
sive cooling of the strips.

In an attempt to eliminate this difficulty
strips were mounted and studied in a moist
chamber, but it was found that strips so
treated gradually shorten and at the same
time lose their irritability. Curiously
enough, this shortening of the strips is’
accelerated during tetanic stimulation.
When a strip that has lost its irritability
in a moist chamber is reimmersed in
Locke’s solution, its irritability returns,
and the beats eventually become as large
as those of a companion strip from the
same heart, but not exposed in a moist
chamber.

On the other hand, attempts to revive
the strip by stimulating it in the bath have
yielded unsatisfactory results. Despite
the fact that care was taken to insulate
the connections up to the points of attach-
ment to the muscle, so much of the current
was shunted by the solution that at best
only minimal stimulating effects could be
obtained.

It has therefore been impossible to de-
termine quantitatively the relative potency
of stimulation and of immersion with ele-
vation of temperature in the initiation of
a series of beats. It is evident, however,
that both are factors, and that of the two
tetanic stimulation is by far the more
potent.

The cause of improvement of beat.—It would
be of the very greatest interest if it could
be determined how tetanic stimulation
effects the improvement in rate and force

110 Joseph Erlanger.

of beat of auricular strips. We were inclined to believe at first that
the phenomenon was related in some way to the treppe process, that
the stimuli and resulting contractions, by acting upon an otherwise
completely or relatively inactive strip, remove resistance to conduc-
tion and heighten rhythmicity. A few experiments made for the
purpose of testing this hypothesis soon showed, however, that it
was untenable. Thus repeated induction shocks, although eliciting
contractions showing a beautiful treppe, have never resulted in
any improvement of subsequent beat, nor started a series, despite
‘the fact that the strength of the single shocks has in some of the
trials been very much greater than that of the tetanus.

On the other hand, the quiescence or apparent quiescence of
the strip that usually obtains during tetanic stimulation might lead
one to believe that both this quiescence and the subsequent im-
provement of reactivity of the heart tissue are affected in some way
through the same mechanism that stops the intact heart and sub-
sequently improves its beat when the peripheral end of the vagus
nerve is stimulated.

In view of this suggestion it was deemed advisable to determine what
effect atropin might have upon the response of strips to tetanic stimu-
lation. We give the protocol of such an experiment:

_ Time
in min,
fe) At this time when the horizontal inferior cava strip was reacting
constantly and beautifully to tetanic stimulation, coil at 9l4,
and while the strip was beating, enough atropin was added to
the Locke’s solution bathing the strip to make it a 0.5 per cent
solution.
4 After making five contractions the strip stopped beating.
6 Not beating. Stimulation, coil at 914, gives initial contraction

only. Stimulation, coil at 914, gives initial contraction and is
followed by a series of slow irregular beats in Si: A block.
14 No further improvement with stimulation coil at 9, at 8%, at 8,

nor at 714.
22 Strip immersed in pure Locke’s solution (free of atropin). |
ie Stimulation with coil at 714 gives decided improvement of beat —

through several successive stimulations. Later, improvement —
was obtained with coil at 814, but not at 914. The beats, how- ©
ever, were never as good as those obtained before atropin was —
administered.

I Pa ont ll 0d

Observations on Auricular Strips of the Cat's Heart. 111

098 Strip was immersed in ancther beaker of fresh Locke’s solution and
again, upon stimulation with the coil at 8, a better beat was ob-
tained, but the beats obtained were at no time as fine as the
splendid series of the pre-atropin stage of the experiment.

This experiment, as well as others of the same kind, demonstrates
that in the case of strips prepared by our method atropin seems to
have no other action than to diminish the ability of the strip to re-
spond with a series of beats after tetanic stimulation. It is possible, al-
though difficult to demonstrate clearly, that the strip treated with
atropin is more apt to respond with contractions while being stimulated
tetanically than the strip not so treated, But even if this could be
demonstrated beyond peradventure it would be quite as justifiable to
conclude from it that it is due to decreased irritability through atropin
action as to the setting aside of an inhibitory mechanism. We have
therefore reached no conclusion as to the way in which tetanic stimu-
lation acts to initiate a series of beats and to improve both the rate
and amplitude of beating strips.

It should be added here that the galvanic current, within the
strengths that have been used (2 volts, 0.75 milleamp.) have been
without appreciable after-influence upon the strips.

Irregular types of response. — The after-effect of stimulation of auto-
matically rhythmical parts of a strip usually consists of a perfectly
regular series of beats such as has been described above. Not infre-
quently, however, and especially when a small part rather than the
whole of the strip is stimulated, atypical series of beats are obtained,
Such atypical responses for the sake of convenience are considered
under two heads.

(1) The first of these groups is characterized by a sudden change
in rate of beat, consisting usually of (a) a decrease, rarely of (b) an
increase.

(a) In the case of the former the series shows the usual initial in-
crease in amplitude and rate. When the beats have attained their
maximum or have passed it and are beginning to decrease, there oc-
curs an abrupt slowing of rate. This new rate may be constant from
the moment of its onset (Figs. 5 and 9); more commonly, however, the
slow rate is developed more or less gradually out of complete stoppage
. of variable duration (see Fig. 10).

(0) The sudden increase in rate, which has been seen only occasion-

MAMAN VAUGHN 14hUt mn Pen Men eg

Ficure 9 (one half the original size). — Showing some of the irregular types of series commonly obtained, including (A) sudden decrease

From Exp. 17.

in rate and (B) sudden increase in rate, in this case with alternating beat.

Joseph Erlanger.

ally, usually appears while the amplitude of beat
is increasing subsequently to stimulation and per-
sists a variable, though usually a short, time, when
there occurs an abrupt return to a slow rate which
is approximately the same as that recorded prior
to the sudden increase (see Fig. 9).

We havenever succeeded in discovering the mech-
anism of the above-mentioned variations from the
usual response. Their resemblance to pictures
often seen in experiments on heart block is very
striking. Thus the sudden stoppage with subse-
quent development of a slow rate is identical with
the result of suddenly and completely severing the
functional connection between a less rhythmical
and a more rhythmical part of the heart. And the
sudden acceleration with equally sudden return to
the former slow rate resembles what is seen when
a partial block suddenly disappears and then re-
appears. Still, with certain exceptions to be men-
tioned below, we have never succeeded in obtaining
with the eye or with recording instruments any evi-
dence of block at these times. It should be added
that in those instances in which there is no stage
of development associated with the change in rate,
the two rates do not bear to each other the exact
aliquot relation that obtains in changing degrees of
partial heart block.

The sudden decrease in rate, sometimes preceded
by stoppage and gradual development of a new
rate, is satisfactorily accounted for, amongst other
ways, upon the assumption that the part of the
strip that has had its rhythmicity increased by
stimulation and which for that reason has been
setting the pace of the strip, suddenly loses its
rhythmicity almost completely, or in some way.
loses its influence over the rest of the strip, and
that then the next most rhythmical part assumes
the function of pacemaker. This it may do with rate

Observations on Auricular Strips of the Cat’s Heart. 113

of beat fully developed at the outset or, owing to the fact that its
rhythmicity has been dominated for a while by another part, it may
gradually acquire its usual rate, much in the same way as happens
when the ventricles are suddenly liberated from the influence of their
normal pacemaker, the auricles, or from an artificial pacemaker, for
example, single induction shocks.”®

i ee rates Se SE sel et es

ne ei Ne LIAR Ro DAMEN SREY UES Oe Cnn

FricureE 10 (two fifths the original size). — Showing the development of a slow rate of
beat out of stoppage subsequent to a series of rapid beats induced by local stimulation.

It is more difficult to account for the sudden increase in rate. Pos-
sibly, however, the stimulus in these cases starts beating rapidly a
limited area, the impulses from which fail for some unknown reason
to reach the rest of the strip. When they do become potent, the rate
of beat of the whole strip is suddenly increased.

These changes in rate resemble those which are commonly observed
in cases of so-called paroxysmal tachycardia, and I have no doubt, as
has very recently been suggested by Lewis ”’ and others, but that cer-
tain of the paroxysms in man are started in much the same way as in
the case of our strips: they are determined by some stimulus acting
either directly or possibly indirectly upon some part of the spontane-
ously rhythmical region of the heart. Our experiments would seem
to indicate that the part acted upon need not be the so-called nodal
regions, as has been assumed by some authors, but any part of the
right auricle inclusive of the septum, but possibly exclusive of the right
appendix and vault.!®

16 ERLANGER and HIRSCHFELDER: This journal, 1906, xv, p. 153.

17 Lewis: Heart, 1909-10910, i, p. 262.

18 Lewis (Loc. cit.) maintains, mainly as a result of his observations on a case
of paroxysmal tachycardia in which the paroxysms were preceded by ectopic
auricular extrasystoles, that “there is no essential distinction between extrasys-
tolic and paroxysmal beats.’”’ If I understand Lewis correctly, I do not believe
his statement is justifiable. I am inclined to believe that there is a distinct dif-
ference between occasional extrasystoles, paroxysmal acceleration, and fibrillation

II4 Joseph Erlanger.

In this connection a case of paroxysmal tachycardia recently de-
scribed by Lewis ’* is of some interest in that there seems to be every
reason for believing that the auricle in this case behaved quite like some
of our strips. Not alone were there the paroxysms of tachycardia, but
in addition these paroxysms ended abruptly and were succeeded by a
pause before the heart resumed its normal slow rate. .

(2) In the second category of atypical series may be included
all of the variations which without doubt are due to the blocking of
impulses.

The types of block and the consequent departures from the typical
serles of beats have been very varied. Some of the blocks undoubt-
edly were entirely artificial and due to injury inflicted while preparing
the strip. There is, however, every reason for believing that most
were natural and were due either (a) to the natural difficulty the
impulse experiences in traversing heart tissue or (0) to the existence of
natural blocking points possibly at the junctions of. developmentally
different segments of the heart.

(a) The former variety of block is frequently seen in strips begin-
ning to beat subsequently to immersion in the bath per se, or subse-
quently to stimulation and immersion. It is then seen that only the
venous end of the strip begins to beat. Gradually, however, and with-
out further treatment of the strip, or as a result of repeated stimulation,
the contraction wave penetrates further and further until the strip beats
throughout its entire length. This phenomenon is probably identical
with that observed when strips of terrapin’s ventricle, in which there
is no reason for believing there are any natural blocking points, are

in that these probably represent three distinct results of irritation of the heart.
The relation between them may be represented as follows: When a constant but
weak stimulus acts upon a heart whose irritability is gradually increasing, or when
a gradually increasing stimulus acts upon a heart with constant irritability, the first
response may be occasional extrasystoles without any alteration of rhythmicity.
Then the stimulus may alter for a time the rhythmicity of the heart tissue and
thus cause the paroxysmal accelerations. Eventually fibrillation, the nature of
which is not understood, may ensue. For this a very strong stimulus or a highly
irritable tissue is necessary. In the present research we have not met with fibril-
lation, probably for the reason that under the conditions of our experiments the
strips possess a very low grade of irritability. If these three conditions are the
same and extrasystolic in nature, the proof that such is the case is still to be
furnished. :
” LEwIs: Heart, 19009, i, p. 43.

Observations on Auricular Strips of the Cat’s Heart. 115

taught to beat by applying stimuli to one end or by immersion in
certain salt solutions. The blocks in this case are almost invari-
ably complete and pass away gradually by the progressive increase in
the extent of the part of the strip that beats.

(b) The second variety of block, namely, that occurring at what
may perhaps be considered natural blocking points, may be either
partial or complete. The records that have been obtained of these
blocks leave no doubt as to the correctness of our interpretation
(see Figs. 11 and 12). The blocks appear and disappear abruptly,
and therefore determine a very abrupt and unmistakable change in
the appearance of the record. As to the cause of these changes in the
degree of block, it would appear that almost invariably they are refer-
able to changes in the strength of the impulse; only rarely they can be
explained upon the basis of altered irritability. Thus, in case the con-
ditions for the occurrence of block are present, it may come on while
the strength of the impulse is presumably decreasing and disappear
when the strength of the impulse is presumably increasing. A glance
at the figures (Figs. 11 and 12) will make clear the basis for this state-
ment, in that it is seen that the block disappears while the amplitude
of beat is increasing subsequently to stimulation and reappears as the
amplitude of the beat diminishes.

On the other hand, a block between the pacemaker and its depen-
dent part may disappear when only the dependent part is stimulated,
and its irritability, presumably, thereby increased.

More rarely a block may appear suddenly and without apparent
cause in the midst of an otherwise perfectly typical series of beats.

Position of blocking point.— Every effort has been made to deter-
mine macroscopically the exact position of the blocks that may be
termed natural. For the sake of safety we have used for this purpose
only those instances in which the block was partial, for the reason
that localization is difficult in the case of complete block.

The vast majority of such blocks were seen in horizontal inferior
cava strips, and in these strips they appeared to be at or near the junc-
tion of the parts we have designated sinus and atrium, perhaps a bit
nearer than this to the venous end of the heart.

Blocks have also been seen some distance to the left of the superior
cava in horizontal superior cava strips, but upon the whole it would
seem that the impulse is conducted more readily from the region of the
superior cava than from that of the inferior cava.

Joseph Erlanger.

116

‘OT “dxq worg ‘apqrdeosed Ajareq are SUOL}OVI}UOD SNUIS 0} ONP J9AV] 9} JO SJUSWMMOAOU OY, “UINII7B pu SNUIs
uoamjoq Ajqeumsaid yoorq jenszed pur ojojduroo Surmoys 103}¥] 949 ‘sdizys BAB JoLayur pus royiadng — *(azIs [eULSIIO ay} SYIJ[IM} WIAes) 7] TANS]

aur

*, -dxq worg ‘o[quydeosed Ajoreq ore suor}9e1}UOD SNUIS 0} ONp J9Ag] ay} JO SJUSMMIAOTT oT, “INIT}e
pure snuis useajoq Ajqeumsoid yooiq [erjsed jo sapesd [ersAes puv o}0[dut09 Surmoys dis vavo Jopoyay — *(azIs [Bul

THT

Z110 oy} SYIYSIO UsAes) TT TANS

ict A one 0

( aL a I

Observations on Auricular Strips of the Cat’s Heart. 117

Admitting, for purposes of discussion, the existence of a natural
blocking point along the whole length of the sino-atrial junction, this
result may be accounted for by the difference in the thickness and ar-
rangement of the tissue surrounding the superior and inferior cava
regions. About the former it is thick, while about the latter it is thin-
ner and trabeculated, the tissue between the trabeculze being espe-
cially thin. |

Spontaneous blocks have been observed, though much more rarely,
between the inferior cava region and the left auricle, between the
superior cava and the right vault, and between the auricle and the
right appendix. A more exact localization of these positions could
not be made on account of the absence of clear landmarks.

That the blocks we have seen may have been the result of mechan-
ical injuries cannot of course be positively precluded. There seems
to be little reason for believing, however, that such was the case, while
the frequent appearance, in different specimens, of block at about the
same place would bespeak nothing but a natural resistance to the
passage of the impulse at that place.

Viewed from the practical standpoint, these results serve to indicate
that in the normal heart there is located somewhat to the left of the
opening of the cave (especially of the inferior cava) into the auricle
a place of relatively high resistance to the passage of the impulse,
but that evidences of this resistance can be brought to light only when
the conductivity of the auricular tissue has been greatly reduced by
subjection to unfavorable conditions, such as have obtained in our
experiments, and by narrowing the conducting tissues. Whether
there are other such regions of low conductivity cannot be positively
affirmed until we shall have the results of a larger number of experi-
ments at our disposal.

There is hardly any need of adding that a more exact localization of
these ‘‘natural’’ blocking points macroscopically but especially micro-
scopically would be of the greatest interest, and should be attempted.
It will not, however, be an easy investigation owing to the fact that
the blocks cannot be produced at will, and even when obtained they
are usually evanescent and in places rendered inaccessible by appa-
ratus or folds of tissue.

118 Joseph Erlanger.

SUMMARY.

An arbitrary division of the cat’s heart is made into sinus, coronary,
septal, atrial, appendicular (right and left), vaultal (right and left), and
pulmonary regions. When treated according to the method herein
described, strips composed wholly or in part of the sinus, septal, cor-
onary, and (or) atrial regions can be made rhythmical, and the beats
have been maintained with intervals as long as nineteen hours. Under
the same conditions the rate of beat developed in any part of these
four regions is practically the same. This property as well as that of
rhythmicity in general is therefore not dependent upon the pres-
ence of nodal tissue. Stimulation of the rhythmical regions in the
main increases the rate of beat and the strength of impulse, whereas
stimulation of the non-rhythmical regions increases the force of their
beat, which must of course be determined from without, or renders
them capable of responding to impulses from without.

The behavior of strips during stimulation and the subsequent re-
sponse of rhythmical strips with regular series of beats and irregular
series of beats, the latter including sudden accelerations and retarda-
tions resembling those seen in certain cases of paroxysmal tachycardia,
and blocks of various kinds and at various places, but especially at or
about the sino-atrial junction, are described and discussed.

Sete

A COMPARISON OF THE TOTAL NITROGEN EXCRETION
OF EITHER KIDNEY IN NORMAL INDIVIDUALS
DURING VARYING PERIODS OF TIME.

By THEODORE B. BARRINGER, Jr., AnD BENJ. S. BARRINGER.

[From the Laboratory of Clinical Pathology, Cornell University Medical College, New York.]

N estimating the functional capacity of a diseased kidney more or
less attention has been given of late to a comparison of the nitro-
gen excretion of the two kidneys, chiefly as shown by the urea output.

If the ability of a kidney to excrete the end products of protein
katabolism is to be made a basis for the estimation of that kidney’s
functional ability, it would seem more rational to determine the ex-
cretion of the sum of the end products as shown by the total nitrogen,
rather than by any one constituent as urea, which Folin has shown
may form between 61 and 87 per cent of the total nitrogen.

As a basis for the urea comparison, Albarran! has reported his
findings in a series of normal persons. Although on a priori grounds
the total nitrogen excretion of either kidney might be expected to
show the same variations as those of its most important constituent,
urea, only isolated determinations have yet been made in normal
‘cases as far as we can-ascertain.

Our work was carried out on eleven young men, all in good health.
In collecting the urine the method of Albarran was employed.

The flute-tipped ureteral catheter, with two lateral openings, was
used. As large a catheter as possible — in a majority of cases 7 F. —
was introduced to a depth of 15 cm. into the ureter. In most of the
cases one ureter was catheterized, a bladder catheter collecting the
urine from the other kidney. Before the collection of urines 1.5 c.c.
of indigo carmine was injected into the ureteral catheter to determine
if extra-catheter flow were present. No cases were included in the
series in which blue appeared from the bladder catheter after this

1 ALBARRAN: Exploration des fonctions renals, 1905, p. 329.
119

120 Theodore B. Barringer, Jr., and Benj. S. Barringer.

injection. In one case the ureteral catheterism was bi-lateral, in
which a bladder catheter controlled the question of extra-cather flow.
In one case the urines were obtained by the Luys Separator.

As arule, the urines for the first half hour following instrumentation
were not collected. If, however, the urine began to run freely from
each catheter after instrumentation, and there were no signs of reflex
oliguria, the collection was begun immediately.

The catheterization occurred at intervals of from one to seven
hours after eating, and the urine was collected for periods of time
varying between twenty minutes and two hours.

In the following table our results are shown in detail:

TABLE I.

NorMAL CASES.

RicHtT KIDNEY, LrEerr Kiney.

Urea-N Urea-N
ieee ap Quantity. Rinse
S* | Amm.-N. gen. | Amm.-N.

Quantity.

c.c, c.c.

27.5 291 .205 32. .278
vf Bs 368 .264 2.3 36
Py) 20.

097 12.

.08

7.5

As regards the quantities secreted by each kidney during the same
length of time: Once they were equal. Six times they varied by less
than 10 per cent. Four times they varied by between 10 and 20
per cent,

Total Nitrogen Excretion of Either Kidney. 121

Albarran found in twenty-six examinations that: Twice the quanti-
ties were equal. Ten times they varied by less than ro per cent. Four-
teen times they varied by between ro and 30 per cent.

As regards the total nitrogen.: In one case the quantities were equal.
In seven cases they varied by less than 1 gm. per litre. In two cases
they varied by between 1 and 2 gm. per litre.

The nitrogen-urea plus ammonia-urea showed in three cases a varia-
tion of less than 1 gm. per litre and in six cases a variation of between
1 and 2 gm.

Albarran found in a series of thirty-nine cases that the wrea itself
differed by less than 1 gm. per litre in twenty-nine cases. In ten cases
there was a difference of between 1 and 2 gm. per litre. Obviously
no basis exists for a comparison between these results and our find-
ings of the excretions of urea-nitrogen plus ammonia-nitrogen.

SOME OBSERVATIONS ON THE PRODUCTION OF
LIGHT BY THE FIREFLY.

By JOSEPH H. KASTLE anp F. ALEX. McDERMOTT.

INTRODUCTION.

i connection with certain investigations on the oxidizing ferments

and biological oxidations, the attention of one of us (Kastle) was
attracted to the subject of light production by the common firefly
as early as 1901. It was found that aqueous extracts of the luminous
organ of the firefly gave no color with fresh tincture of guaiacum.
With guaiacum containing small amounts of hydrogen peroxide,
however, a faint blue coloration was obtained, and, as is the case
with such a great variety of living tissues, the hydrogen peroxide was
decomposed. In other words, these preliminary observations pointed
to the absence of an oxidase in the luminous organ of this insect, and
to the presence of small amounts of peroxidase and catalase. During
the summer of 1909 the opportunity presented itself for making some
additional observations on the production of light by this insect.
For the most part these were made before we had had the opportunity
to thoroughly familiarize ourselves with the extensive literature of
the subject,* and hence include much that had already been done
before. We have been able, however, to add a little here and there
to the great wealth of observations on this interesting phenomenon,
and to confirm much of the older work on the subject. In what
follows we propose to consider the production of light by the luminous

* During the past year one of us (McD.) has attempted to compile a complete
bibliography of the literature of this subject, and already over eight hundred
references have been obtained, dating back in some instances to the times of
Aristotle and Pliny the younger. Obviously anything like a complete survey
of this extensive literature is beyond the scope of the present communication, and
hence only such references have been made use of as seemed to us necessary
for an understanding of our present knowledge of the subject.

122

Observations on the Production of Light by the Firefly. 123

organs of the common firefly (Photinus pyralis Linn.), under the
influence of mechanical, physical, and chemical stimuli.

MECHANICAL AND PHYSICAL STIMULI.

Mechanical stimulation. — It has long been known that the produc-
tion of light by living things is greatly influenced by mechanical and
physical stimuli. Thus Spallanzani* observed that light production
by the Lampyride js greatly intensified by scratching or stabbing the
luminous organ with a needle. KdOlliker™ also found that light is
emitted by the luminous organ if it be divided or crushed, and also if
it is pulled to pieces or subjected to slight pressure. Similarly Fara-
day ” found the luminous material of the glow-worm to glow actively
on being pressed with a knife. Humboldt ?° also observed that a
luminous medusa which had been bound to a tin plate by a strip of
metal actively lightened as the result of any motion or disturbance
of the plate. Macartney * also observed the luminous meduse to
emit light when the water containing them was subjected to agita-
tion. As the result of his studies on this subject, Heinemann ” finds
mechanical stimuli to be the most active of all stimuli to the pro-
duction of light in living forms. Artaud? also showed that luminifer-
ous sea water at rest lightens again when it is disturbed. That such
is the case is evident from the production of light in tropical seas,
following in the wake of a boat or other object moving through the
water. Peters,*! from his work on the phosphorescent Ctenophores,
concluded that the combined action of darkness and agitation was
one condition which would result in light production by these organ-
isms, but that neither alone could be relied on as a positive stimulus
to the photogenic function. On the other hand, Kolliker * found, in
the case of the higher phosphorescent animals, that the normal move-
ments of the animals themselves are apparently without influence on
the production of light. According to Agassiz*® phosphorescent
medusz are sensitive to shock, and according to Todd *° all mechanical
and chemical stimuli which ordinarily produce pain cause the lumi-
nous organs of the firefly to lighten. Pfliiger ** observed that when
one cuts off the head of a Lampyris the light is extinguished, but
that when after a time the motions of the rump begin again, the light
reappears, although weaker than at first.

124 Joseph H. Kastle and F. Alex. McDermott.

According to Tilesius #8 and Meyen”’ repeated stimulation exhausts
the light-producing power of photogenic organisms, and Macaire *
found that sudden noises have the power to darken the light of the
firefly.

The following observations of our own on the effect of mechanical
and physical stimuli on the production of light by the detached organs
of the common firefly (Photinus pyralis Linn.) are also of interest in
this connection. The luminous organs were removed from the insects
by means of a delicate pair of scissors. As a general thing, these
luminous organs at once burst into glow as the result of removal from
the insect. In some cases this glow spread uniformly over the whole
surface of the luminous organ; in other instances it was confined to
those portions of the luminous organ nearest to the line of the cut.
If allowed to remain undisturbed, these detached luminous organs
gradually ceased to glow and remained quiescent practically indefi-
nitely. They were found to retain their power to glow actively, how-
ever, for several hours after removal from the insect, in spite of the
fact that they had dried out considerably in this time, and in one
instance a luminous organ which had been preserved in a I per cent
trikresol solution was found to glow strongly eight hours after removal
from the insect. Squeezing between the thumb and finger, a sudden
blow with a match-stick, tapping with the head of a pin, or pricking
or scratching with a needle or pin, dropping through a distance of
three or four feet to the floor, and plunging into ice water were all
observed to cause these quiescent, dark, luminous organs to burst
actively into a glow.

As a matter of fact, it has been found that percussion furnishes us
with a simple test for judging of the photogenic activity of the lumi-
nous organ of the firefly. Thus we have observed repeatedly that the
moist luminous organ of the firefly which does not emit light on per-
cussion cannot be made to lighten by any other means.

Temperature. — The production of physiologic light is also greatly
influenced by temperature. Thus Macaire* observed the light of
Lampyride to be extinguished above 52° C. and under 12° C., and
that heat which is far below that of boiling water destroys the light-
producing material of the firefly. He also states that the light of the
Lampyride is extinguished at 10° R. (12.5°C.). At 22° R. (27.5° C.)
he found these animals to lighten again. At 33° R. (41.3° C.) the light

Observations on the Production of Light by the Firefly. 125

was strongest, and at 46° R. (57.5° C.) it was extinguished. He also
found it to be impossible to restore the power to lighten in those in-
sects which had been heated to 47°-50° R. (58°-63° C.) He found
further that the free luminous organs of the Lampyridz glow most
actively at 33° R. (41.3° C.); at 42° R. (52.5° C.) the light was ex-
tinguished and the organ itself took on the appearance of cooked egg.
This author also observed that heat cannot cause the emission of
light in the luminous organ which has been kept im vacuo, but that it
lightens as soon as air is admitted. Jousset de Bellesme™ also found
that increasing warmth stimulates the Lampyride to lighten. This
author, however, considers this to be an indirect phenomenon, due to
the fact that heat excites the animal. He also observed that if the
animal be killed by exposure to a high temperature, the light is ex-
tinguished forever. Bongardt observed that up to 4o° C. the light of
fresh organs of the firefly increased somewhat with the rise of tempera-
ture, whereas above this temperature it became weaker, and ceased
at 68° C. This author also observed that below 23° C. these animals
do not lighten. At 48° C. also they seem to be dead, but still lighten
again occasionally; at 59° C. their power to lighten is permanently
destroyed. Kdlliker * also investigated the effect of temperature on
this phenomenon. According to this author, heating the luminous
organs to 40° to 60° R. (50° to 75° C.) produces a constant bright
light. He also found that cooling to —3° to —5° R. (—3.75° to
—6.25° C.) caused the organ to lighten, but only rarely and not
so certainly. He also observed that sudden changes of tempera-
ture, such as that caused by removing the insect from a piece of
ice to the hand, nearly always react upon the organism, causing it
to lighten.

According to Pfliiger,** these temperature effects are character-
istic of all light-producing material of animal origin. Thus with
molluscs (Pholaden) light is produced by moderate warmth and is
destroyed by boiling. This author observed that the light emitted
by putrid fish disappears at o° C., but reappears on warming and is
permanently extinguished by boiling. Artaud?! also observed that
luminous sea water lightens best at 43° C., and according to Michae-
lis 28 the water of the Ost See cannot be heated above 24° R. (30° C.)
without extinguishing its luminosity. Peters *! observed that the
phenomenon of phosphorescence in Ctenophores was produced at

126 ©Joseph H. Kastle and tb. Alex. McDermott.

temperatures between 9° and 37° C., the optimum temperature being
about 21.5° C., the temperature of sea water.

Electricity. —- Similarly, most observers are agreed that light is
produced by the luminous material of photogenic life forms as the
result of electrical stimulation. Thus Macaire” found the Lampyris
(Luciola italica) to emit an intense light as the result of electrical stim-
ulation. He also observed that the clear, translucent light organs of
this insect emit light under the influence of the electric current as long
as the current is passing, but that dead, opaque organs no longer re-
spond to the electrical stimulus. So also he observed that the electric
current is powerless to produce light in living organs which are kept
in vacuo, but that on the admission of air the current at once causes
such organs to glow. Jousset de Bellesme™ also found the electric
current to be an exciter of phosphorescence in the Lampyride, but
whether a primary excitation or a secondary one due to a general
irritation of the animal could not be determined. KéGlliker* also
found that electrical stimulation of the light organ of the firefly re-
sulted in the production of an intense light. Pfaff * observed that
phosphorescent sea water appears full of light when an electric current
is passing through it, the light apparently emanating from minute
particles in active motion. Heinemann ’° has also devoted consider-
able attention to the study of the electrical stimulus on phosphorescent
animals, and Pfliiger ** has called attention to the fact that neither
the contraction of the muscles of lower animals nor the emission of
light by photogenic life forms is brought about by the electric spark
or shock, but only by the passage of a continuous current.

Vacuum. —- The effect of the vacuum on the production of light
by photogenic life forms has also been investigated by different
observers with different results. Thus Carradori* observed that the
Italian luciole continues to shine in a barometric vacuum. On the
other hand, Pfliiger ** cites Heinrichs 1” to the effect that the light
emitted by living things disappears im vacuo to reappear again on ex-
posure to air. Similar results were obtained by Macaire* with
Lampyris (Phausis) splendidula. He observed that zm vacuo neither
heat nor the electric current could cause the production of light in
the luminous organ of this species, but that they glowed at once on
the admission of air. So also it is said that the light of decaying fish
is extinguished in a vacuum, and Dubois *:° observed the lighting
of Pholas and Pyrophorus to be suspended therein.

Observations on the Production of Light by the Firefly. 127

Our observations on the effect of drying zm vacuo on the photogenic
process are also of interest. A number of live fireflies were placed in a
large petri dish, in a vacuum desiccator, over sulphuric acid. As
soon as the desiccator was connected with the vacuum pump
(giving a vacuum of 27.5 inches) the insects began to show signs of
distress, as manifested by greatly increased activity. At the same
time they began gradually to elongate from head to tail, to such an
extent that the luminous organ protruded almost or quite its entire
length beyond the elytra, which ordinarily cover it over completely
on the dorsal side. At the same time the insects began to lighten
vigorously and rapidly, and in a short time the entire luminous organ
of each insect was glowing brightly and continuously, whereas ordi-
narily, as is well known, light production by this species is an inter-
mittent phenomenon. The detached luminous organs of the insect
showed essentially the same conduct so far as the production of light
im vacuo is concerned. On exposing a number of them to the vacuum
they began glowing brightly and continuously, one after another,
until all of them were in a state of glow, which lasted an hour or longer.
In this connection it is of interest to note that all of the luminous
material is not consumed on drying im vacuo, since on moistening the
dry material with water it again emits light.

The production of light by photogenic life forms is also brought
about by the action of chemical stimuli. In the literature accessible
to us we have been able to find references to the action of no less than
sixty-eight different substances. These include various gases and
vapors; also liquids and solutions of acids, alkalies, salts, and alkaloids.

CHEMICAL STIMULI.

In order to test the conduct of the luminous organ of the firefly
towards various gases the following mode of procedure was adopted:
The luminous organs of the insect were removed by means of scissors,
and kept upon a watch-glass until they had become quiescent and had
ceased to glow. They were then placed in short glass tubes open at
both ends, about 2 cm. in length by o.5 cm. in diameter, and stopped
loosely at both ends by means of glass wool. The receptacle for the
- gas consisted of a test tube with a gas inlet tube sealed into the bottom.
The bottom of the tube was loosely stopped with glass wool, and the

128 Joseph H. Kastle and F. Alex. McDermoit.

upper end of the test tube was closed with a stopper carrying an open
capillary tube. The gas whose effect was to be observed was passed
in at the bottom of the tube and allowed to escape at the top. When
the tube was filled with the gas, the rubber stopper carrying the capil-
lary tube was removed as quickly as possible, and the short glass
tube containing the luminous organ of the insect was dropped into
the larger tube containing the gas, and the stopper again inserted.
In this way the conduct of the luminous organ in a current of any
given gas or in a quiet atmosphere of the gas could be observed at
will.

In studying the effect of vapors such as chloroform, ether, etc., a
small amount of the volatile substance was placed in a small test tube.
A loose plug of glass wool was then inserted above, but not touching
the liquid or substance, and the test tube closed with a cork. After
the tubes had stood for a sufficient length of time for the vapor of the
substance to diffuse into the upper part of the tube, the small glass
tube containing the luminous organ of the firefly was introduced, and
the test tube again closed with the cork. For hydrofluoric acid a
lead tube was employed, and the luminous organ was suspended in
the gas in a little basket made of platinum wire, and the conduct of
the luminous organ observed by looking down into the tube from
above. In testing the effect of solutions of acids, salts, alkaloids, etc.,
the solution was either injected into the live insect by means of a
very fine hypodermic needle, or the cut surface of the excised luminous
organ was brought in contact with the given solution on a test plate.

The following are the principal points of interest regarding the
action of chemical stimuli on the biophotogenic process.

Air. — The production of light by living things takes place in at-
mospheric air. Indeed according to many observers oxygen is essential
to its production, as manifest by the abundant supply of trachea in
the luminous organs of insects, and by the fact that the light is in-
tensified by an increase in the amount of oxygen supplied the tissue.
Faraday ” observed the luminous material of the glow-worm and
firefly to glow on exposure to air, such exposure always causing a
fresh emanation of light. According to Milne-Edwards,”? physiologic
light is produced only under the influence of oxygen. Dubois ® found
even the dry luminous organs to glow in air at a pressure of 600
atmospheres. On the other hand, according to Bongardt,? a stream

Observations on the Production of Light by the Firefly. 129

of air or a current of an indifferent gas seems to retard the lighting
process. He concludes that in hydrogen or carbon dioxide it is not
the indifferent gas itself which causes the light to diminish, but the
current of gas. Still other references to this subject are given under
oxygen. It has also been our experience that atmospheric air at
room temperature is no direct stimulus to light production by the
luminous organ of the firefly, unless the organ has been previously
immersed in an atmosphere of an indifferent gas such as nitrogen,
or some reducing gas such as hydrogen sulphide. Thus luminous
organs of the firefly were frequently kept in the air for an hour
or longer under the conditions already described on page 127,
without showing any luminosity, and yet at the end of this time the
organ glowed brightly on percussion. On the other hand, a luminous
organ which had ceased to glow in hydrogen sulphide glowed strongly
on being brought out into the air, and this phenomenon, namely,
alternate extinction of the light in hydrogen sulphide and glowing in.
the air, could be repeated several times on the same luminous organ.
Oxygen. — According to Bischoff,? oxygen is absorbed and carbon
dioxide evolved in the glowing of certain luminiferous rhizomorphs.
According to Macartney,” Spallanzani** observed that in oxygen
glow-worms shine more brilliantly than in air. Forster found the
light of certain living things to be more intense in oxygen than in
common air. Jousset de Bellesme”! observed the Lampyridz to lighten
intensely in oxygen. Dubois !° found that Pyrophorus conducts itself
in oxygen just as in air, only the intensity of the light seemed to be
increased. According to Pfliiger,** the lighting of dead fish first takes
place with the absorption of oxygen and the evolution of carbon dioxide.
He also states in another connection that the lighting of dead fish, etc.,
takes place only in respirable gases. This author states that dead fish
glow in water containing oxygen, but not in freshly boiled water;
they glow, however, in the latter medium also on the admission of
air. Watasé*! claims to have proved that the lighting of the firefly
is directly due to oxidation ; on crushing the luminous organs on a
glass slide and immersing in carbon dioxide, the light was extinguished;
when the slide was removed from the carbon dioxide and placed in
oxygen, the light reappeared. According to this author, this phenom-
enon could be repeated several times in succession on the same speci-
men. Miss Townsend *’ has also found that the photogenic tissue of

130 06©6-: Joseph _H. Kastle and F. Alex. McDermott.

Photinus marginellus responds definitely to the action of oxygen.
The light emitted by this tissue increases in brilliancy when placed in
oxygen, and tissues, the light of which has been wholly extinguished
in carbon dioxide, become instantly luminous when placed in oxygen.
This author also observed that during life all the tracheoles are filled
with air. On the other hand, Kélliker “ found oxygen to have no real
exciting action on the resting luminous organ of the firefly, and that
the live insects lightened in oxygen only after immersion therein for
an hour or longer. However, they then lightened very brilliantly.
Giesbrecht * also concludes that the lighting of marine animals
occurs without the aid of free oxygen, and Bongardt? also found the
intensity of the light of certain fireflies to diminish in a current of
pure oxygen, so that at the end of forty minutes only two insects were
emitting light, and these but feebly, whereas, after shutting off the
current of gas, the insects lightened intensely after one and six hours.
Macartney * cites Sir Humphry Davy as authority for the statement
that the light of the glow-worm is not increased in oxygen or chlorine,
nor diminished in hydrogen. Carradori* considered that oxygen
could have no effect upon the luminosity of the Luciola italica, since
this insect continued to shine when immersed in oil, which he claimed
could contain no free oxygen.

We found the luminous organs of the common firefly to glow in
oxygen, but not very brilliantly. The glow persisted for some time,
finally dying out. On immersion in hydrogen the organs which had
lost their glow in oxygen did not glow. Such organs, however, were
found to emit light on percussion at the conclusion of the experiment. —
A live insect placed in oxygen glowed feebly, but with no sudden and
intermittent flashes * of light such as characterize the light emission
by this insect under normal conditions.

* As is well known, the light of our common firefly (Photinus pyralis Linn.)
is emitted intermittently at more or less regular intervals in distinct and definite
flashes. In other words, the emission of light might almost be said to partake
of the nature of a luminous explosion or coruscation. As we shall see, this sudden
and intermittent emission of light can also be brought about by means of certain
chemical stimuli. Usually, however, the light production as brought about by
chemical stimuli is not of this character, but shows itself as a distinct and per-
sistent glow which slowly and gradually spreads over the luminous organ and

which continues without interruption for a considerable interval, sometimes for
an hour or even longer.

Observations on the Production of Light by the Firefly. 131

_ Nitrogen. — Bischoff ? observed that certain luminous rhizomorphs
did not entirely lose their luminosity when brought into an atmos-
phere of nitrogen. Macaire* states that the light of the living organ-
ism is extinguished in indifferent gases. According to Spallanzani,®
no living organism produces light in non-respirable gases. Jousset de
Bellesme ~ observed the same conduct of Lampyride in nitrogen as
in carbon dioxide, and Pfliiger ** states that the light of dead fish is
extinguished in nitrogen and reappears on the admission of air. In our
own experiments we found that the luminous organ of the firefly did
not glow in nitrogen during a period of fifteen minutes; such organs,
however, were found to emit light on percussion after removal to the
air.

Hydrogen. — According to Bischoff,’ the light of luminous rhizo-
morphs is entirely extinguished in hydrogen. Jousset de Bellesme *
observed the same conduct of Lampyride in hydrogen as in nitrogen
and carbon dioxide. Pfliiger * also calls attention to the fact that the
light of dead fish is extinguished in hydrogen, but reappears on the
admission of air. On the other hand, Bongardt* found Lampyris
noctiluca to lighten in hydrogen; the light gradually became weaker
and ceased to be emitted after about fifty minutes. After four hours
the insects lightened after shutting off the supply of gas and then
lightened again on turning on the gas the next morning. We observed
the luminous organ of the firefly to glow in hydrogen, but not very
brilliantly. However, the weak glow persisted for some time, finally
dying out. These organs glowed again when placed in oxygen.

Carbon dioxide. — Jousset de Bellesme*! found Lampyridz not to
lighten in carbon dioxide. If, however, they were kept in this gas
several hours and were then placed in a current of air, they lightened
intensely. Bongardt? observed that Lampyris noctiluca was not
killed by carbon dioxide, but that it lightened less intensely. Accord-
ing to Pfliiger,** the light produced by living organisms is extinguished
in carbon dioxide. He also pointed out that the light-producing sub-
stances of dead fish and decaying wood are destroyed on immersion
in water saturated with carbon dioxide. Watasé* found carbon
dioxide to extinguish the light of the crushed luminous organ of the
firefly, and Miss Townsend *° found the luminous tissue of Photinus
- marginellus to be extinguished in this gas.

We observed that when the luminous organs of the firefly are brought

132 Joseph H. Kastle and F. Alex. McDermott.

into carbon dioxide, they do not glow at first; after some time, how-
ever, they glow faintly; this glow persists for some time and finally
dies out. After an hour’s exposure to the gas they glowed on percus-
sion in the air. In certain of our, experiments with inert gases results
were obtained which seemed to indicate that the emission of light
by the luminous organ in these gases was more the result of a mechani-
cal stimulus due to variations in temperature and gas pressure than
to the chemical action of the gas. In fact, our results show a distinct
analogy to the influence of air currents and slight changes of tempera-
ture on the strychninized frog.

Carbon monoxide. — Bongardt* found Lampyris noctiluca to
lighten in carbon monoxide after a few minutes’ exposure. At the
end of ten minutes the insect ceased to lighten.. After three hours in
the gas the insect was apparently dead, but after sundown those ex-
posed to carbon-monoxide still lightened. In carbon monoxide we
observed the resting luminous organs of the firefly to glow dimly and
the glow to be slow in appearing.

Chlorine. — Macaire”* found the light-producing material of living
organisms to be permanently destroyed by chlorine. On the other
hand, Kélliker * found chlorine to act as a stimulus to the lighting
of the firefly. We observed that the quiescent luminous organs of the
firefly glowed faintly in chlorine and then went out. The organs
exposed to this gas were found to be almost dead to percussion after
a short exposure, whereas a control which had been kept in the air
for the same length of time glowed very brightly _on percussion.

Nitrous oxide. — According to Macaire,” “‘le gaz oxide d’azote
produces nearly the same effect as oxygen on the production of light
by Lampyris splendidula, namely, a brighter light than air. We
have observed that on placing a live firefly in nitrous oxide it lightened
once. The luminous organ then became quiescent. Later it emitted
a very bright, steady glow, which was maintained for some time.
When brought out into the air, it made only a few irregular move-
ments with its legs during the first ten minutes, but later recovered
completely and was finally lost sight of. The detached luminous
organs were found to glow slowly but brightly in nitrous oxide, the
light emitted being more yellow in color than that ordinarily produced.

* One can scarcely be sure from this name that MAcArRE meant nitrous
oxide. — J. H. K.

Observations on the Production of Light by the Firefly. 133

Nitric oxide. — In nitric oxide Bongardt ® found live insects to move
about unquietly, and after half an hour no more lighting was observed.
In four minutes after closing the tube three of the insects lightened
intensely. On conducting the nitric oxide again through the tube the
emission of light ceased in eleven minutes. We observed the lumi-
nous organs of the firefly to glow feebly when placed in this gas and
then die out. They glowed again on being brought into the air.

Nitrogen tetroxide. — We have not been able to find any references
in the literature to the action of nitrogen tetroxide. We have found
that the conduct of the luminous organs of the firefly in this gas is
similar to that in nitric oxide, only the glow is somewhat stronger.

Ammonia.—In ammonia we observed the luminous organs to
glow brightly after a short delay.

Hydrogen sulphide. — According to Macaire,” the luminous organ
. of the firefly is entirely deprived of its hght-producing power by
hydrogen sulphide. Pfliiger®* also states that the lght-producing
substances of decaying fish and wood are destroyed by solutions of
hydrogen sulphide. Jousset de Bellesme* observed that hydrogen
sulphide extinguished the light of Lampyris immediately, leaving the
photogenic cells intact in form, but robbing them of their photogenic
function. In one instance we found that the luminous organ of the
firefly did not glow in hydrogen sulphide, but did so on being brought
out of this gas into the air. In another instance we found a luminous
organ to glow feebly in hydrogen sulphide and then to die out; it
also glowed again on bringing it out into the air. This phenomenon
could be repeated at will several times, the organ ceasing to glow in
hydrogen sulphide but glowing again on exposure to the air.

Suphur dioxide. — Carradori*® found sulphur dioxide to extinguish
the light of the luciole; Dubois ® made the same observation on the
cucuyo (Pyrophorus noctilucus), and Macaire* on the glow-worm.
Of all the substances which we have thus far studied sulphur dioxide

has been found to be the most quickly poisonous to the light-producing |

function of the firefly. When the detached luminous organs are placed
in sulphur dioxide, they do not glow, and if glowing when placed in it
they are immediately extinguished. Luminous organs which had
been exposed to sulphur dioxide were found not to glow in oxygen,
and were found to be dead to percussion when removed from this gas.
These phenomena were repeated several times on different luminous

134 Joseph H. Kastle and F. Alex. McDermott.

organs, the light of the glowing, luminous organs going out almost
instantly on exposure to this gas.

Carbon bisulphide. — Kolliker * found carbon bisulphide to have no
effect upon Lampyris. Heinemann” found the light organs of Mexi-
can cucuyos to be killed in carbon bisulphide after a short time. We
have observed that the resting luminous organs of the firefly glow
very strongly in the vapor of carbon bisulphide. The glow is slow in
appearing, but very bright. One of the tubes containing carbon
bisulphide and the luminous organ was accidentally dropped upon the
floor; the glowing, luminous organ broke into a number of pieces, each
of which glowed very strongly, showing distinct coruscations. Of
all substances which we have thus far examined, carbon bisulphide is
probably the most powerful exciter of luminosity in the luminous
organ of the firefly, and gives rise to phenomena which in a darkened
room are very striking and beautiful.

Cyanogen. — We found the luminous organs of the firefly to glow
in cyanogen gas. They soon went out, however, turned brown, and
were found to be dead to percussion after a short exposure to the gas.
The experiment was repeated several times with like results.

Hydrocyanic acid. — According to Bongardt* the luminous organs
of Lampyris noctiluca are killed by exposure to hydrocyanic acid, and
the live insects do not lighten in this gas, but five hours after appar-
ent death one of these thus exposed did lighten again intensely, and
another three hours after apparent death in the gas. Heinemann ™
found that on bringing a moistened piece of potassium cyanide in
close proximity to the glowing organ of the firefly, it goes out in a
short time and is apparently dead. Weakly luminous organs were
found to be killed by such an exposure without showing any preliminary
stimulation. According to Michaelis,?* ten minutes’ exposure to
hydrocyanic acid weakens the luminescence of glowing sea water,
and after thirty minutes the light is entirely extinguished.

We found that the resting luminous organs of the firefly glowed
strongly in hydrocyanic acid gas and then died out, only, however,
to glow again strongly. Certain of the organs glowed for some time,
and after a few minutes’ exposure to the gas were found to glow on
percussion after they were taken out into the air. Such organs were
observed to smell strongly of hydrocyanic acid, indicating that a
certain amount of the substance had been actually absorbed.

Observations on the Production of Light by the Firefly. 135

Iodine cyanide.—In the vapor of iodine cyanide the luminous
organs were found to glow faintly and for some time, after which they
died out. Such organs were then found to be colored light brown,
and did not glow on percussion.

Hydrofluoric acid. —In hydrofluoric acid gas we found the lumi-
nous organ of the firefly to give a slow but distinct glow. After
remaining in the gas for some time it glowed brightly, and immer-
sion in the solution of the acid failed to extinguish it.

Ether. — Heinemann observed that ordinary ether acts as an
irritant towards the luminous material. It kills the luminous organ,
however, in four to five minutes, converting it into a solid, heavy
mass. Ko6lliker* also found ether vapor to stimulate light produc-
tion. Pfaff * observed sea water to lighten on the addition of ether.
According to Macaire,” ether destroys the light-producing material
of the firefly.

We etherized a firefly which was not lighting. A short time after
being placed in the ether vapor it showed one short but vivid corusca-
tion, and then one or two faint flashes of light, after which the lumi-
nous organ burst into a fine, steady glow which lasted for some time,
dying out after several minutes. The detached resting luminous
organ of the firefly in ether vapor also gave one or two preliminary
minute flashes, and then burst into a full, steady glow lasting for some
minutes. These interesting observations were repeated a number of
times, always with a like result.

Chloroform. — Kolhker #3 found chloroform vapor to act as a stimu-
lus to light production by the Lampyride. Heinemann " also found
it to act as an irritant, similarly to ether. In order to determine the
effect of chloroform on the luminous activity of the firefly, an active
insect was placed in the vapor of chloroform. It at once showed loss
of ordinary bodily activity; one or two bright flashes of light were
emitted, and then the luminous organ burst into a strong, steady
glow which persisted for half an hour. At the end of this time both
the insect and the luminous organ were apparently dead. This obser-
vation was repeated a number of times.

The resting luminous organ of a firefly was placed in the vapor of
chloroform; in a short time it showed a few faint flashes, and then
burst into a full, steady glow, dying out more quickly than in ether.
This phenomenon could be repeated at will. Luminous organs from

136 ©. Joseph _H. Kastle and F, Alex. McDermott.

fireflies, one of which was still alive after twenty-four hours’ cap-
tivity in a plain glass bottle, and the other dead, were placed in chloro-
form vapor; both glowed, but the glow was slower in developing and
fainter than in the case of luminous organs from freshly captured,
active insects.

Carbon tetrachloride. — When placed in the vapor of carbon tetra-
chloride, the luminous organs of the firefly glowed quickly and strongly,
the glow being followed by a series of brilliant and repeated corusca-
tions, and with the production of bright points and pulsations of light
throughout the entire luminous organ, persisting for a considerable
time. This substance is one of the most powerful exciters of luminosity
in the firefly that we have thus far met with.

Ethyl chloride. —'The luminous organ of the firefly was found to
glow after some delay in the vapor of ethyl chloride. It soon went
out, however, and showed no luminosity on being brought out into
the air, and responded only feebly to percussion.

Ethyl bromide. — In ethyl bromide the luminous organ of the firefly
glowed feebly after remaining in the vapor for some time.

Bromoform.—n the vapor of bromoform the luminous organ of
the firefly glowed slightly after remaining in the vapor for some time.

Iodoform.—In the vapor of iodoform at room temperature the
luminous organ of the firefly glowed very faintly after remaining in
the vapor for some time.

Ethylene bromide. — In the vapor of ethylene bromide the luminous
organ of the firefly was found to glow brightly and persistently for a
long time.

Ethyl alcohol, — Giesbrecht “ states that ethyl alcohol acts as an
irritant to the luminous material. Heinemann ' found that absolute
alcohol instantly destroys the power of the Mexican cucuyo to lighten.
Ko6liiker 8 found alcohol of 45 per cent and higher concentration to
act as a stimulus. Pfaff * observed sea water to lighten on the addi-
tion of alcohol, and Artaud! obtained similar results. On the other
hand, Macaire * found alcohol to destroy the light-producing material
of the firefly. According to our own observations, the detached
luminous organs of the firefly glow brightly in the vapor of ethyl
alcohol, finally showing a succession of brilliant luminous pulsations,
and then dying out.

Methyl alcohol. — The results obtained were similar to those obtained

Observations on the Production of Light by the Firefly. 137

with ethyl alcohol. With both of these alcohols the luminous effects
were very striking.

Amyl alcohol. —In the vapor of amyl alcohol the glow obtained
with the luminous organ of the firefly was faint and slow in develop-
ing; it then became somewhat brighter, and then died down.

Allyl alcohol.— In the vapor of allyl alcohol the luminous organ
was slow in glowing. It then began glowing in one corner of the
luminous organ; after a time, however, the glow spread over the
entire organ, which then glowed uniformly and brightly for some
time.

Acetone. —In the vapor of acetone the luminous organ of the
firefly glowed brightly after a short exposure and then died down.

Formaldehyde. — According to Giesbrecht,“ formaldehyde excites
certain luminous marine forms to luminescence. We have also ob-
served that the luminous organs of the firefly glow steadily in the vapor
of formaldehyde, but only for a short time.

Illuminating gas. — The luminous organs of the firefly were found
to glow feebly but persistently in illuminating gas.

Acetylene. — In acetylene the luminous organs of the firefly glowed
faintly and then went out; they then glowed again faintly and again
died out; glowing intermittent. The organs glowed brightly on re-
moval to the air and also on percussion.

Benzene. — In benzene Heinemann "° found the light of the Mexican
cucuyo to be unchanged after half an hour; they were killed, how-
ever, in three quarters to one hour. In benzene we found the luminous
organs of the firefly to glow brightly and for some time, the glow finally
dying out.

Petroleum. — Heinemann * observed that in petroleum the light
of the Mexican cucuyo continued to be given out for some time. We
have found that in the vapor of petroleum ether the luminous organ
of the firefly glowed quickly and showed a succession of distinct light
flashes.

Phenol. —In the vapor of phenol the detached luminous organ
of the firefly showed a slight glow in five minutes; this continued
dimly for some time, when a few bright spots appeared which gradu-
ally spread over the entire organ, and in forty-five minutes the entire
organ was in a bright and steady glow. At the end of one hour and
forty minutes the glow had died down.

138 Joseph H. Kastle and F, Alex. McDermott.

Ortho-cresol. — Kélliker > found creosote to act as a stimulus to
the lighting of the luminous organs of the firefly. In the vapor of
ortho-cresol we observed that a few faint bright spots had made
their appearance in fifteen minutes; in forty minutes the entire organ
was in a bright, steady glow, which was maintained for an hour and
forty minutes, when the light went out.

Para-cresol. —In the vapor of this compound the detached lumi-
nous organs of the firefly showed a faint glow after half an hour; then
a succession of bright points appeared, which gradually spread until
the entire organ was in a state of bright, steady glow, which was still
bright at the end of one hour and forty minutes.

Amyl nitrite. —In the vapor of amyl nitrite a live firefly glowed
strongly for five to ten minutes, without the appearance of any sepa-
rate or distinct flashes or coruscations. When removed from the
nitrite vapor, this insect was apparently dead, and the luminous
organ failed to emit light on percussion, indicating that the amyl
nitrite acts as a specific poison to the luminous material. The de-
tached luminous organs were also found to glow strongly in the vapor
of amyl nitrite, and on percussion after removal from the vapor they
showed only faint luminosity.

Mononiirobenzene.— With nitrobenzene some very interesting
results were obtained. When placed in the vapor of this substance,
luminous organs of the firefly glow faintly after a short time; the glow
gradually becomes brighter, and the organ then shows a succession of
brilliant flashes, after which it settles down into a bright, steady glow,
which in some instances lasted as long as seven hours. The luminous
flashes which are observed when the luminous organ of the firefly is
placed in the vapor of this compound do not cover the entire organ at
any one time, but appear to run over it rapidly, reminding one of the
spread of combustion through a mass of moist gunpowder.

Essential oils. — The conduct of resting luminous organs of the
firefly was also tested towards the vapor of a number of essential oils,
with the following results. In oil of wintergreen the organ showed a
slight glow, confined to small areas and soon dying out; it gave a
strong glow on percussion after removal from the vapor. Similar
results were obtained with oil of cloves, oil of cinnamon, oil of ber-
gamot, and oil of rosemary. In the vapor of oil of eucalyptus the organ
exhibited a strong glow, slow in developing. The detached organs

Observations on the Production of Light by the Firefly. 139

failed to emit light in the vapor of oil of peppermint, but exhibited a
fairly strong glow on percussion after removal from the vapor. With
oil of lavender beautiful phenomena were observed; in the vapor
of this oil the organs exhibited a strong glow, slow in developing, and
also a series of flashes or distinct light pulsations, appearing irregularly
in spots over the luminous organ, and gradually spreading over the
entire surface, lighting it up very brightly in patches, similar in gen-
eral effect to the spread of a conflagration or the setting off of a num-
ber of explosives by means of a common fuse.

In addition to: these gases and vapors, whose effect on the photo-
genic process has been described in the foregoing, the effect of a large
number of other substances, in the liquid condition or in solution, has
been investigated by ourselves and others. These include aqueous
solutions of various acids, alkalies, and salts, and also alkaloids and
such liquid substances as glycerine, etc. As might be expected, the
results obtained by different investigators are in many instances con-
flicting. These differences are doubtless to be accounted for by
reason of the fact that the various substances tested were employed
at different concentrations or are due perhaps to some idiosyncrasy on
the part of the particular photogenic organism studied. Thus Gies-
brecht “ states that glycerine is a stimulus to light production, whereas,
according to Heinemann," it is exceedingly toxic in its action on the
luminous organs of the Lampyride.

Lack of space prevents any detailed account of these observations.
It is perhaps sufficient to say in this connection that, in addition to
those substances whose effects on the photogenic process have already
been described, the following have been found to stimulate the pro-
duction of light by various luminescent life forms: strong acids, such
as hydrochloric, nitric, sulphuric, chromic, etc., and various salts,
such as the halogen and neutral salts of the alkali metals and those of
the alkaline earths. Also sodium carbonate, sodium sulphate, and the
diphosphate, potassium hydroxide, potassium carbonate, potassium
nitrate, potassium cyanide, potassium ferrocyanide, potassium per-
manganate, calcium chloride, silver nitrate, mercuric chloride, and
even chemically indifferent substances such as cane sugar. All of
these have been found, by various observers, to act as stimuli to
light production. Dilute, aqueous ammonia has also been found to
be a powerful stimulus; indeed, according to Watasé,” a true lumi-

140 Joseph H. Kastle and F,. Alex. McDermott.

nous tissue will glow in dilute ammonia when all other stimuli have
failed. According to Koélliker, all nerve stimuli cause the emission of
light when allowed to act on a luminous organ, and conversely all |
substances which prevent the transmission of the nerve impulse act
as poisons towards the luminous organ. Among the substances which
have been found to be either toxic or without action on the photo-
genic process may be mentioned the following; certain strong mineral
acids, acetic acid, salts of certain of the heavy metals, such as those
of copper, etc., very dilute solutions of sodium chloride, ammonium
carbonate, magnesium sulphate, conine, and strychnine. According
to Macaire * and also according to Pfliiger,®® all chemical agents which
coagulate albumin or bring about its chemical decomposition destroy
the light-producing material of luminous plants and animals.

Without entering into the details of all of our own experiments,
it may be said that strychnine, adrenalin, hydroxylamine sulphate,
hydrazine sulphate, the nitrites of sodium, potassium and bafium,
sodium hydroxide, sodium acetate, sodium bicarbonate, disodium
phosphate, sodium fluoride, and sodium sulphite, all acted as powerful
stimuli to light production, whereas pure water, sodium chloride,
sodium bromide, sodium nitrate, sodium benzoate, potassium nitrate,
potassium iodide, barium chloride, etc., showed but little or no effect.
Strychnine, hydroxylamine sulphate, and the nitrites were especially
remarkable in their effect upon the light organ of the firefly. When
a drop of a solution of strychnine sulphate was injected into a live,
active firefly, it was observed that invariably the insect gave one or
two strong flashes of light, after which the luminous organ became
quiescent and the insect apparently dead. In a short while, how-
ever, the entire luminous organ began glowing, and a great many
bright scintillations or bright points of light followed one another in
rapid succession over its entire area. This phenomenon was repeated
a number of times with other specimens of the insect. Hydroxyla-
mine sulphate, adrenalin, and the nitrites of sodium, potassium and
barium, also acted as powerful stimuli. Injections of these substances
into the live insect, or the placing of the detached luminous organs
in their solutions, invariably resulted in the production of a strong,
steady glow, the nitrites especially causing the emission of light from
the luminous organ when other chemical stimuli had failed.

So far as their effect on the luminous organ or the photogenic ma-

Observations on the Production of Light by the Firefly. 141

terial is concerned, it seems that substances are roughly divisible into
four groups, namely, (1) those substances which distinctly stimulate
the light production, called by some authors “exciters’’ (Erregern);
(2) those substances which distinctly inhibit luminescence, probably
by poisoning the luminous organ or photogenic substance (poisons) ;
(3) inert substances, such as nitrogen, hydrogen, etc., which inhibit
luminescence only so long as they are present in excess in the atmos-
phere surrounding the insect or photogenic substance, but which are
without any well-defined toxic action on the photogenic substance
(non-toxic inhibitors); and (4) substances which have little or no
effect on the light production (indifferent substances).

It is also evident from the foregoing that no definite relationship
exists between chemical composition and power to excite the living
photogenic material to luminescence, nor can the effect of these various
chemical stimuli be distinctly attributed to any particular group or
ion, without it is the nitro group. It is of interest to note, however, in
this connection that the anesthetics and related compounds act as
powerful stimuli to the photogenic process.

ON THE EFFECT OF WATER ON THE PRODUCTION OF LIGHT BY THE
PHOTOGENIC MATERIAL OF THE LIVING ORGANISM.

Of the many substances whose effect on the production of light in
the living organism has thus far been investigated, water apparently
plays the most essential and interesting réle. The fact that, in com-
mon with all living tissues, the luminous organs and light-producing
cells of animals and plants normally contain large amounts of water
tended naturally to preclude the recognition of the essential part
played by this compound in the production of light in the living organ-
ism. Thus, according to Kélliker,?’ water has no effect on the pro-
duction of light by the firefly. It came gradually to be recognized,
however, that water plays an essential part in the photogenic process.
Thus Carradori,t in 1798, mentions the fact that the light of the
Italian luciole is suspended by drying, but is again revived by soften-
ing in water, though not after too long a period of desiccation. He
mentions that Reaumur, Beccaria, and Spallanzani had observed the
same thing with regard to the Pholas and Meduse. Carus® made a
similar observation in 1864, and Panceri®® in 1872. According to

142 Joseph H. Kastle and F, Alex. McDermoit.

Bongardt,’? Panceri*° pointed out that Pholas and Phyllirhoé which
had been dried out for ten days, lightened again when moistened with
water. In 1887 Dubois’ confirmed Panceri’s observation on the
Pholas dactylus. Jousset de Bellesme* found the Lampyride to
lighten more frequently in moist than in dry weather, — even the dis-
membered organ after four days. Giesbrecht ™ filtered certain cope-
pods on gauze and observed that they began to lighten as soon as
water was poured over them. He also found the secretion of the lumi-
nous glands to lighten on being brought in contact with water. Ac-
cording to Pfliiger,** the fact that the light-producing material can be
dried and then made to glow again by moistening with water speaks in
favor of the view that such material is actually living. According to
this author, this phenomenon is not shown by the light-producing
organs of the firefly, but only with the light-producing material of
the Pholas, which is very tenacious of life. On the other hand,
Bongardt * found that the dry powder obtained by drying and crushing
the luminous organs of Lampyris noctiluca retains its power of glow-
ing on moistening with water for over a year. This author concludes
that a material is produced in the light organ of this species which
lightens when it is brought to the proper degree of moisture. Miss
Townsend *° also observed that when the luminous organs of the fire-
fly are crushed and dried and reduced to powder, the powder lightens
under the influence of air and moisture, thereby proving, according
to this author, that the production of light is independent of the life
of the cell.

All things considered, the remarkable effect of water in exciting the
dried material of the luminous organs of photogenic life forms to
luminescence is certainly one of the most striking phenomena in the
whole field of biochemistry, and is obviously of such interest as to
warrant further investigation. It therefore occurred to us to repeat
these observations on the dried material of the luminous organs of
the firefly, in the hope of extending our knowledge of this particular
phase of the subject.

It is known, of course, that if the fresh luminous organs of the fire-
fly be rubbed up in a mortar, the whole mass glows actively for some
time as the result of mechanical stimulation. As the mass dries,
however, the light gradually dies out, but on the addition of water
the glow reappears. We have found it impossible, however, to pre-

Observations on the Production of Light by the Firefly. 143

. pare a luminously active dry powder by the complete drying of the
luminous organs of the firefly in the air, a fact which may account for
Pfliiger’s idea, namely, that a material of greater vitality than the
luminous organ of the firefly is required for the preparation of dry,
luminously active material. Luminous organs of the firefly which
had dried out spontaneously in the air at summer temperature were
found not to emit light when ground to a powder and moistened with
water. Similarly a fresh luminous organ of the firefly was crushed
between folds of Japanese bibulous paper and allowed to dry spon-
taneously in the air at room temperature for two or three days. The
dry material thus obtained did not lighten on moistening with water.
On the other hand, by drying whole insects or fresh luminous organs
in vacuo over sulphuric acid, it has proved an easy matter to obtain
dry material which glows vigorously on moistening with water, and
which retains its photogenic activity for long periods of time. With
the view of throwing still further light on the preparation of the dry,
luminously active material, an equal number of live, active fireflies
and active luminous organs of fireflies were dried in a vacuum over
sulphuric acid. Attention has already been called to the peculiar and
interesting conduct of the live fireflies 77 vacuo, and to the fact that
under the influence of the vacuum all the luminous organs burst into
a steady glow which lasts for some time. After eighteen hours in the
vacuum the insects were found to be dead and as brittle as glass. The
two sets of dry luminous organs thus obtained were ground to fine
powders. The material obtained from the luminous organs which
had been detached from the insects previous to drying 7m vacuo was
labelled (1), whereas that obtained from the luminous organs of the
fireflies which had been dried whole was labelled (2). On moistening
small amounts of powders Nos. (1) and (2) with water, both glowed
instantly, but the light emitted by (2) was feeble as compared with
that emitted by (1). Both continued to glow for one hour and thirty
minutes; at the end of this time both were again moistened with water.
_ With (1) the glow was renewed and lasted for an hour longer, whereas
(2) failed to glow a second time. Hence a more active material may
be obtained by drying the detached luminous organs than by drying
the entire insect. Taking everything into consideration, this is prob-
ably what one would be inclined to expect. The fact that the lumi-
nous organs of the firefly, whether still a part of the insect or detached,

144 Joseph H. Kastle and F, Alex. McDermott.

glow so long and persistently in a vacuum, points, of course, to such an
expenditure of the light-producing material as to greatly diminish the
luminous power of the dry material obtained by this method, and
naturally suggested the advisability of drying the luminous organs
in hydrogen, or im vacuo and hydrogen, with the view of preventing
this waste of the photogenic substance. In fact, one of our experi-
ments, in which a number of luminous organs were being dried in
hydrogen and vacuo with considerable diminution of the amount of
light emitted, gave every promise of enabling us to overcome this
waste of photogenic material, when, unfortunately, concentrated
sulphuric acid was sucked over into the tube containing the luminous
organs, and, occurring as it did towards the end of the firefly season,
put an end to attempts in this direction.*

In the mean time a quantity of the dried powder of the luminous
organ of the firefly had been obtained by drying the luminous organs
in vacuo in the usual way. With the view of learning what we could
with regard to the stability of the dry material, small amounts of
it were put up in sealed glass tubes in air and im vacuo. These tubes
were prepared on August 10, 1909. On January 4, 1910, two of the
tubes were opened and the material moistened with water. Both
specimens, namely, the one which had been preserved in air and that
which had been preserved in vacuo, glowed actively for fifteen minutes.
Two of the tubes were again tested at the time this communication
went to press (September 9, 1910), and both specimens were again
found to be active. It is evident, therefore, that, when dry, the photo-
genic material of the firefly retains its power to emit light through the
action of water, for a period of thirteen months. It is proposed to
keep the remainder of the tubes for longer periods and to test them
at intervals of six months or a year, until the supply is exhausted,
in the hope of obtaining further information on this phase of the sub-
ject. Already, however, certain facts of interest have been brought
to light concerning the conduct of the dry material. In our first
work with the powder obtained by drying the luminous organs of the ~
firefly in vacuo, during the summer of 1909, it was observed that while
the dry material always glowed actively on the first addition of water,

* Since the above was written, it has been found possible to prepare very
active photogenic material, by drying over sulphuric acid in hydrogen, under
diminished pressure.

Observations on the Production of Light by the Firefly. 145

it failed to glow again, on moistening, after drying a second time.
In other words, it appeared from our first experiments that the dry
material resulting from the first drying 7m vacuo contained only a
certain limited amount of photogenic material which, on moistening
with water, spent itself in the emission of light; or, in still other words,
that the reaction between the dry photogenic substance and the water
was complete so far as the emission of light energy is concerned. We
have since obtained results, however, which indicate that such is not
the case, but that the dry material which has once emitted light through
the action of water may, after drying again, emit light a second time
on moistening, and may even emit it a third time by the action of
water. Thus the specimens which had glowed on moistening with
water on January 4, 1910, were allowed to dry in the air at room
temperature. On January 8, 1910, they were found to be quite dry,
and on moistening again with water, both of them glowed freely and
for about ten minutes, although not so intensely as they did the first
time that they were moistened. These specimens were then allowed
to dry in the air again, and on January 12, 1910, they were moistened
again (the third time), when one of them, the specimen which had
originally been preserved iu vacuo, glowed quite strongly again, the
glow lasting for fully five minutes. The two specimens were then
allowed to dry out again, and were again moistened on January 18,
Ig10; neither of them emitted light. On January 10, 1910, two of
the specimens which had been preserved in sealed tubes in air were
moistened (1) with distilled water, and (2) with N/2o0 solution of
sulphur dioxide respectively. The specimen which had been moist-
ened with distilled water glowed brightly for thirty-five minutes.
The specimen which had been moistened with N/20 sulphur dioxide
was slower in becoming luminous, glowed only feebly, and ceased
glowing in three minutes. The two tubes were then allowed to dry
in the air, and on January 14, 1910, both were moistened with distilled
water; (1) glowed feebly for about two minutes, and then went out,
whereas (2) did not show any glow at all. The two specimens were
then allowed to dry in the air again, and on January 18, r1o10, they
were again moistened (the third time); neither specimen showed any
luminosity.

It is evident from these experiments that the latent energy of the
photogenic material obtained by drying the luminous organs of the

146 Joseph H. Kastle and F. Alex. McDermott.

firefly in vacuo is not always completely liberated by a single appli-
cation of water, or, if such is the case, then a certain amount of the
photogenic material is regenerated on drying at room temperature
under the conditions already described. The fact, however, that
the material ultimately loses its power to emit light on being brought
into contact with water indicates that it is actually consumed during
the process, and that it cannot be produced continuously from the
other substances present under the conditions at hand. The fact
that the material lightens so feebly and for such a short time when
moistened with a dilute solution of sulphur dioxide, and that, when
so treated, it fails entirely to glow on moistening with water after
drying in the air, again illustrates the poisonous nature of this sub-
stance to the photogenic process, and agrees with all of our other
observations on the effect of this substance on the light-producing
power of the firefly. ,

In order to test the conduct of the dry photogenic material towards
water, in the presence of various gases, the following experiments
were carried out. Small amounts of the dry photogenic powder, ob-
tained by drying the luminous organs of the firefly 7m vacuo over
sulphuric acid, were placed in small receptacles made of a single thick-
ness of Japanese bibulous paper. These were then placed in a tiny
basket of platinum wire suspended from the end of a small separatory
funnel, and so arranged that the contents of the basket could be
moistened with water from the separatory funnel. The stem ofthe
separatory funnel, carrying the basket, was then placed in a large
glass tube by means of a doubly perforated stopper, through the
second opening of which was inserted an exit tube for the escape of
gas from the large tube. The latter was about three inches in length
and was drawn out at the lower end into a smaller tube about three
inches long and one quarter of an inch in diameter. The small tube
served as an inlet tube for the introduction of any gas with which
we desired to fill the central chamber of the apparatus in which the
dry photogenic material was suspended. By means of this simple
device we were able to study the effect of freshly boiled and cooled
distilled water on the dry photogenic material in atmospheres of the
following gases:

Hydrogen. — The dry powder showed no luminosity in dry hydro-
gen. When moistened with a drop or two of water, it glowed for five

- ete OP Ne Er

Observations on the Production of Light by the Firefly. 147

minutes, and then the light died out. The admission of the vapors
of carbon disulphide and carbon tetrachloride, both of which were
found to greatly stimulate the fresh luminous organ of the firefly,
caused no return of the light when admitted along with the hydrogen.

On shutting off the current of hydrogen, however, for about five
minutes, the moist material began glowing again; on again admitting ~
hydrogen, the light died out, and on again shutting off the hydrogen,
the light reappeared.. This phenomenon was repeated several times,
and was doubtless due to the slow diffusion of air into the tube con-
taining the moistened photogenic material.

Carbon dioxide. — The dry powder exhibited no luminosity in carbon
dioxide. On moistening it with a drop or two of water the material
glowed for four minutes, at the end of which time the light died out.
Air was then admitted, when the material glowed again in two minutes.

Sulphur dioxide. — The dry powder showed no luminosity in sul-
phur dioxide. Even after moistening it with a drop or two of water
it still did not become luminous, and the subsequent admission of
air and air with carbon disulphide failed to produce a glow. The
photogenic material was evidently killed by sulphur dioxide.

Hydrogen sulphide. —The dry powder showed no luminosity in
hydrogen sulphide. After moistening with a drop or two of water
it glowed for two minutes in this gas and then went out. Air was ad-
mitted, and after some delay the moistened material again glowed
brightly.

Oxygen. — The dry powder showed no luminosity in pure oxygen.
On moistening with a drop or two of water the material instantly
lighted up with a bright glow which continued practically undimin-
ished for half an hour.

Air. — As already pointed out, the dry powder does not glow in
the air. On moistening it with a drop or two of water, it glowed
rather dimly for an hour. This experiment served as a control on the
other five, and furnished us with a basis for comparison.

Sulphuric acid. — Neither concentrated nor dilute sulphuric acid
caused the evolution of light from the dried photogenic organs of the
firefly. In the concentrated acid the pale yellow color of the ventral
surface of the dried luminous segments of the abdomen of the firefly
became first light and then dark green and finally dark blue, and a
transient blue coloration was produced in the acid.

148 Joseph H. Kastle and F, Alex. McDermott.

Nitric acid. — Neither concentrated nor dilute (10 per cent) nitric
acid caused the evolution of light from the dry photogenic tissue.
The concentrated acid stained the tissue a deep yellow.

It would seem therefore, from these experiments, that in the pro-
duction of light by the firefly two substances are actively concerned,
in addition to the photogenic material itself, namely, water and oxygen.
In other words, the light results from the action of oxygen and water
on the photogenic material. As to the precise mechanism of these
changes we at present know nothing. The fact that some light is
emitted when the dry photogenic material is moistened in an atmos-
phere of certain gases other than oxygen, such as hydrogen, carbon
dioxide, etc., indicates, in all probability, that the gases contained
small amounts of air, or that the water employed contained small
amounts of dissolved oxygen, or that the dry photogenic material
itself contained sufficient oxygen, in loose combination, to evolve a
limited amount of light when the dry material is moistened in such
atmospheres.

As one would naturally expect, the effect of drying the photogenic
material of the firefly is to greatly increase its stability towards heat.
That such is, the case is indicated by the following observation: A
small amount of the dry material, enclosed in an envelope of bibulous
paper, was placed in a test tube. The test tube was then closed with
a cork and heated for ten minutes in boiling water. When removed
from the tube at the end of this time and moistened with water, it
glowed quite as brightly as some of the unheated material, indicating
no loss of activity by exposure to the temperature of boiling water
for ten minutes.

SUMMARY.

Our present knowledge of the photogenic process may be prety
summarized as follows:

So far as the purely chemical aspect of the process is concerned,
three things seem to be concerned in the production of light by living
things. These are the photogenic substance, water, and oxygen. As
yet nothing is known regarding the precise chemical nature of the
photogenic substance. Like other biologically active substances, ‘it
is characterized by extreme irritability. Under the influence of cer-
tain chemical stimuli, especially of such substances as ether, chloro-

Observations on the Production of Light by the Firefly. 149

form, carbon bisulphide, carbon tetrachloride, and mononitro-
benzene and the nitrites of certain metals, the production of light
by the luminous organ of the firefly is a continuous process, extend-
ing over considerable periods of time, whereas normally the emission
of light by the firefly is an intermittent process of short duration,
partaking of the nature of a luminous explosion.

It is an interesting fact that the majority of poisons act as transient
stimuli, and that but few of these are so suddenly destructive in their
action on the photogenic material as to prevent the production of
light altogether. Of all substances thus far investigated sulphur
dioxide seems to be the most toxic in its effect on the photogenic
material. Greatly diminished atmospheric pressure invariably results
in the emission of light both by the living firefly and by the detached,
resting, luminous organ. Lastly, the photogenic material which has
been dried im vacuo over sulphuric acid retains its power to emit
light by moistening with water for a period of thirteen months and
perhaps even for longer intervals. ,

BIBLIOGRAPHY

1 Artaup: Annales maritimes et coloniales, 1825, p. 364; Bulletin des sciences
mathematiques, 1826, vi, p. 129; Journal fiir Chemie und Physik, 1828, lii, pp.
319-323.

2 BiscHoFF: Journal fiir Chemie und Physik, 1823, xxxix, pp. 259-305.

3 BoncarptT: Zeitschrift fiir wissenschaftliche Zodlogie, 1903, Ixxv, pp. I-45.

4 Carrapori: Annales de chimie, 1798, xxiv, pp. 96-101; Annalen der Physik,
1799, i, pp. 205-213; Philosophical magazine, 1708, ii, pp. 77-80; Giornale di
fisica, chimica e storia naturale, 1808, i, pp. 269-280.

5 CARRADORI: Giornale di fisica, chimica e storia naturale, 1809, li, pp. 247-264.

6 Carus: Comptes rendus de l’Académie des sciences, 1864, lix, pp. 607-608.

8 Dusotis: Bulletin de la société zodlogique de France, 1886, xi, pp. 1-275.

® DuBois: Comptes rendus de l’Académie des scfences, 1887, cv, pp. 690-692.

10 Dusors: Comptes rendus de |’Académie des sciences, 1896, cxxiii, pp. 653-
654.

1 EHRENBERG: Abhandlungen der kéniglich-preussische Akademie der Wis-
senschaften, Berlin, 1834, pp. 411-575.

2 FarapAy: Life and letters by Bence-Jones, i, pp. 90, 91, 125, 141,142, 144-146.

13 Forster: Observations sur la physique, 1783, xxiii, p. 24; Magazin oe
Wissenschaften und Literatur, 1783, Pt. ii, p. 281. cay

144 GIESBRECHT: Mittheilungen aus der zodlogische Station der Neapél, oe
1895, xi, pp. 648-690.

150 Joseph H. Kastle and F, Alex. McDermott.

18 GrotrHuss: Annales de chimie, 1807, lxiv, pp. 19-41 (38-40).

16 HEINEMANN: Archiv fiir mikroskopische:Anatomie, 1872, vili, pp. 461-471
(see also Archiv fiir die gesammte Physiologie, 1873, vii, pp. 365-366; Archiv.
fiir mikroskopische Anatomie, 1886, xxvii, pp. 296-382; Pyrophores, Vera Cruz,
1887).

1” HEINRICHS: Die Phosphorescenz der Kérper, Niirnberg, 1820.

18 HENNEGUY: Comptes rendus de la société de biologie, 1888, v, vili ser., pp.
707-708.

19 Hotper: Living lights, Chas. Scribners’ Sons, N. Y. and Lond., 1887,
8vo.

20 Humporpt: Tableaux de la nature; Paris, 1851, ii, pp. 60 ef seq.

21 JOUSSET DE BELLESME: Journal de l’anatomie et physiologie, 1880, xvi, pp.
121-169; Comptes rendus de l’Académie des sciences, 1880, xc, pp. 318-321;
Revue médicale de France et étranger, 1880, i, pp. 287-201.

2 Tves and CoBLENTz: Bulletin of the Bureau of Standards, Washington,
IQIO, Vi, pp. 321-336.

23 KOLLIKER: Verhandlungen der Wiirzburg physikal-medicinische Gesell-
schaft, 1857, viii, pp. 217-224; Journal of microscopical science, 1858, vi, pp.
166-173.

24 McDermott: Popular science monthly, 1910, lxxvii, pp. 114-121.

2 Macarre: Annales de chimie et de physique, 1821, xvii, pp. 151-167; Anna-
len der Physik, 1822, lxx, pp. 265-280; Quarterly journal of science, 1822, xii,
pp. 181-182; Journal de physique, 1821, xciii, pp. 46-56.

26 MACARTNEY: Philosophical transactions of the Royal Society, 1810, c, pp.
258-293; Annalen der Physik, 18109, lxi, pp. 1-35; Journal of natural philosophy,
chemistry, and the arts, 1811, xxviii, pp. 41-55; Journal fiir Chemie und Physik,
1814, X, Pp. 409-444.

*7 Mrven: Novorum actorum academize cesaree Leopoldino-Caroline Ger-
manice nature curiosum, xvi, Supplement.

28 MICHAELIS: Ueber das Leuchten der Ostsee, Hamburg, 1830.

29 Mitne-Epwarps: Lecons sur l’anatomie et la physiologie comparée de
Vhomme et des animaux, 1863, viii, lec. 68, pp. 83-120.

%° Pancert: Annales des sciences naturelles, 1872, xvi, v ser., pp. 1-67.

31 PETERS: Journal of experimental zodlogy, 1905, lix, pp. 235-242.

8 PraFF: Journal fiir Chemie und Physik, 1828, lii, pp. 311-321.

38 PrLUcer: Archiv fiir die gesammte Physiologie, 1875, xi, pp. 222-263 (see
also Ibid., 1875, X, pp. 251-367).

34 Quoy and GarmarD: Annales des sciences naturelles, 1825, iv, p. 8.

% SPALLANZANI: Memoria della Societa Italiano, Verona, 1794, vii, pp. 271-
290.

6 SPALLANZANI: Chimico esame degli esperimenti del Sig. Gottling, Professor
a Jena, sopra la luce del fosforo, etc., Modena, 1796.

87 'TIEDEMANN: Physiologie de homme, Paris, 1831, ii, pp. 516-551.

88 Titesius: Neuer Annalen der Wetteranischen Gesellschaft fiir die gesammten
Naturkunde, 1818, iv, p. 74.

Observations on the Production of Light by the Firefly. 151

39 Topp: Quarterly journal of science, literature, and the arts, 1824, xvii, p. 269;
Ibid., 1826, xxi, p. 241.

49 TowNSEND: American naturalist, 1904, Xxxviiil, pp. 127-151.

41 Wataseé: Biological lectures, Wood’s Hole, 1895, pp. 101-118.

4 WartaseE: Biological lectures, Wood’s Hole, 1898, pp. 117-192.

43 Acassiz: Memoirs of the American Academy of Arts and Sciences, 1874, x,

PP. 357, 398.

ACAPNIA AND SHOCK.—VII. FAILURE OF THE
CIRCULATION.

By YANDELL HENDERSON.
[From the Physiological Laboratory of the Yale Medical School.]

CONTENTS.

PaGE

I. What factor in the circulation really failsinshock? ...........- 15
Tf. ‘The veno-pressor mechanism. . . ~ 2.9. 2 se «2s oe rr 160
If. The development of acute oligemia ;. .. ~~. = = = =). soe 167
IV. Hyper-tonus of the heart in shock ..... ... «2 © «= s si see 171
V. Conclusions . 0 5). ses 2 ee le es se oe 172

TI. Waat Factor IN THE CIRCULATION REALLY FAILS IN SHOCK?

AILURE of the circulation is the commonest mode of death.
When the process is judged by the arterial pulse, there appears
to be a progressive weakening of the heart beat. Such a decline char-
acterizes the approach of the end after many abnormal conditions.
It often follows intense and prolonged pain. It may occur at the
height of an acute infectious disease. At first the rate of the heart
beat is rapid. Its amplitude diminishes, while arterial pressure is
nevertheless maintained at a normal level, or even above normal.
After the pulse has become ‘“‘thready,” arterial pressure sinks rapidly.
Thereafter the heart rate may be quick or slow; but unless it is
extremely slow the amplitude of the beats is small, and becomes
progressively smaller. Formerly this process was regarded as con-
sisting essentially in failure or fatigue of the contractile force of the
heart. Even to-day it is generally so denominated among clinicians,
— for lack of better terms and more critical conceptions of the factors
involved.

The later stages of this process are spoken of by surgeons as “‘shock.”
In their usage this term might be defined as a condition of the circula-
tion essentially similar to that induced by extreme hemorrhage, —
although there may have been no apparent loss of blood. When we

consider that ‘‘shock”’ means literally a jolt or jar or impact, and
1§2

Acapma and Shock. 153

that it was originally used to include concussion of the brain, it is
evident that in its modern sense the word is illogical and ought ulti-
mately to be superseded by a clearer term. It is, however, convenient
for the purposes of this paper to speak of the three most common
conditions terminating in a so-called “‘failing heart” as hemorrhagic
shock, traumatic shock, and toxemic shock. The terms imply that
all three conditions are fundamentally alike in their bloodlessness.

It is orthodox at the present time to say that traumatic and toxemic
shock are not due to a loss of power in the heart itself, but that these
conditions consist ,in a fatigue, or inhibition, or failure of some sort
in the vaso-motor centres. This doctrine was crystallized in its modern
form by two independent, yet nearly simultaneous, investigations,
— each a classic in its own field, — that of Crile ! on traumatic shock,
and that of Romberg and Passler? on toxemic shock. These two
pieces of work were nearly identical in methods of observation, in
the phenomena recorded, and in the conclusions reached, although
in one failure of the circulation was induced by trauma, and in the
other by bacterial toxins. Thus a single statement, as above, covers
the ‘‘vaso-motor failure” theory of both forms of shock. Is this
theory logical and true? To me it appears half true and half false.
Both investigations afford ample evidence that the heart does not
fail; but the same evidence proves no less conclusively, if the view
to be supported in this paper is correct, that the vaso-motor centres
are equally tenacious of their functional capacity.

The problem of the circulatory disturbances in shock has usually
been dealt with as if a demonstration that ‘the heart does not fail”’
were equivalent to a proof that ‘‘the vaso-motor nervous system does
fail.” * This alternative involves the assumption that there are two,
and only two, principal mechanical factors in the circulation, — the
contractions of the heart and the peripheral resistance of the arterioles
controlled by the vaso-motor nervous system. This assumption is, I
believe, an error of the greatest importance both theoretical and

1 CritE: Surgical shock, 1899; also Blood pressure in surgery.

2 RomBERG and PAssLeR: Deutsches Archiv fiir klinische Medicin, 1899, lxiv,

052.
‘ 2 The literature of toxaemic shock was recently reviewed and discussed in a
paper read before the Section of Medicine of the American Medical Association,

June 8, 1910, by W. G. Macattum. It will soon be published in the Journal of
the American Medical Association.

154 Yandell Henderson.

practical, but an error for which neither Crile nor Romberg and
Passler were responsible. They applied to their problem that con-
ception of the circulation which the physiology of to-day teaches.
This was their sole mistake. Both Crile and Romberg and Passler
concluded (correctly, I believe) that in shock the circulation fails in
the same manner as after hemorrhage, and that the heart fails be-
cause too little blood is supplied to it through the veins. Both found
that intra-venous infusion restored for a time a normal arterial pres-
sure and heart action. And unfortunately both labelled this true
picture with the misleading formula, — the only formula for it offered
by current physiology, — “‘vaso-motor failure.”

Present knowledge regarding the vaso-motor nervous system indi-
cates that its control is exercised — mainly at least — upon the finer
branches of the arterial system. It regulates the frictional resistance
to the outflow of blood from the arteries into the capillaries. Upon
the latter vessels it has, so far as known, no direct influence. A
relaxation of the arterioles of any region induces indirectly a mechani-
cal distention, not direct active dilatation, of the capillaries supplied.
Information regarding a central nervous control over the tonus of
veins is extremely meagre. In a later paper I shall attempt to show
that no direct central nervous control of the venous reservoirs exists.4
Now the failure of vascular tonus in traumatic and toxemic shock is
almost wholly in the venous system. Both Crile and Romberg and
Passler saw and emphasized this fact. It seems not to have occurred
to them that they were dealing with the failure of a mechanism as
yet unrecognized by physiology. In this they were not alone. For
half a century physiologists have been so dazzled by Claude Ber-
nard’s discovery of the vaso-motor nervous system that they have
neglected to emphasize the fact that the circulation must involve a
third factor in addition to the heart and the peripheral resistance of
the arterial system. Otherwise it would be as unstable as a stool
balanced on only two legs. It must include a mechanism, or mecha-
nisms, regulating the volume of the blood, and determining the venous
supply to the right heart. It is this veno-pressor mechanism, I be-
lieve, and neither the heart nor the vaso-motor nervous system, which
is the essential element in the failure of the circulation in shock.

* For a summary of the experiments on which this negation is based, see section
TT. on p. 161.

Acapma and Shock. 155

Crile’s view of traumatic shock has been almost universally accepted
by surgeons. The results of the work of Romberg and Passler have
not, at least in America, been adopted to an equal extent among
‘internists. This has been due, however, to neglect, and not to oppo-
sition. Janeway says that their “conclusions have, in my opinion,
never been controverted. . . . They clearly demand that we shall
in most cases abandon the idea of cardiac death at the height of acute
infectious diseases, such as pneumonia, typhoid fever, diphtheria,
and the septic fevers; though sudden death during convalescence may
be due at least in part to the later development of lesions in the
heart muscle. In place of heart failure, we must write vaso-motor
failure, or collapse . . . the heart stopping only because so little
blood is returned to it.”” Similarly Crile ® says that ‘‘the vaso-motor
centre in a state of shock may be designated as paralyzed, or exhausted.
The bulk of the blood does not circulate freely through the arterial
system. A large portion accumulates in the venous trunks, a state
equivalent to an intra-venous hemorrhage. The arterial system
bleeds into the dilated venous system. The symptoms closely mimic
those of hemorrhage. Most clinicians concede that differentiation
is impossible.”

A peculiar mingling of truth and error is contained in these quota-
tions. They represent shock as identical with hemorrhage, and in
the same sentence as identical also with failure of the vaso-motor
nervous system. The expression “‘a hemorrhage into the veins”
has become widely current among surgeons, yet it is a complete con-
tradiction in terms. It is illogical to speak of any condition as being
‘a hemorrhage’’ if it involves an increased flow of blood into the
veins, for hemorrhage consists essentially in loss of blood from the
veins. A diminished venous stream to the right heart, an incomplete
diastolic filling of the ventricles, and a consequent reduction in the
volume of the systolic discharges, — these are the immediate results
of a lessening of the volume of blood in the body.’ As hemorrhage

5 JANEWAY, T. C.: New York medical journal, Feb. 2, 1907.

6 CritE: Shock and collapse, KEEN’s Surgery, 1906, i, p.g29. (The quota-
tion is abbreviated.)

7 A paper presenting in a lucid manner the mechanics of hemorrhage was
read before the Section of Pathology and Physiology of the American Medical
Association, June 7, 1910, by CARL J. WiccERs. It will soon be published in the

156 Yandell Henderson.

progresses, the amplitude of the pulse is correspondingly diminished;
but until the output of the heart becomes extremely small, arterial
pressure is maintained by a compensatory constriction of the arterioles
by the vaso-motor nervous system. It makes no difference whether
the bleeding vessel be vein or artery, the fundamental condition from
which the fall of arterial pressure finally results is the depletion of the
venous reservoirs and an insufficient supply to the right heart. On
the contrary, the expression ‘‘a hemorrhage into the veins” implies
an altogether different and — unless the aorta itself be severed —
an erroneous conception of the mechanics of hemorrhagic shock. It
suggests that the easier egress of the blood through the cut artery is
the direct cause of the fall of pressure in the arterial system. It
suggests also that the condition of the circulation after hemorrhage
is similar to a general relaxation and dilatation of all the arterioles
of the body, such as results from severing the spinal cord just below
the bulb. In point of fact, however, when this operation is performed
upon a cat, although arterial pressure may sink to 60 mm. Hg or
even much lower, because of the diminution of the peripheral resist-
ance in the arterial system, the amplitude of the pulse is not reduced.
It may even be increased. The tonus of the venous reservoirs is not
affected. Hence the blood continues to flow to the right heart in un-
diminished volume; and the heart, being thus filled to its normal
capacity during diastole, discharges its full stroke during systole.
Clear evidence that fall of arterial pressure after haemorrhage is due
to diminished venous supply to the right heart and not to abolition
of peripheral resistance in the arterial system is afforded by the fact
that no considerable fall of arterial pressure occurs if the blood be
re-injected into a vein as fast as it runs out of a severed artery. There
are more points of difference than of similarity between the experi-
mental condition of ‘‘spinal shock” (z. e., true vaso-motor failure)
on the one hand and traumatic, toxemic, and hemorrhagic shock on
the other.

It is so easy to record arterial pressure and so difficult to measure
the minute-volume of the arterial blood stream that one is inclined
to forget that the pressure in the arteries is really a phenomenon of

Archives of internal medicine. See also voN DEN VELDEN, R.: Archiv fiir ex-
perimentelle Pathologie und Pharmakologie, 1909, lxi, p. 37 (bibliography on
hemorrhage.)

Acapnia and Shock. ee =

only secondary importance. Because of the technical perfection of
the method introduced by Ludwig we are prone to lose touch with
the great doctrine of Harvey. The primary function of the circulation
is the volume of blood pumped onward by the heart in unit time. It
is not the pressure of the blood within its vessels which keeps the
tissues alive, but the quantity of oxygen and other nutriment which
the stream supplies. Above a moderate degree arterial pressure is
an expense rather than an asset. Within wide limits a high arterial
pressure is compatible with a diminished stream, and a low arterial
pressure with a normal flow. Thus it is misleading to say that “an
abnormally low [arterial] blood pressure is the essential phenomenon”
of traumatic shock.* The essential arterial condition in shock is the
greatly diminished stream which both Crile and Romberg and Passler
described. This reduced blood flow results in, instead of resulting
from, lowered arterial pressure. Arterial pressure falls, not as
in ‘‘spinal shock” because the peripheral resistance is diminished,
but because of the fact that when the volume of the stream has
dwindled beyond a certain limit the utmost activity of the vaso-motor
centres fails to afford a peripheral resistance sufficient to maintain
the pressure.

Both Crile and Romberg and Passler laid great stress on the fact
that stimulation of a sensory nerve fails to induce in a shocked sub-
ject that reflex rise of arterial pressure which always occurs in a
normal animal. They interpreted this as clear evidence of lessened
functional capacity in the vaso-motor nerve centres. I believe, on
the contrary, that this observation (which I have repeatedly verified)
really indicates that in shock the vaso-motor centres are in a condi-
tion of maximal activity. They are already constricting the arterioles
as much as possible in the effort to compensate for the lessened out-
put of the heart and to maintain arterial pressure. This is the normal
function of these centres.? When a sensory nerve is stimulated in
shock, no increase in the constriction of arterioles results, because
they are already constricted to the utmost.

Both Crile and Romberg and Passler found that even in profound
-shock arterial pressure returns to a normal level when the venous
supply to the right heart is restored to normal volume by an intra-

8 CrILE: KEEN’s Surgery, 1906, i, p. 926.
® Cf. PorTER and Quinsy: This journal, 1908, xx, p. 505.

iho ° Yandell Henderson.

venous infusion of saline. They interpreted this (correctly, I believe)
as proof that the heart is still functionally capable. But does not
this experiment prove also, and with equal conclusiveness, that the
peripheral resistance of the arterial system and the nervous mechanism
controlling this resistance are likewise still functionally capable? If
the vaso-motor nervous system were inactive and the arterioles relaxed,
the blood would run out through the capillaries too easily for any
pressure to be developed. Think of the tissues of the body as similar
to a house afire. Think of the heart as a steam fire-engine; of its
arteries as lines of hose; of the arterioles controlled by the vaso-motor
nervous system as the nozzles of the hose. The engine can throw no
stream upon the fire, nor maintain a pressure in the hose, even though
the nozzles be constricted to the utmost, unless the supply of water
from the cistern or the street water main be adequate. Similarly,
arterial pressure cannot be maintained merely by the heart and the
peripheral resistance of the arterial system. As well might one think
of lifting a weight by means of a lever with no fulcrum! The heart
can discharge during systole only so much blood as distends its cham-
bers during diastole. The diastolic filling of the right heart depends
upon the volume of the stream flowing to it through the veins and
upon the distending pressure which this stream affords. Venous pres-
sure ts, so to speak, the fulcrum of the circulation.

The object of this paper and of others to follow is to show that
shock, as surgeons use the word, is due to a failure of this fulcrum.
Because of the diminished venous supply the heart is not adequately
distended and filled during diastole. Hence the picture of a ‘failing
heart’’ revealed by the pulse. For the same reason arterial pressure
ultimately sinks in spite of an intense activity (not because of failure)
in the vaso-motor nervous system, and in spite of an extreme con-
striction (not because of relaxation) of the arterioles. Finally the
blood stream is so much diminished that it is inadequate to supply
oxygen to the tissues, and death quickly ensues.?°

The distinction between the functions of the vaso-motor and of the veno-:
pressor mechanisms is illustrated in Fig. 1. In this diagram a force pump
(A) is placed at the bottom of a pit or well. These conditions correspond
to the heart working under the negative pressure of the thorax. The top

The fatal process is exemplified in the experiments described in section III,
on p. 167 et seq.

Acapnia and Shock. 159

of the pit is zero pressure. If the depth of the pit is varied rhythmically, as
shown by the dotted line below the pit, the principal influence of the respi-

Ficure 1.— A diagram illustrating the
functions of the three essential factors
in the circulation. It shows the heart
(A) as a force pump placed at the bot-
tom of a pit to correspond with the neg-
ative pressure of the thorax. The top
of the pit (OO) is taken as the level of
zero pressure. The pump discharges
into an air chamber (B) affording an
elasticity like that of the arteries, and
against the peripheral resistance of sev-
eral small nozzles (C), which play the
part of arterioles. The vaso-motor ner-
vous system (not shown in the diagram)
controls, the calibre or lumen of these
nozzles, and thus regulates the arterial
pressure and the relative volume of the
various jets. The jets fall into tanks or
reservoirs (D), corresponding to the va-
rious capillary areas of the body (e. g.,
the intestine, skeletal muscles, head,
etc.), from which a system of drains ‘or
veins supplies the pump. The pump ex-
erts no suction, but is filled during dia-
stole by the venous pressure pushing the
plunger upward. In systole the force
represented by the arrow drives the
plunger down. For discussion see small
type in text.

ratory movements upon the circula-
tion is thereby imitated. The pump
is incapable of exerting any suction
whatever. Its plunger is not drawn up
actively, but is pushed up by the ve-
nous inflow during the diastolic filling
of itschamber. In systole it is driven
down actively by the force indicated
by the arrow. An air chamber (B)
affords the arterial elasticity. The
small nozzles (C) at the ends of the
hose pipes, or arteries, play a part
analogous to the vaso-motor mecha-
nism. The height to which the jets
from these nozzles rise is an expres-
sion of the arterial pressure. A
narrowing of the nozzles, or vaso-
constriction, will raise arterial pres-
sure; a dilatation of the nozzles will
lower it. Dilatation of one and con-
striction of others will vary the distri-
bution of the stream. But so long as
the pump acts efficiently and with a
full venous supply, the total volume of
the aortic stream will be the same, no
matter whether the nozzles be con-
stricted to induce arterial hyperten-
sion or dilated to the utmost by
vaso-motor failure. In this scheme
vaso-motor failure cannot induce
stagnation.

The jets fall into tanks (D)," corre-
sponding to the capillaries of the
various organs of the body. From
these reservoirs large drains unite to
form the ‘venous system supplying
the pump. A hemorrhage may be
produced in one or other of two

1 This is the “pre-ventricular reservoir’ of v. RECKLINGHAUSEN: Archiv fiir
experimentelle Pathologie und Pharmakologie, 1906, lv, p. 476, and 1907, lvi, p. 1.

160 Yandell Henderson.

ways, — either by diverting one of the jets from its tank, thus imitat-
ing a hemorrhage from an artery, or by opening the cock on the
tank at the left, thus causing a hemorrhage from a vein. ‘The effects of
these two procedures will be exactly the same. The volume of liquid
stored in the reservoirs will be diminished. Neither the arterial pressure
nor the arterial stream will be influenced until the supply in the venous
reservoirs is considerably depleted. When the supply no longer affords a
sufficient venous pressure to distend the heart (7. e., to push up the plunger
with normal rapidity and to a normal extent during diastole), then the out-
put of the pump will be diminished. If the nozzles are now actively con-
stricted and the peripheral resistance is thus sufficiently increased, arterial
pressure may, however, be maintained at a normal height for a little while
longer; but the arterial pulse will be “ thready.”” Finally, when the limit of
such vasomotor compensation is reached, and the nozzles are as small as
they can be made, a slight additional diminution in the output of the pump
will involve a sudden and extensive fall of arterial pressure.

After the arterial pressure has been thus lowered by extreme hamor-
rhage, it may be restored by pouring liquid into the venous reservoirs. If
it is restored, this fact affords proof both that the pump has not broken
down and that the nozzles are not relaxed. Neither cardiac failure nor
vaso-motor failure has occurred. It is evident that to speak of a ‘* hemor-
rhage into the veins because of vaso-motor failure” would be, at least in
this scheme, to confuse two distinct conditions. The only conditions which
will accurately simulate the phenomena of hemorrhagic, traumatic, and
toxeemic shock are such as will induce a failure in the supply and pressure
from the venous reservoirs to the heart.

II. THe VENo-PRESSOR MECHANISM.

The nature of the veno-pressor mechanism is a peculiarly difficult
experimental problem. From such data as I have been able to col-
lect during the past five years it appears to consist of two sets of
factors. One is tonic; the other osmotic. Both depend upon the
maintenance of a normal tension of CO, in the fluids of the body and
of a normal alkali-acid equilibrium in the tissues.” The respiratory
centre, by regulating the CO, content of the arterial blood within
narrow limits of variation," exerts an indirect but powerful control

22 HENDERSON, L. J.: Ergebnisse der Physiologie, 1909, vill, p. 254.
8 HALDANE and PriestLEy: This journal, 1905, xxxii, p. 252.

Acapma and Shock. 161

over the veno-pressor mechanism. Any considerable accumulation
of CO, above normal augments the venous pressure. Excessive
pulmonary ventilation tends to lower it. Acute acapnia diminishes
the volume of the blood as effectively as does an extensive hmor-
rhage. This condition of bloodlessness, or oligeemia, or exsanguinity
results from the passage of fluid from the blood vessels out into the
tissues.

The details of the experiments carried out in this laboratory upon
the veno-pressor mechanism will be presented in a later paper. Neither
the tonic nor the osmotic element in the mechanism is subject to any
direct control by nerve centres, either spinal or bulbar. Adrenalin
exerts no direct influence upon it. These statements are based upon
five sets of experiments which may be summarized as follows: ™
(1) After section of the vagi, vigorous stimulation of the splanchnic
nerves causes a considerable rise of arterial pressure, but produces a
barely perceptible effect upon the pressure in the systemic veins.
(2) Injection of adrenalin, after vagus section, has likewise an insig-
nificant influence upon venous pressure, unless the dosage is so large
as to raise arterial pressure to the point at which the heart’s action is
interfered with. (3) Section of the spinal cord just below the bulb
causes a marked fall of arterial pressure, but no immediate drop, on
the contrary sometimes a rise, in venous pressure. (4) Stimulation
of the severed cord, under curare, restores arterial pressure to a normal
level and maintains it at this level so long as the stimulation is con-
tinued. Simultaneously there is either no rise of venous pressure or
a slight rise and immediate relapse even during continued stimulation.
(5) When a cat is decapitated by the method of Sherrington ® and
is supplied with a jet of oxygen into the trachea by the method of
Volhard,® CO, accumulates in the body. If the stream of oxygen
is slow, a gradual rise of venous pressure develops. Arterial pressure
continues unaltered; and until a high degree of hypercapnia has
developed the heart rate remains unaffected. If the gas jet into the
trachea consists of oxygen containing a large percentage of CO, the

M4 Partly quoted from a preliminary report presented by the writer before the
American Physiological Society in December, 1908, This journal, 1909, xxiii,
p. xxx. Cf. Kaya and Stariinc: Journal of physiology, 1909, xxxix, p. 347.

15 SHERRINGTON, C. S.: Journal of physiology, 1909, xxxvili, p. 375.

7 VoLHARD: Miinchener medizinische Wochenschrift, 1908, No. 5.

162 Yandell Henderson.

rise of venous pressure develops in a few minutes. The venous pres-
sure falls again in the course of a few minutes if artificial respiration
is supplied. It falls also, but less rapidly, if the lungs are ventilated
with a rapid stream of oxygen alone or with air (as in the Meltzer 1”
insufflation method). This phenomenon of hypercapnial venous
hypertension affords a convincing demonstration of the existence of
the veno-pressor mechanism. It appears to be a crucial experiment
for this mechanism very much as section and stimulation of the cervi-
cal sympathetic nerve in the rabbit were crucial for the recognition
of the vaso-motor nervous system.

The tonic element in the veno-pressor mechanism consists appar-
ently in the tonus of the tissues rather than in that of the walls of the
veins, — at least my experiments on excised veins have yielded
inconclusive er negative results. In the third paper of this series ®
the powerful” influence of CO. upon the tonicity of the intestine was
described. The experiments of Lee’? have shown a similar relation —
of CO,.to the tonus (or Treppe) of skeletal muscle. The increase of
tissue tonus induced by hypercapnia squeezes the blood out of the
capillaries. In the intestine it can be seen to render the tissue paler.
Local acapnia induced by exposure of the intestine to a stream of
warm moist air rapidly results in an intense congestion and stasis.
It is noteworthy that Romberg and Passler found that animals in
toxeemic shock exhibited a marked rise of arterial pressure when
partially asphyxiated, — a rise in some cases equivalent to that fol-
lowing a liberal infusion of saline. This effect may fairly be attributed
to the veno-pressor mechanism and not to the vaso-motor, since arte-
rial pressure was scarcely at all affected by vigorous stimulation of an
afferent nerve. It is a fact which doubtless many investigators have
observed that when an animal has been under artificial respiration
for a considerable time and the circulation has begun to fail, a
marked and lasting improvement results from a period of asphyxia.
The well-known effects of asphyxia upon the normal circulation
are in part due to the influence of hypercapnia upon the veno-pressor
mechanism.

Mettzer and Aver: Zentralblatt fiir Physiologie, 1909, xxiii, pp. 210 and
442; also Brept and RoTHBERGER, [bid., p. 327.

18 HENDERSON, Y.: This journal, 1909, xxiv, p. 66.

19 Ler, F. L.: This journal, 1907, xviii, p. 267.

Acapma and Shock. 163

The osmotic element in the veno-pressor mechanism appears to be
even more important than atonicity in the failure of the circulation
in shock. The blood does not merely stagnate in the veins and capil-
laries. Ample and conclusive evidence demonstrates that both in
traumatic and toxemic shock the volume of the blood in the vessels
is greatly diminished. The balance of osmotic forces, the colloidal '
imbibition of the proteins, and the other physico-chemical conditions
determining the passage of fluid back and forth through the walls
of the capillaries are upset. As a result, the distribution of water
between the blood, the lymph, and the cytoplasm of the tissues is
altered. In traumatic shock the excessive passage of liquid from the
blood into the tissues is, I believe, initially induced by acapnia. In
toxeemic shock it appears to depend either upon acapnia or upon
some similarly acting disturbance of chemical conditions in the tissues
(e. g., the protein-alkali-acid-CO, equilibrium of L. J. Henderson,
loc. cit.) involved in the febrile process.

Forty years ago H. Fischer ?° noted that the condition of the cir-
culation in profound traumatic shock is essentially like that in the
last stage of cholera. In cholera the volume of the blood is enor-
mously diminished; and intra-venous infusions of normal saline are ©
of very marked temporary benefit. The fluid is, however, soon lost
from the blood vessels into the intestines or into the tissues. Recently
Rogers 7! has demonstrated a loss of 64 per cent of the serum of the
blood in the toxeemic shock of cholera. The degree of prostration of
the patient he finds to be proportional to the diminution of the blood
volume. In the blood remaining in the vessels the chlorides are so
much reduced that intravascular hemolysis sometimes occurs. Intra-
venous infusions of hypertonic saline have lowered the mortality in
the shock stage of cholera by 50 per cent.

Many years ago Leyden” observed a relative retention of water
during fever, —7. e., an increase in the percentage of water as com-
pared with the solids in the tissues. Since then a literature too large
for more than a brief citation here has gathered about the topic.

® FiscHEeR, H.: Ueber dem Shok, in Vorkmann’s Sammlung klinischer Vor-
trage, 1870, No. 10.
. | Rocers, L.: Philippine journal of science, 1909, iv, p. 99; also Proceed-
ings of the Royal Society, 1909, Ixxxi, B, No. 548.
# LEYDEN : Archiv fiir klinische Medicin, v, p. 366.

164 Yandell Henderson.

Krehl * says: ‘‘Why the patient with fever fails to excrete the
extra liquid in his tissues is not known, though we suspect that it is
because the extra water is retained chiefly within the cells themselves
and does not reach the blood or lymph. Possibly the physico-chemical
properties of the cells or their secretory activities are changed.”
Parallel with such tissue changes a concentration of the blood has
frequently been noted both in patients and in experiments on animals.
Among others Sandelowsky* has recently noted it in pneumonia;
and Oppenheimer and Reiss have found it to occur in scarlet fever. As
a rule, studies upon this topic do not apply directly to the problem of
fever collapse, since at the time the observations were taken the
subjects were merely ill, not dying in shock. They are sufficient,
however, to indicate that some degree of oligemia is frequently, if
not always, an accompaniment of toxemia.”

Sherrington and Copeman” have shown that in traumatic shock
the specific gravity of the blood is markedly increased even before
the arterial pressure has fallen to a low level. Mummery,?’ who is
an adherent of the ‘‘vaso-motor failure” theory, has found that in
spinal shock the specific gravity is distinctly diminished.”* The only
significance which the latter observation bears for the problem of
traumatic shock is the additional evidence which it affords that
traumatic and spinal shock are two totally distinct conditions.

Nearly all modern writers on traumatic shock have recognized more
or less explicitly the occurrence and the importance of oligemia. It
was, however, a long step forward when J. D. Malcolm” first boldly
supported this idea to the exclusion of vaso-motor failure. He has
pointed out the illogical character of the arguments for the ‘‘vaso-
motor failure”? theory. He bases his argument upon keen and exten-

*3 KREHL, L.: Pathologische Physiologie, fourth edition, translation by HEw-
LETT under the title Clinical pathology, second edition, 1907, p. 422 (bibliography).

24 SANDELOWSKY: Deutsches Archiv fiir klinische Medizin, 1909, xcvi, p. 445 ;
OPPENHEIMER and ReEIss, same volume.

*» Acapnia is also an accompaniment. See Kreut, L.: Pathologische Physi-
ologie, third edition, 1904, pp. 459 and 475.

26 SHERRINGTON and CopEMAN: Journal of physiology, 1893, xiv, p. 52.

77 Mummery: Lancet, 1905, i, pp. 696, 776, 846.

28 Mummery and Symes: Journal of physiology,1907, xxxvi, p. Xv.

29 Matcorm, J. D.: Transactions of the Medical Society of London, 1909,
XXxii, p. 289; and Lancet, 1905, i, ii, pp. 573, 618, 737, 922; and 1907, i, p. 497.

Acapma and Shock. 165

sive clinical observations made especially during the early days of
antiseptic surgery. He has laid particular emphasis upon the point
that when saline is infused into the veins of a subject in shock the
increase in the volume of the blood and the improvement of the
heart’s action are only temporary. The fluid is rapidly transferred
into the tissue spaces, into the pleural cavity, etc., and the heart is
then again insufficiently supplied with blood from the veins. The
first section of this paper is in part a reproduction of Malcolm’s
reasoning.

Seelig and Lyon*® have recently rendered the great service of
bringing forward absolute proof that in traumatic shock the arterioles
are more than normally constricted, instead of being relaxed. They
have shown that the strength of the impulses transmitted from the
centres over the vaso-motor nerves is greater, instead of being less,
than under normal conditions. In Professor Lyon’s laboratory experi-
ments have recently been carried on which demonstrate conclusively
that in shock the volume of blood in nearly all of the important organs
of the body is greatly diminished.*!

As to the etiology of the exsanguinity of shock Malcolm has sug-
gested that the intense vaso-constriction induced by pain squeezes
the fluid of the blood out through the walls of the vessels into the
tissue spaces. This idea is in general accord with the views regarding
lymph formation expressed by so competent an authority as Starling.”
Nevertheless I cannot agree in this simple mechanical explanation of
the edematous process in shock. If the high pressure causes exuda-
tion, the low pressure of profound shock should involve re-absorption.
Lymph is not a mere filtrate from the blood. The increase in the water
content of the tissues in shock is probably intra-cellular quite as
much as it is inter-cellular or lymphatic. I believe that it depends
upon chemical rather than upon physical conditions.

The retention of chlorides in the acute infectious diseases liable to
terminate in shock is probably associated with the oligeemia of these

30 SEELIG and Lyon: Journal of the American Medical Association, 1909, lii,
DP. 45.

31 The results of thése experiments were exhibited at the last meeting of the
American Medical Association, June, 1909.

% STARLING, E. H.: Herter lectures (N. Y., 1908) on the fluids of the body
(published in Chicago, 1909).

166 Yandell Henderson. ‘

fevers. In passing out of the blood into the tissues the chlorides
carry water to dissolve them, and thus induce a diminution of the
volume of the blood. Important in this connection is the single obser-.
vation of Hamburger * recently extensively repeated and verified by
Luckhardt,* that the CO, content is ordinarily less, and the chlorine
content greater, in lymph than in blood. Luckhardt concurs in Ham-
burger’s suggestion that the lymph gives up the CO,-ions, which it
has received from the tissue cells, to the blood in exchange for twice
the amount of Cl-ions. This mechanism offers a simple explanation
for the retention of chlorides and the development of oligaemia under
the influence of acapnia. When the CO; content of the blood is dimin-
ished by excessive pulmonary ventilation, the more rapid passage of
CO,-ions out of the tissues into the blood involves a correspondingly
increased passage of Cl-ions out of the blood into the tissues. The
increased osmotic pressure thus developed in the tissue fluids with-
draws water from the blood and induces oligemia. Somewhat the
same process appears to be in part responsible for the policythemia
occurring coincidently with acapnia under lowered barometric
pressure.”

This mechanism does not, however, afford a complete explanation
of that crisis in the development of shock which Crile’s experiments
have shown to be a characteristic feature of failure of the circulation.
Up to a certain point the failure is gradual. The amplitude of the
pulse diminishes, Arterial pressure may fall somewhat, or may be
maintained. Finally, however, a time arrives when within a few
minutes the pressure drops to an extreme low level. Up to this point
intra-venous infusion of saline is retained within the circulation fairly
well, and the subject is usually capable of recovery. Thereafter the
subject is irrecoverable; for saline infusions pass out of the circula-
tion into the tissues nearly as rapidly as they can be injected into a
vein,

The experiments to be reported in the next section appear to afford

8 HAMBURGER, H. J.: Zeitschrift fiir Biologie, 1893, xxx, p. 142; Beitrige zur
pathologischen Anatomie und zur allgemeinen Pathologie, 1893, xiv, p. 448;
and Osmotischer Druck und Ionenlehre, 1904, ii, p. 54.

*4 LucKHARDT, A. B.: This journal, roro, xxv, p. 345.

© See Mastnc and Morawitz: Hohenklima und Blutbildung, Deutsches
Archiv fiir klinische Medizin, 1910, xcviii, p. 301. On acapnia and increased
hemoglobin, see Warp, R. O.: Journal of physiology, 1908, xxxvii, p. 378.

Acapnia and Shock. 167

an adequate explanation of this crisis. They show that in traumatic
shock acapnial exsanguinity develops gradually until the arterial
blood stream is diminished to the point at which it is insufficient to
supply the oxygen needed by the tissues. The venous blood at this
time, therefore, contains little or no oxygen. The blood stream of
normal life has a two thirds factor of safety; in other words, the venous
bloed normally contains about two thirds as much oxygen as does the
arterial.*® In the absence of an adequate supply of oxygen a con-
dition of asphyxial acidesis rapidly develops in the tissues. The acid
substances thus formed cause the proteins of the tissues to imbibe
water from the blood in much the same manner that fibrin swells in
dilute acids. Hence the fatally rapid transudation of fluid from the
blood vessels.”

Thus acute oligemia develops in two stages. The foregoing para-
graphs suggest their nature. What is here said is, however, merely
a superficial and approximate preliminary statement of the hypothe-
sis of acapnial oligemia. It is of course very far from being the whole
story of the complex processes probably involved.

Ill. Tot DEVELOPMENT OF ACUTE OLIGZMIA.

Experiments were performed on twenty dogs. Some were mor-
- phinized and initially anesthetized with chloroform. The majority
were merely etherized. All were tracheotomized. Arterial pressure
was recorded by means of a Hiirthle manometer connected with the
carotid artery. Respiration was recorded by means of a spirometer
connected with the trachea for a few seconds at a time during the
observation periods. Samples of arterial blood were drawn from the
femoral artery. Venous samples were taken from the right heart by
means of a sound inserted through the right jugular. They were in
all cases 3 c.c. each and were analyzed by the method of Barcroft
and Haldane.*®

After a preliminary normal period the abdomen was opened widely

36 The same factor of safety appears to determine circulatory failure or recov-
ery in CO poisoning. See RincER, A.I.: Proceedings of Society for Experimental
Biology and Medicine, 1909, vi, p. 68.

7 This idea was first suggested to me by a paper (q. v.) on the nature and cause
of edema by Martin H. FisHer: Journal of the American Medical Association,
1908, li, p. 830.

88 BARCROFT and HALDANE: Journal of physiology, 1902, xxviii, p. 234.

108 Yandell Henderson.

and the abdominal viscera handled continuously in a stream of warm
moist air. In some cases one of the sciatic nerves was exposed and cut,
and its central end occasionally stimulated either mechanically or
electrically. An intense congestion developed in the tissues and a
stasis in the veins of the exposed viscera. The mechanical irritation
of the intestines induced a continual hyperpnoea, a rapid heart rate,
and high arterial pressure. The amplitude of the pulse diminished
gradually to extreme threadiness. Such arteries as were exposed
became constricted to less than half their initial diameter. Often
the cannula used to draw blood from the femoral artery at the begin-
ning of the experiment could not be inserted after shock had devel-
oped; and a smaller cannula had then to be used. While the animal
was still in fairly good condition, the withdrawal of the blood samples
had no perceptible effect upon the arterial pressure; but after the
pulse had become very narrow, and even before the decisive fall of
pressure, the loss of even this small amount of blood produced a notice-
able depression. :

It usually required two or three hours to induce shock by the methods
employed. After the first hour it required watchfulness and a con-
tinual irritation of some afferent nerve to prevent the subjects from
relapsing into a fatal apnoea, as described in the fifth paper of this
series.*® A change in this respect was noticeable after venous anox-
hemia had developed. They then breathed rapidly even when irrita-
tion was discontinued. This hyperpnoea was probably due to the
presence in the blood of the acidosis substances resulting from tissue
asphyxia, since the CO, content of the arterial blood was only about
half the normal amount.

The results of the blood gas analyses show that the arterial CO:
content was usually diminished by the hyperpnoea to about one-
half the normal, and the venous CO: content to about two thirds
or three quarters of the normal. In some experiments the oxygen
content of the arterial blood increased noticeably ‘during the de-
velopment of shock. This is probably to be explained as due to the
relative increase of corpuscles and hemoglobin in unit volume of
blood induced by the passage of serum out of the vessels into the
tissues. The oxygen content of the venous blood diminished progres-
sively until it contained in some cases only a mere trace of oxygen.
Assuming that the oxygen consumption of the tissues continued the

39 HENDERSON, Y.: This journal, 1910, xxv, p. 395.

Acapma and Shock. 169

same as in the preliminary normal period, the degree of venous anox-
hemia affords an index of the diminution in the volume of the blood
stream. The arterial pulse record underwent a corresponding diminu-
tion in amplitude, but the systolic arterial pressure was usually well
maintained, until the period of approximately complete venous
anoxheemia (3 per cent of oxygen or less) was reached. From these
observations it appears that up to this point an increased vaso-motor
activity (i. e., augmented frictional resistance in the arterioles) com-
pensated the arterial pressure for the diminution in the stream dis-
charged by the heart because of the failing veno-pressor mechanism.

Up to the crisis of venous anoxhemia it was found to be possible to
induce a rapid recovery of the animals by the methods described in the
third paper of this series (Loc. cit.). They consisted in pouring saline
saturated with CO, into the abdomen and closing the cavity, infusing-
saline likewise saturated with CO, into a vein, and attaching a bag or
long tube to the trachea so that the animal partially rebreathed its
expired air. It was found to be safer to use oxygen for this rebreath-
ing, instead of air, since the subjects at this time were peculiarly sus-
ceptible to the ill effects of lack of oxygen.

After the crisis these measures were inadequate. Respiration was
indeed improved; and this benefit was retained so long as the re-
breathing was continued. Temporarily the arterial pressure and pulse
was likewise restored to normal height and amplitude. In the course
of a few minutes, however, they sank again to a mere flutter only a
little above zero pressure. By repeated and increasingly liberal in-
fusions the circulation could be maintained for a half hour longer;
but sooner or later death from circulatory failure resulted. In some
cases a quantity of saline equal to one fifth of the animal’s weight was
thus administered, — without saving the subject. At autopsy the
tissues were found to be edematous, the spleen distended, the arteries
empty, the veins not well filled, the left heart contracted, and the
right partially relaxed.

In Fig. 2 are reproduced the graphic records obtained in one of
these experiments. In the table (pp. 170-171) are summarized the
analytical and other data of twelve experiments illustrating the con-
ditions of tissue respiration under which shock does, and does not,
develop. When the mixed venous blood of the right heart contains
only 3 to 5 percent of oxygen, it is fair to assume a complete venous
anoxhemia in some organs, —e. g., the abdominal viscera.

“soy 9I1Y} J9}Je Yooys

Yandell Henderson.

ON ‘A[PAISSAOXa poyyeeIq IOAN 907 oot
“pow1oF
-o10yy9 pue paziurydiom Ajdoaq 09 cor
*yooys podojaaop-T]a GIG OF
‘pozturydiow | 06 SzI
*yooys Jo yuowdopaaop ur Ape q oy SIT
“pozyurydio yy no) SET
*yooys jo yuowdoyaAop ut Apreq OST 09
“WIIOJOIO]YD YIM
pozVVIUI visoyyseur ‘pozrurydioyyy SL val
‘oynurTU “BE “WU
, 59}0N god 9ye1 | oanssoid
ylvoyy eloqly

170

Ctr |+ OTT FOE
mag S°6 SSP
O'SF 8'8 6°62
S29 TOI V9
67S ZT V7
V6P VPI O1F
7°67 901 SH7
Dg w= SOK €°9S
%09 %0 i709)
“snous A,

"[euloyIy

*yuao Jad ‘sjoa soses pool,

8°C7
fcr
OT?
£02
9'VE
VIe
6°02
9°61

76)

‘sojdures
Pood

*S-LO6T
quoulLiedxa

jo 93eq

‘IHOIAM SOTTY OJ OL 9 0 SOOM “AAYAN OMLVIOS AHL IO NOILVIOWILG A GNV VUAOSTIA
TVNINOGaY FHL JO NOMVIIMA. GNV AXNSOdXY AT GAONGNI VONdUYAdMAP[ AHL AM HOOHS JO INAWAOTAATQ: AHL ([] AIdWvS) ONrana
GNV (J GIdWVS) TAOATA AAASSAAG IVIVALAY AHL ANV ‘SALVY ASTAG AHL ‘SASVS) GOOTGY SNONTA GNV IVIMALAY AHL ONIMOHS AAV],

‘I ATV

171

‘QUOTE JOYA JopuN oJaM spafqns 9y} poye}s VSIMIJOYy}O UIyM ydaoxq 7

O'S IV

‘Z “BLT 928 OEE IV
( g6¢ “d ‘Axx “QTOT
(‘{eumnol siyy, ‘f9) vaoude je}eq

Sule]

SUTIe,T

*ypoys dopAsp jou
Pip pure ‘AjaAIsseoxa 9Y}eaAIq 0}
pasnjol YOTYM [RUMI SNOIOSIA VW

yyeop rea

Acapnia and Shock,

*aye] soyn
-ulll Moy @ vaoude je}Z} Usppns

*yooys pedoyaaop-]fo
*pazimiydio jy

OLT
OL

027
CL

OOT
Sct
OST
set

9°60
6
ee

os
cSt
00

Ve
6'6¢
Por
VLE
6°02
VIZ
Toe
o'S¢
Ov
9°0P
V'9C
o'CP
oT
9 CP

9°S7
C07
SST
6ST
L6t

7 ARI

1z Ae

ST Av

* 1 ke

gy ‘dy

OT “APTN

0) a

"oz Ant

Yandell Henderson.

£72

“Aypides payrey ureSe oanssosd yerojze ‘sanssi} ay} 07U UOIsNjur ay} Jo WoLyepnsuvs} prides oy} 0} SuIMQ *AI04
-1suvI} AIOA ‘IOAOMOY, ‘SEA UOTYE[MIIIO dy} 07 WYoUeq oy, “UoKe[NoID pue uoljerdsar yoq Ut qUIUIAOICWI }eVII3 d]eIPIWUT 94} SMOYS
piooai oy, “Seq zoded v ut uoShxo oyyeoiq-a1 07 apeu sem qoaiqns oy} pue ‘u1eA rejnsnf oy} ofur payoefur A[Mojs o1aM auUITes pazeu
-oqieo Jo sazjowIyuad sIqnd AWYYy pur perpuny ouQ “Yyep Jo yuIod oy} yw sem yoolqns ay} pur ‘sdsed 07 podueyo Wt MON “MOT[eYS o10Ur
pure o1ow Suimoseq apm ‘prides ponunuoo pry uoresrdser ay} Mou [UL “SFT jo “uIUL QZ ATUO Jo ainsseid & 3¥ 19}}Np slow v sem asind ayy
jun ‘Aypidesr yay} W. ‘JaAeMOY ‘s9\JeoIOyT, “poysturmmpun ya sev svm oinssoid [ere Ysnoyye ‘QuIT} Sq} 7 as_nd oy} Jo ssouT[eUs 94}
dJON ‘“pasojd uawOpge oy} put paoridas aJaM BIBDSIA 9Y} MOTTE puodes oy} PY “BeoUdIodAY snonuyyUoS poonpurl SIyL ‘rojenb & pue moq
UL JO} Ie SIOW WILM JO UIvOI]s B UL poypULY o19M VIODSIA oY} Pue ‘pouado sem USWOpge 94} MOIIV JSG OY} TV “Ajoyeinooev UMOYS JIB
syyeorq oy} Jo sopnqzydue aarjeper oy} yey OS ‘JoJaWOIIds oIJEUUNJOA ev Aq poaps0dI uoneiudsoy ‘as—nd ainsseid pyoied ‘spuoses ut
ouLy, “Jaye Jopun ydoy premsoye pur ‘WIIOJOIOTYO YRIM pozoyjswue AT[eITUL sopry 1 JO So “ZO6T ‘eq Avy jo Juounsodxy —°*7 aang

Pt, Tremp yen —00!
Mata omtmatenth mT yyy YWANNSYYNWY* ost
—002
Wh
ot EU Se ba ek TL SA Eg ory Sa apg AT re at re ee

Acapnia and Shock. 173

IV. Hyprer-TONUS OF THE HEART IN SHOCK.

It is impossible in any one paper of ordinary length to discuss all
the factors concerned in acapnial shock. The behavior of the heart
is, however, so important that it must be noticed here. Crile and
Romberg and Passler concluded from their experiments that the heart
does not fail in shock. The only types of failure which are generally
recognized at the present time are those in which the ventricles do
not contract adequately. It is noteworthy that the pumping action
of the heart can be abolished by excessive contractility also. If for
any reason it relaxes less readily than normally in diastole, the resist-
ance thus offered to the distention of its chambers by the venous
inflow will result in a diminution of the systolic stroke.

The heart appears to differ from all other organs in that its tonus
is increased, instead of being diminished, by acapnia. Rapidly in-
duced acute acapnia augments cardiac tonus to such an extent that
death from cardiac tetanus results, as was shown in the first paper of
this series.4° In a less rapidly developing acapnia, such as occurred
in the experiments described in the present paper, hyper-tonus of the
heart is probably a secondary but important contributing factor in
the failure of the circulation. This hyper-tonus causes the ventricles
to relax less readily during diastole at the same time that the force
of the venous stream is diminished by the failing veno-pressor mechan-
ism. Thus two factors work together to lessen the pumping action of
the heart.

This behavior on the part of the heart itself appears to have been
first noticed in animals and described by Howell." Boise* has also
observed it clinically. He regards it indeed as the essential element
in traumatic shock. Jerusalem and Starling * have shown by ex-
periments on the isolated mammalian heart that the CO: content of
the blood must be maintained at a certain height if the pumping

4 HENDERSON, Y.: This journal, 1908, xxi, p. 143; also 1910, xxv, Pp. 325.

41 HowELL, W. H.: Contributions to medical research dedicated to V. C.
Vaughan, 1903, p. SI.

# Boise, E.: Transactions of the American Gynecological Society, 1908, p. 7;
and American journal of obstetrics, 1907, lv, p. I.

43 JERUSALEM and STARLING: Journal of physiology, 1910, xl, p. 279.

174 Yandell Henderson.

action of the ventricles is to be normally carried out. In acapnia
they find that the diastolic relaxation is incomplete, and the output
is therefore minimal.

V. CONCLUSIONS.

Traumatic shock and toxemic shock are in all essential features
identical. In both the circulation fails in the same manner as after
hemorrhage. It is illogical to call this condition ‘‘vaso-motor fail-
ure,” for the same evidence which shows that in shock the heart
is still functionally capable demonstrates also that the vaso-motor
mechanism is in a high degree of activity. The true vaso-motor
failure of spinal shock, after a high section of the spinal cord, is an
altogether distinct and different condition. The essential failing
element in traumatic and toxemic shock is the — hitherto unde-
scribed — veno-pressor mechanism.

The veno-pressor mechanism (to be more fully described in a later
paper) consists in part of the tonus of the tissues, and in part of osmo-
tic processes. The tonus of the contractile tissues is largely dependent
upon their content of CO,. This tonus prevents stasis by compressing
the capillaries. When it is diminished by acapnia, the blood stagnates
in the venous reservoirs. The tension of CO, within the body is regu-
lated by the respiratory centre. Thus this centre exerts an indirect
but powerful control over the veno-pressor mechanism.

The osmotic element in the veno-pressor mechanism is the regula-
tion of the water content of the blood, the lymph, the tissue fluid, and
tha cytoplasm of the tissue cells by the CO. which they contain.
Acapnia upsets the normal balance of osmotic forces and induces a
passage of water out of the blood into the tissues. When the blood
stream has been thus diminished to the point at which it supplies to
the tissues less oxygen than they demand, and the venous blood
therefore contains little or no oxygen, a condition of asphyxial acido- |
sis results. The proteins of the tissues swell and imbibe water from
the blood. If saline is infused into a vein under these conditions, it
passes rapidly out of the blood vessels into the tissues. Fatal olige--
mia quickly ensues.

Dogs were brought into shock by exposure and aeration of the ab-
dominal viscera and by the hyperpncea thus induced. When fatal

Acapma and Shock. 175

apnoea vera was prevented by continual sensory irritation, all (except
a few which were so deeply morphinized or were otherwise so resistant
as to refuse to breathe excessively) died from failure of the circula-
tion. During the development of shock large arteries (e. g., the
femoral) became constricted, not relaxed, — confirming the observa-
tions of Seelig and Lyon as against the supporters of the ‘‘ vaso-motor
failure’ theory. The cause of the failure of the circulation was a
diminution in the volume of the blood — oligeemia, or exsanguinity,
— similar to that observed clinically in traumatic shock by Malcolm,
and similar also to that occurring in the acute infectious diseases.
Blood gas analyses and arterial pressure records show two stages in
this process. . The first is due to acapnia and is curable. The second,
which develops from the first, is the fatal failure of the circulation
quickly following (in shock as in carbon monoxid poisoning) the
occurrence of venous anoxhemia.

The heart in shock does not fail, —if by failure is meant beating
more and more feebly until it stmks into prolonged diastole. On the*
contrary, during the tachycardia early in shock the systoles empty the
ventricles more completely than normally. With the increased tonus
induced in the heart by acapnia the ventricles require more than a
normal venous pressure for their diastolic distention, — at the same
time that the force of the venous stream is diminished by the failing
veno-pressor mechanism. Thus hyper-tonus of the heart, as held by
Howell and by Boise, and confirmed by Jerusalem and Starling, is an
important contributing element in circulatory failure.

The essential sequence of events in acapnial shock is: (1) hy-
perpnoea; (2) acapnia; (3) failure of the veno-pressor mechanism
(or sometimes sudden fatal apnoea); (4) venous anoxhemia, tissue
asphyxia, and acidosis; (5) acute oligamia.

I am indebted to my colleague Prof. F. P. Underhill for valuable
assistance and criticism in this work, and to Dr. T. B. Barringer and
Mr. S. C. Harvey for collaboration in the related topics, here referred
to, which will be published in later papers.

Note. — The following additional references may be useful to any one caring
‘to look up the literature bearing on the topics here discussed: Viscosity of blood
increased by COs, Zeitschrift fiir klinische Medizin, 1909, lxviii, p. 177; MULLER,
’ W.: Mitteilungen aus den Grenzgebieten der Medizin und Chirurgie, 1910, xxi,
Pp. 377. Alkalinity of blood influenced by COz, HENDERSON, L. J., and Sprro, K.:

176 Yandell Henderson.

Biochemische Zeitschrift, xv, p. 114; Boycotr and Cutsorm: Biochemical jour-
nal, 1910, v, p. 23. Influence of CO: on colloids, BAyLIss: Proceedings of the Royal
Society, 1900, Ixxxi, B, No. B, 548. Chlorides in pneumonia, v. Wyss, H.: Zeit-
schrift fur klinische Medizin, 1910, lxx, p. 183. Water content of tissues and alter-
ations in the circulation in acute infectious diseases, SCHWENKENBECKER and
Inacaxki: Archiv fiir experimentelle Pathologie und Pharmakologie, 1906, ly,
p. 203, and 19090, lx, p. 166; MrvEr, F., same volume, p. 208; Lusx, G.: The
Science of Nutrition, second edition, 1909, p. 331; Saline infusions in acute dis-
eases, LENHARTZ, Deutsches Archiv fiir klinische Medicin, 1899, lxiv, p. 189;
and in cholera, NicHots and ANDREws: Philippine journal of science, 1909, iv,
B, p. 81.

METABOLISM OF DEVELOPMENT. —II. NITROGEN BAL-
ANCE DURING PREGNANCY AND MENSTRUATION OF
THE DOG.

By JOHN R. MURLIN.

[From the apse an Laboratory of the Cornell University Medical College,
New Vork City.]

HIS paper contains the results of a part of the experiments done
within the past two years and a half with the object of deter-
mining what influence the development of the embryo may have on
the protein metabolism of the mother dog. Starting with a chance
observation made in 1907,! that the output of nitrogen in the urine of
the fasting pregnant dog is much more profoundly influenced by a
small amount of ingested carbohydrate than it is in the urine of a
non-pregnant dog, the writer has since been interested both in the
quantitative and the qualitative alterations in the metabolism pro-
duced by pregnancy and menstruation. The first paper of the series ?
dealt with the quantitative effects of pregnancy on the energy metab-
olism. The present one deals with quantitative effects of pregnancy
and menstruation, and the next will take up their qualitative effects, .
on the protein metabolism.

NITROGEN BALANCE IN PREGNANCY.

The purpose of studying the nitrogen balance in pregnancy is, pri-
marily, to determine the advantage or disadvantage to the maternal
organism; secondarily, to determine the source and the amount of the
protein materials entering into the product of conception. One would
suppose a@ priori that from the mdment of fertilization of the ovum,
and in order to provide material for its growth and development,
nitrogenous materials from the food would be retained and be built

1 Murtin: Proceedings of the Society for Experimental Biology and Medicine,
1908, v, p. 72.
* Morin: This journal, 1910, xxvi, p. 134.
177

178 John R. Murlin.

up into the new protoplasms. But another possibility exists. The
new growth may require protein materials not furnished by the food,
or the maternal organism may not be capable of assimilating suffi-
cient food protein to meet the additional demands. In either case
nitrogenous materials would be taken from the mother’s own body,
and a comparison of the nitrogen intake with the nitrogen output for
the entire period of pregnancy ought to reveal this fact. If, for ex-
ample, the animal should remain in nitrogen equilibrium throughout
the pregnancy, we should say at once that the embryos had been
produced at the expense of the maternal tissues, or, if produced to any
extent at all from the food proteins, had occasioned a corresponding
loss from the maternal tissues to the outside world. The exact amount
of the total loss could be determined by analyzing the entire birth —
embryos, placenta, membranes, and fluids — and adding any extra
nitrogen from involution of the uterus found in the urine immedi-
ately after parturition. If there should be a certain amount of nitro-
gen retained from the food, but the amount discharged at birth were
in excess of this, the difference would represent the amount lost from
the maternal tissues either to the embryonic tissues or to the outside
world. If the amount issuing at birth together with the amount
which could be accounted for by growth of the uterus and mamme
should be just about equal to the amount retained from the food, one
would be justified in saying that the nitrogen retention is for the ex-
press purposes incident to the formation and growth of the embryo.
Any nitrogen retained beyond this amount would be so much clear
gain to the maternal organism.

Should an animal which had been in nitrogen balance before preg-
nancy show on the same diet, during any considerable portion of the
period of pregnancy, a condition of minus nitrogen balance, one could
say that the katabolic processes of the mother’s body were for the
time more active than in the non-pregnant condition. It would not,
of course, signify that the nitrogen loss came from the embryo or
from the surrounding maternal structures, since in both these situa-
tions, so long as development goes on normally, the synthetic or ana-
bolic processes must be in the preponderance. The most natural
interpretation would be that materials are being mobilized for the
purposes of embryonic growth.

Should either a net loss of nitrogen from the mother’s body for the

Metabolism of Development. 179

whole period of pregnancy or a condition of minus balance for any
particular portion of the period prove to be invariable and characteris-
tic of pregnancy, one would say with some assurance that the embryo
depends to a greater or less extent upon the maternal tissues rather
than upon freshly absorbed food protein for building materials. Such
a fact would be a suggestive one from the standpoint of heredity,
since, if the embryo (after fertilization of the ovum) is of necessity
formed to any extent at all from the maternal proteins, a possible
mechanism for the transmission of the species (not the individual)
characters, supplementary to that of the pre-organization of the
germ cells, is introduced.

HISTORICAL.

The first serious attempt to determine the influence of pregnancy on
the nitrogen balance was that of Hagemann.’ His experiments, two
in number, were performed on dogs. When the amount of nitrogen
delivered at birth in the embryos, placenta, etc., was considered,‘ it
was seen, according to Hagemann’s interpretation, that there was a
net loss from the mother’s body. Hagemann concluded, therefore,
that ‘‘the formation of the fetus occasioned a loss of protein from the
mother’s body, even when there was abundance of nourishment”
(sehr reichliche Néhrung).

Ten years later Ver Ecke *® reported numerous (nineteen in all) sim-
ilar experiments on rabbits. His figures show that in many of them
there was a net loss of nitrogen from the mother’s body when the young
were born. From which he draws the rather dramatic conclusion that
“gestation constitutes a sacrifice of the individual for the species.”’
Both Bar® and Jagerroos,’ the only observers since Ver Ecke who

3 HaAGEMANN: Archiv fiir Anatomie und Physiologie, 1890, p. 577, and
Inaugural Dissertation, Erlangen, 1891.

4 This was not determined directly, but was estimated from BiscHorr and
VOLKMANN’s figures.

5 VER Ecke: Mémoires courronnés, Académie Royale de Médicine de Belgique,
1901, xv, No. 7. A good digest of this paper by Macnus-Levy is to be found in
Von Noorpen’s Handbuch der Pathologie des Stoffwechsels, 1906, i, pp.
403, 404; English translation, ‘‘ Metabolism and practical medicine,” i, p. 373.

5 Bar, P.: Lecons de pathologie obstetricale, 1907, ii, p. 271.

7 JAGERROOS: Archiv fiir Gynaekologie, 1902, lxvii, p. 517.

180 John R. Murlin.

have worked on animals, are inclined to discount this conclusion pretty
strongly, because of the difficulties of carrying out metabolism ex-
periments on rabbits with sufficient accuracy. My own examination
of Ver Ecke’s original paper confirms their opinion.

_ The next year Jagerroos published results of four experiments on
dogs. In three out of the four — one on low, one on high, and one on
a moderate amount of protein — (and in all cases an abundance of
potential energy in the food) there was, according to Jagerrdos, a net
loss of protein from the mother’s body as the result of the pregnancy.
The one dog which may have shown a net gain or at least a condition
of equilibrium (counting the nitrogen lost at birth) was given an ex-
cessive diet both as regards protein and potential energy and aborted
two weeks before term.

At times in all of the experiments there was more nitrogen in the
excreta alone than in the food. Jagerroos says: ‘In these periods of
negative nitrogen balance the tissue protein (Organeiweiss) therefore
must have served as building material for the embryos. In all my
experiments, on the other hand, the food protein (circulierendes
Eiweiss) has at times likewise served as building material, if a net
sparing may be counted as a direct gain to the embryo. The question
whether tissue or food protein is used in the formation of the fetus is
therefore to be answered by saying that probably tissue as well as
food protein can serve this purpose.’’ Whether a condition of negative
balance, that is, of increased protein katabolism, is characteristic of
any portion of the period of pregnancy, is a question which Jagerroos
leaves undecided.

Bar criticises the work of Jagerroos at two points: (1) he thinks the

diet selected was not’ suitable, and inveighs pretty heavily against |

the continuous use (in Experiment I) of only cane sugar and meat.
(2) Jagerroos, following Hagemann, overestimated the amount of
nitrogen in the placente and membranes. The facts amply justify
Bar’s criticisms. Two of Jigerroos’ experiments must be discarded
entirely so far as the net result of the pregnancy is concerned,
because the dogs aborted and ate the young. The other two Bar
shows are scarcely demonstrative of the conclusions which Jagerroos
draws.

Bar’s own experiments are unquestionably the most satisfactory,

all points considered, which have yet been done on the metabolism

Metabolism of Development. ISI

of pregnancy in animals. In five of his experiments on dogs he de-
termined the weights of the young at birth and, by muzzling the
female, obtained also the weights of the placentee and membranes.
In four experiments the young and adnexa were analyzed. Since
Bar’s results have never been published in the current literature and
we shall have occasion to refer to these figures again, they are pre-
sented here.

: : Per cent of N : N in adnexa
yi Bee de ney in young at per 100 of N in

gm.
24.59

6.809
21.51
8.679

The percentage of N in the young alone agrees very well with the
figure given by Jagerroos (2.17 per cent), but the relation of N in the
adnexa to that in the fetus is only one fourth of the amount assumed
by Hagemann (0.127 instead of 0.46).

Bar also sacrificed two of his pregnant dogs, one at the thirtieth
day (middle) of gestation, the other at the forty-fifth day, in order to
obtain the amount of N required by the entire ovum at different stages
of the pregnancy. These data enabled him to plot a curve repre-
senting the amount of nitrogen diverted to the ovum, and, by differ-
ence, to obtain the amount lost or gained by the mother’s own body
at these successive stages. Similar curves for the rabbit are even
more complete. His figures for the dog are as follows:

At thirty days a single ovum contained 0.162 gm. N.
At forty-five days a single ovum contained 2.09 gm. N.
At term a single puppy contained 4.62 gm. N.

The mothers were covered in the three cases by the same sire, were
kept on the same diet (proportioned to weight) and under the same
general conditions; but the dog killed at the end of thirty days was

182 John R. Murlin.

only about half as large as the one killed at forty-five days. The
puppies analyzed at term and used in the table above were from the
dog which (in the next pregnancy) was killed at forty-five days. It
is possible, therefore, that the figure given for thirty days is propor-
tionally a little too low in comparison with the other figures.

Bar’s first dog gave birth to four puppies and ended her pregnancy
with a net gain of 5.24 gm. Where this was deposited Bar does not
attempt to say, but presumably in the uterus* and mamme. His
second dog gave birth to only a single puppy and ended the pregnancy
with a net gain of 27.02 gm. N. The third experiment was marred
by the occurrence of diarrhoea, vomiting, and loss of appetite from
the twenty-first to the thirtieth days, notwithstanding which and the
birth of seven puppies the pregnancy ended with a net loss of only
1.27 gm. N. Bar presents results from two rabbits which he believes
are trustworthy, showing a retention of nitrogen from the beginning to
the end of the period of gestation. His conclusion from all his animal
experiments is that, “‘given a sufficient ration, the healthy mother’s
body can supply the needs of the developing fetus without drawing
upon its own capital, even when those needs are extreme.” °

On the basis of his own observations together with those of Hage-
mann and Jagerroos, Bar recognizes two periods in the nutrition of
the pregnant organism, — the first extending from the time of fer-
tilization to the middle of gestation (thirtieth day in the dog) and the
second from the middle to the end. The dividing line corresponds with
the beginning of active growth of the embryo.!° The second period is
characterized by a continuous and progressive retention of nitrogen
up to the very day of parturition. In the first he distinguishes two
phases: (a) one of “retention” immediately following fertilization and
lasting until the second or third week (in the dog), and (6) one of
‘saturation,’ characterized by a diminished retention of nitrogen, a
state of equilibrium, or even a loss of nitrogen, which continues until
about the middle of gestation. This latter period he is convinced is
not due to any diminution in the absorption of the nitrogenous mate-

§ Bar analyzed the uterus and mamme of the dog killed at term and found
3 gm. N in the uterus and 11.3 gm. in the mammez. He estimates that the three
dogs reported in full would require from 6 to 14 gm. for these tissues, according to
weight. Loc. cit., p. 292.

9 Bar: Loc. cis. 204.

10 Cf. RUBNER: Archiv fiir Hygiene, 1908, Ixvi, pp. 180 e¢ seq.

Metabolism of Development. 183

rials, but to a real increase in the processes of disassimilation. Bar
offers no explanation, but mentions the possibility of proteolytic en-
zymes, or cytolysins produced by the embryo and acting upon the
maternal tissues to produce a true mobilization of protein materials
(une véritable mobilization des albumines)." This period he points
out corresponds in time with the period of morning sickness in women.

Aside from this work on animals a number of trustworthy observa-
tions * have been made on the nitrogen balance in pregnant women,
but all save one of them have been confined to a few weeks at most,
and all save two to the very last weeks of the pregnancy. All have
observed a retention of nitrogen varying in amount from 1.5 to 7 gm.
per day, according to the stage of pregnancy and the amount of nitro-
gen ingested.

Summarizing all of the work reported to date, one may say that
with respect to the first half of pregnancy the weight of evidence is
that a period of diminished retention or of minus nitrogen balance is
likely to occur in animals at about the stage which corresponds to the
period of morning sickness of women ™ (see tabulated summary, p.200).
With respect to the second half there is perfect agreement that more
nitrogen is regularly ingested than is excreted, provided a “‘sufficient”’
amount of nitrogen and potential energy is supplied. In Jagerroos’
experiment No. IV 0.2 gm. N and 70 calories per kilo were not suffi-
cient to maintain equilibrium except for one week (the third) out of
the seven that the dog lived. In his experiment No. I 0.8 gm. N and
75 calories per kilo were sufficient to maintain a plus balance every
week after the fourth. In all the other animal experiments a plus
balance prevailed throughout the latter half, and in the single experi-
ment of Hoffstr6m on a woman it prevailed from the beginning of the
' fifth lunar month uninterruptedly to the end of gestation.

Whether the total amount of nitrogen retained is in all cases sufficient

SPA. Lac. Gi. Pp. 288.

12 ZACHARJEWSKY: Zeitschrift fiir Biologie, 1894, xxx, p. 368; SCHRADER:
Archiv fiir Gynaekologie, 1g00, lx, p. 534; SILLEVIS, J.: Jets over de Stofwiselling
der Gravida, Academie Poefschrift, Leyden, 1903, cited by Horrstr6m, Loc. cit.;
SLEMONS: Johns Hopkins Hospital reports, 1904, xii, p. 111; Haunt: Archiv fiir
Gynaekologie, 1905, Ixxv, p. 31; Bar: Loc. cit., p. 243; Horrstrom: Skandana-
visches Archiv fiir Physiologie, 1910, xxiii, p. 326.

13 No one has yet attempted to follow the nitrogen balance in pregnant women
earlier than the sixteenth week.

184 John R. Murlin.

to cover the requirement for growth and development is quite a differ-
ent matter. Hagemann and Ver Ecke were convinced by their, ex-
periments that in gestation the embryo must appropriate to itself a
portion of the maternal protein materials; or, in the words. of Ver
Ecke, “‘gestation constitutes a sacrifice of the individual for the spe-
cies.”’ Jagerroos admits the possibility of this, but does not look upon
it as necessary. He does not regard the embryo as a parasite, but
admits that it may utilize maternal proteins if food proteins are not
available. Bar regards gestation not as a sacrifice of the old indi-
vidual for the new, but as an instance of ‘‘ homogeneous and harmoni-
ous symbiosis.” * It is not an occasion of loss, but of profit to the
maternal organism as well as to the embryo.

AUTHOR’S EXPERIMENTS.

The experiments about to be reported were completed without
any knowledge that Bar’s magnificent work had been done. None
of those alluded to here seem ever to have been published until
they appeared in his ‘“Lecons,” and that work became known to the
writer only by reading Hoffstrém’s paper. It seems regrettable that
his results should not have been given more extended circulation
chrough the medium of some good journal. It is the more gratifying,
however, to find the results reported in this paper in harmony, in the
main, with those of so excellent an observer. Bar admits that he neg-
lected the influence of menstruation because he could not predict the
return of the cestrual cycle. That deficiency the following experiments
will supply. It is of much more importance than the writer at first
realized to have one’s dog in equilibrium, or at least to know the pro-
tein condition of the animal, before the beginning of the sexual phe-
nomena. Not only Bar, but Hagemann as well, was led astray by
this lack of knowledge. Bar also neglects the influence of the total cal-
orific content of the foods in the interpretation of his own experiments
as compared with those of Hagemann and Jagerroos.

In planning the following experiments the chief improvement over
those of Jigerroos aimed at was in the matter of suitableness of the
diet. Jagerroos could not be certain that the periods of minus nitrogen
balance were not due entirely to lack of appetite, and this in turn to

M4 JAGERROOS: Loc. cit., p. 298.

Metabolism of Development. 185

the unsuitable character of the diet and to confinement. With the
exception of the early part of Experiment I and a day or two of Ex-
periment V of the present series, no trouble was experienced with poor
appetite, and there has been no sign of malnutrition such as the birth of
dead puppies. In Experiment I only one puppy was born, but this
may be due to the fact that copulation occurred only once. In all the
other experiments copulation was permitted on two or three successive
days, and four was the smallest number of puppies born.’ The cage used
nearly all the time was a commodious one, permitting considerable
freedom of movement by dogs of the bull-terrier breed (12 to 16 kgm.
weight) such as those here employed. Bull-terrier dogs are most
satisfactory for metabolism experiments both on account of their dis-
position and their hardihood.

The chief reason, however, for the good fortune in keeping the dogs
in satisfactory nutritive condition which has attended these experi-
ments seems to lie in the diet selected. Ground beef-heart (using only
the thick ventricular wall free of both pericardium and endocardium)
has been found to be a most satisfactory form in which to give protein
to dogs, because the N-content is very constant ” and because, mixed
with cracker meal, lard, bone ash, salt, and enough water to make the
whole mass slightly wet, it imparts a peculiarly agreeable taste. Only
two dogs have been used in the experiments about to be reported, but
all of the many dogs which have been given the diet have taken it
greedily every day for as long as ten weeks without any sign whatever
of distaste or malnutrition.

The average gestation period in the dog is just nine weeks (sixty-
three days). In the nitrogen balance experiments below the excreta
were collected in weekly periods and a balance was struck at the end
of each week.

Experiment I. Dog A. Bull-terrier weighing about 12 kgm. was copu-
lated in the laboratory by a strong male of the same breed on April 22,
1908. Collection and analysis of the excreta and analysis of the food
were begun immediately. The urine was collected for the most part by
the catheter and was separated into twenty-four-hour periods by wash-
ing the bladder. Immediately after this the dog was weighed, then fed

15 Tt is only by careful selection of hearts, securing every day organs of about
the same weight, and weighing the meat immediately after cutting and grinding,
that one can depend on the N-content without analysis.

186 John R. Murlin.

(only once in twenty-four hours) and taken for a walk. The urine for
the day was made up to a standard volume, and after a thorough mix-
ture an aliquot part (one fourth or one fifth) was put aside for the
weekly urine. In this experiment the balance of the daily urine also
was saved on two days of each week, and a fairly complete analysis was
made (see following paper of this series). On the same two days the
dog lived in the respiration apparatus for twenty-two out of the twenty-
four hours."

This dog was very cleanly (according to the standard of dogs) about
her person, so that it was unnecessary to collect the vaginal discharge
for analysis.

For convenience in preparing the food an attempt was made to
supply protein in the form of meat powder, a commercial preparation
known as ‘‘meatox.” Varying quantities of this preparation with lard
and cane sugar were given the first week in an attempt to suit the
dog’s appetite; but at the end of two weeks the dog refused to take
this diet, and the change was made to beef-heart instead of the meat
powder. Separate analyses of the ground ox-heart, freshly cut, showed
the following N-content: 3.00, 2.98, and 3.00 per cent respectively.
For the second two weeks the daily diet consisted of

275 gm. beef-heart (3% N and 5% fat) 8.250 Nand 339.2 cal.

40 lard = 0.0 a2
40 “ -cane sugar = 0.0 156
to “ bone ash
t ‘* common salt
Total = 8.250 867.2 cal.

At the beginning of the fifth week another change was made: 50 gm.
cracker meal, containing 1.48 per cent N and yielding 196.2 cal., were
substituted for the 40 gm. cane sugar. The entire diet then furnished
8.990 gm. N and 907.4 calories, of which 24 per cent was supplied by
protein. This was continued every day and was taken eagerly up to
July 15, seventeen days after parturition, when the experiment ended.
The one puppy, born on the sixty-third day (June 26) and weighing
285 gm. as soon as dry, was apparently normal in every respect but one,
namely, the absence of one eye. It was allowed to suck for a few days
and then was taken away in order to reduce the dog’s metabolism to
the base level of entire sexual rest.

16 See Experiment I, Murtin: This journal, 1910, xxvi, p. 134.

Metabolism of Development. 187
TABLE I.

Doc A. Frrst PREGNANCY. NITROGEN BALANCE BY WEEKS.

- BW es Nitrogen Nitrogen Nitrogen
Week of gestation. Calories in food. foat a oesta, Falance:
1908
i. 4180.8 44.970 45.945 —0.975
April 23 to 29 (50 cal. per kg.)
18 66 $e 505:0 54.905 57.341 — 2.436
April 30 to May 6 (58 cal. per kg.)
i. 6152.8 54.555 56.844 —2.289
May 7 to 13 || (71 cal. per kg.)
TA 6070.4 57.750 58.075 —0.325
May 14 to 20 (69 cal. per kg.)
V. 6351.8 62.190 58.937 +3.153
May 21 to 27 (69.8 cal. per kg.)
VI. 6351.8 62.930 59.882 +3.048
May 28 to June 4 (68 cal. per kg.) |
VII. 6351.8 62.930 57.316 +5.614
June 5 to 11 (66 cal. per kg.)
VIII. 6351.8 62.930 57.216 +5.714
June 12 to 18 (64 cal. per kg.)
IDG 6351.8 62.930 56.317 +6.613
June 19 to 25 (62.5 cal. per kg.)
I. Post partum 6351.8 53.940 54.533 —0.593
June 27 to July 2 (65 cal. per kg.) (6 days) (6 days)
July 14 to 15 66.3 8.990 4.576 +4.414

One puppy born. Weight, 285 gm.

The nitrogen balance for Experiment I is givenin Table I. In the
second column of the table are given the apparent total calories of
energy in the food and the calories per kilogram calculated on the basis
of the average weight of the dog for the week. In the other columns
are shown the total nitrogen in the food, the amount found in the ex-
creta, and finally the plus or minus balance. It will be observed that .
a minus balance prevailed throughout the first four weeks, then changed
to a plus balance, which increased gradually up to the day of parturi-
tion. The figures given for the nitrogen output do not include the
nitrogen lost by falling off of hair or epidermal scales. The result by
weeks is not thereby affected, however, for, while the analysis for this
quantity of nitrogen would have made the minus balance in the early
weeks a little greater, it is certain that it would not have changed the
_ plus balance of the fifth nor any of the subsequent weeks to a minus
balance.

There can be no doubt that the total supply of energy was sufficient

188 John R. Murlin.

to maintain the dog in energy equilibrium, for the total energy pro-
duction was determined for this pregnancy by calculation from the
output of C and N, and at no time, not even in the last week of preg-
nancy, was the production from the dog’s body equal to the supply
from the food. (see Table I of previous paper).

The total amount of nitrogen lost from the mother’s body ie the
first four weeks was 6.025 gm. The total amount retained for the
last five weeks was 24.142 gm. —a net retention for the entire period
of gestation therefore of r8.127 gm. The puppy was not analyzed at
birth because it was desired for the purpose of the respiratory metab-
olism experiments reported in a previous paper. Figures are avail-
able from Bar’s and Jagerroos’ papers, however, which will enable us
to calculate the amount of nitrogen it contained accurately enough
for our present purpose. Taking the mean of all their determinations
(2.16) as the percentage composition of the puppy in the present case,
the total amount contained (weight, 285 gm. ) would be 6.15 gm.
Adding 12.7 per cent for the membranes, placenta, and fluids (see
Bar’s table, p. 181), or 0.78 gm., we have 6.93 gm. as the amount
delivered at birth. The excreta on the day of parturition and for five
days thereafter contained, on the average, 0.5 gm. more nitrogen than
the food. This extra nitrogen, which is practically all in the urine, has
been regarded by Jagerroos and others as belonging to the birth. In
this dog the extra nitrogen for the first day or two represents blood,
placenta, amniotic fluid, etc., ingested at parturition, and the rest of
it is doubtless traceable to involution processes, 7. e., represents nitro-
gen stored temporarily in the uterus. On the sixth day after parturi-
tion in this case the extra nitrogen in the excreta had fallen to 0.06 gm.
This point therefore may be taken as representing the end of the most
active involution, though it is not to be supposed that the uterus has
yet reached the minimal size of sexual rest. At any rate, counting
0.5 gm. a day for five days, we have 2.5 gm., at the most, to be added
to the 6.93 gm., making a total of 9.43 gm. N as the final product of
gestation. In this pregnancy, therefore, the mother has furnished all
the protein materials to the embryo and has retained for her own body
(18.127 — 9.43) 8.69 gm. nitrogen besides. In other words, the
mother’s body is in a better condition as regards protein at the end of
the pregnancy than it was in the beginning. It does not follow, how-
ever, that the maternal tissues have not been drawn upon for build-

Metabolism of Development. 189

ing materials. For the first four out of nine weeks, without taking any
account of losses through the skin, there was a condition of negative
balance, which, according to the interpretation given on page 178,
means that some of the material contributed to the embryo during this
time probably had its origin in the mother’s own tissues.

Experiment II. Dog B. A brindle-colored female with a predominant
strain of bull-terrier. blood and weighing about 13 kgm. was copulated
by a bull-terrier male on February 1, 2, and 3, 1909. The daily routine
of collecting and separating excreta was exactly the same as in Experi-
ment I. The food given to this dog was the same in general character
as that given to Dog A in Experiment I, and was designed to furnish
about 7o cal. per kilogram of body weight. After a few days’ trial the
following diet proved to be satisfactory and was given every day from
February 3 to April 4 at the same hour:

250 gm, beef-heart (2.73% N1 and 5% fat) = 6.825 N and 307.5 cal.

80 “ cracker meal (1.48% N) Tee 213.6
4o “lard 372.0
to ©“ bone ash
ce
2 common salt
Total = 8.009 993.1 cal.

20% protein calories.

Four puppies were born on April 4, the sixty-third day from the first
copulation. They were all large and well formed, weighing in the
neighborhood of 350 gm. each. They were allowed to suck for about
three weeks and then were taken away for the purpose of reducing the
mother’s metabolism to the level of sexual rest. The diet during these
three weeks contained more meat than before, but otherwise was of
about the same general character as above. It was not analyzed. That
it was probably sufficient so far as energy is concerned, is shown by the
fact that the mother dog maintained her body weight (13.1 kgm. on
May 7 as compared with 13.0 kgm. just after parturition), although
the puppies grew rapidly, weighing on the 2oth of April 1220, 1280, 1180,
and 1080 gm. respectively. On the 3d of May the original diet was re-
sumed and was continued until the rath.

17 The apparently lower percentage of N in the beef-heart of this experiment
was due to an error in the standardization of the acid. The same error applies
to all the N determinations of this experiment, but not, of course, to the balance
figures.

190 John R. Murlin.

The results of this experiment are given in Table II. It is seen that
the minus nitrogen balance prevailed in this case throughout at least six
of the nine weeks of pregnancy. It is necessary to say “at least,” be-
cause it is possible that the amount of nitrogen lost by shedding of
hair, epidermal scales, and vaginal secretion (this dog was not so
cleanly about her person as Dog A) would have exceeded the 1.801 gm.,
the amount apparently retained in the seventh week. Chittenden 1*
reports a dog of about the same weight and the same general charac-
ter of coat which lost on the average 0.21 gm. N per day through the
hair. Assuming the same rate of shedding in this dog, the total loss
for the week through this channel alone would amount to 1.47 gm.
There can be no doubt, however, about any other week.

TABLE II.

Doc B. NirroGEN BALANCE By WEEKS.

: i: Nitrogen Nitrogen Nitrogen
Week of gestation. Calories in food. in food: in Qeeretat balance.

1909

I. 5382.1
February 3 to 9 (56 cal. per kg.) 53.287 63.116 —8.829
Il 6851.7

56.063 60.893 —4.830

February 10 to 16 (72 cal. per kg.)
_ 6851.7 56.063 «| 62.031 —5.968

February 17 to 23 (71 cal. per kg.)
a fue 56.063 64.508 — 8.445

February 24 to March 2 (71 cal. per kg.)
¥ ee 56.063 62.594 —6.531

56.063 60.064 —4.001
56.063 54.262 +1.801

March 24 to 30 (63 call perke.) | soo 47.042 “9-021
Ix 3972.4 32.036 25.786 +-6.250
March 31 to April 3 (61 cal. per kg.) (4 days) (4 days) .
2 weeks after lact. 5958.6 48.054 40.006 +8.048
May 7 to 12 (76 cal. per kg.) (6 days) (6 days) ;

March 3 to 9 (70 cal. per kg.)
VI 6851.7

March 10 to 16 (68 cal. per kg.)
VII 6851.7

March 17 to 23 (66 cal. per kg.)
VIII. 6851.7

Four puppies born. Weight, 1400 gm.

For the first six weeks there was a total loss by the excreta of
38.604 gm. N. For the last three weeks there was a total retention

18 CHITTENDEN: Nutrition of man, New York, 1907, p. 250.

Metabohsm of Development. IOI

over the amount discharged by the excreta of 17.022 gm. The four
puppies weighed at birth about 1400 gm. and contained (2.16 per cent;
see p. 181) not less than 30.2 gm. N. Taking into account the nitrogen
loss by the fluids, membranes, and placente (12.7 per cent of N in
fetus), the total amount delivered could not have been less than 34 gm.,
and was probably quite a little more than this. The net result thus
was a loss of at least (38.604 + 34.0) — 17.0 =) 55.6. This case
therefore presents conditions just the reverse of those presented by
the previous case.

The contrast with this of the metabolic conditions in sexual rest is
very striking. The six-day period from May 7 to 12 inclusive, five
weeks after parturition and two weeks after the dog ceased nurs-
ing her brood, shows a positive balance of about the same propor-
tions as that of the eighth week of gestation. Every trace of milk
had disappeared from the mamme by this time, and the processes
of involution were presumably long since at an end. ‘The
nitrogen retained therefore must have gone to the maternal tissues
themselves. 5.

This high retention doubtless was due in part at least to the impov-
erishment resulting from the heavy drain upon the maternal tissues
which had occurred in the pregnancy period; for while the diet during
the period of lactation may have been sufficient for the demands of
that period, it probably was not sufficient to enable the mother to
recoup herself completely for the previous losses. The metabolism
at this time therefore probably does not represent the base line of
perfect sexual rest, but only an approach thereto. Unfortunately the
case could not be followed: further.

The results of Experiments I and II are represented graphically in
Fig. 1. The continuous line in each case represents the protein con-
dition of the mother, assuming that she was in nitrogen equilibrium
(i. e., that the ordinary demands of the tissues for protein building
materials had been kept up) at the beginning of the pregnancy. The
preliminary decline in protein condition at the beginning of Experi-
ment I had been fully compensated for at the end of the sixth week.
From this time to the end of pregnancy there was a steady gain. In
Experiment II the decline continued until the end of the sixth week,
and was not yet compensated at the end of pregnancy by something
over 20 gm. N.

The discontinuous lines in the charts represent the amounts of

IQ2 John R. Murlin.

nitrogen diverted to the embryos and their adnexa at successive
stages of the gestation period. These curves are based on Bar’s
figures (page 181) for the nitrogen content of the entire ova at the
thirtieth and es fifth days and the end of gestation, due allowance

36 i‘ being made for the differences in weight

of the puppies. In the first experiment
Ae / from which only one puppy was born it is
e / clear that the pregnancy ends with a con-
i siderable advantage, in protein condition,
i to the mother herself (differences between
f the two curves); while in the second ex-
i periment it ends with the mother’s body
considerably depleted.
It is highly significant, as Bar has pointed

I2 out, that the improvement in general pro-
16 tein condition should begin at about the

6
DAYS oF SG il

time ‘‘active growth”’ of the embryos be-
comes of some consequence from a com-
parative standpoint. Rubner?® also, from
a consideration of the dynamic conditions

aes l7 2 38%49 68 ip gestation, has emphasized the fact that
Ficure 1.— Unbroken lines the embryo is of no significance, relatively
represent the protein condi- sneaking, until about the middle (0.4) of
tion of the pregnant dog at A ke Stee
successive stages of the preg. the period. Only after the t rtieth day,
nancy, and broken lines the in the dog, does the amount of nitrogenous
amount of nitrogen diverted materials required by the embryo become
to the embryos. Experiment 5 A :

_ appreciable. At about the same time in

I, Dog A, one puppy born ; d ; :
weight, 285 gm. Experiment Experiment I nitrogen began to be retained.
II, Dog B, four puppies born; In Experiment II this point is marked by

weight, 1400 gm. a decrease in the loss (6.5 gm. N in the fifth

as compared with 8.4 gm. in the fourth week). From this time the
condition is progressively better, and from the sixth week to the end
the rate of improvement is almost exactly that of the growth rate (7. e.,
the two curves are nearly parallel). One would perhaps be justi-
fied in saying that all of the nitrogen retained during the last three
weeks found its way to the embryos.

Since the total amount of energy in the food of the two dogs was

19 RUBNER: Loc. cit.

pee.

Metabolism of Development. 193

almost exactly proportional to their weights (see tables on page 187
and rgo) and the nitrogen intake was not widely different, it is signifi-
cant also that the total amount of nitrogen lost up to the end of the
fourth week in the two cases is nearly proportional to the weight of
new-born delivered (6 : 28 :: 285 : 1400 nearly).

This itself strongly suggests that the cause of the minus balance
is the presence of the embryos. We have as yet no proof of this,
however, because, as already noted at page 184, we are not certain that
the dog would have been in nitrogen equilibrium on the same diet
just previous to menstruation. With a much higher supply of po-
tential energy and a somewhat lower supply of nitrogen, Bar found
in all his dogs a condition of plus balance in the first two weeks of
pregnancy; but not having observed it himself, he did not correctly
interpret the influence of menstruation. Realizing the importance of
this factor, the following experiments were planned:

INFLUENCE OF MENSTRUATION ON THE PROTEIN CONDITION OF THE DOG.

Experiment III. Dog A. Third pregnancy. Having had this dog under
observation through two previous pregnancies, I was able to predict

TABLE III.

EXPERIMENT III. Doc A.

N N Total N N N
sok ee ee we in urine in feces excreted balance balance
mee per week. | per week. | per week. | per week. per day.
1909.
March 20 to 26 55.268 2.901 58.163 +3.767 +0.538
March 27 to April 2 57.360 2.901 60.261 +2.669 +0.381
Last week
of sexual rest Gy/aAl 3.847 60.968 +1.962 +0.280
pril 3 to 9
April 10 to 16 55.209 3.494 58.703 +4.227 +0.604
Active menstruation
Apa 17 to 23 t 52.102 3.453 55.555 46.375 | +0911
First of pregnancy
; April 23 to 29 55.448 3.333 58.781 +4.149 +0.593
econd of pregnancy
April 30 to May 6 rf 58.790 2.631 61.421 +1.509 +0.215

1 The food contained every day 8.990 gm. N and 907.4 calories of energy (66 to
70 cal. per kgm.).

194 John R. Murlin.

the return of the cestrual cycle pretty exactly. The dog was accordingly
placed on a measured diet about four weeks in advance of the expected
menstruation. Collection and analysis of excreta began at the same
time. First menstrual blood was observed on April 16, and on April 23
first copulation occurred.

The diet was exactly the same as that given during the last five weeks
of Experiment I. The routine also was the same as in that experiment,
except that the dog was not used for respiration experiments, and the
excreta were collected for weekly periods only. Four puppies were
born on June 26, the sixty-fourth day of gestation.

We see clearly that the diet, which was precisely the same as given
to this dog during the last five weeks of her first pregnancy, was en-
tirely adequate to maintain her in nitrogen equilibrium during sexual
rest. It is equally clear that the minus balance which prevailed in
the first two weeks of her first pregnancy (Table I) was due to the
inadequate diet given at the beginning and not to anything peculiar
to the early stages of pregnancy. Because of an engagement to work
in another laboratory, it was impossible for the writer to follow the
nitrogen balance throughout the gestation period. The experiment is
sufficient, however, to disprove Hagemann’s statement that a change
from a plus to a minus nitrogen balance takes place at the “‘moment”’
of conception. Copulation occurred on April 23, just a week from
the first appearance of menstrual blood, but bleeding did not cease
until about the 27th of April.

The influence of the menstrual period is, clearly, to cause an increased
retention of nitrogen. Schrader,?° who studied the nitrogen balance
through the menstrual period of several women, and Schéndorff,”
who followed the same, quite incidentally, in a single dog, were of
the opinion that this retention is a direct compensation for the loss
of blood. Potthast * and Ver Ecke,?? however, thought that some
other explanation is necessary. The various studies on the influence
of hemorrhages of other kinds on the nitrogen metabolism have given
discordant results.

20 SCHRADER: Zeitschrift fiir klinische Medicin, 1894, xxv, p. 72.

*t ScHONDORFF: Archiv fiir die gesammte Physiologie, 1897, lx, p. 395.

~% Porruast: Beitrige zur Kenntniss des Eiweissumsatzes, Inaugural Disser-
tation, Leipzig, 1887.

3 VER EckE: Loc. cit.

Metabolism of Development. 195

Hawk and Gies,”4 whose experiments were most painstakingly per-
formed, with due regard paid to the effects of anesthesia, state in
their conclusions that ‘‘external hemorrhage equal to from 3 to 3.5 per
cent of the body weight of dogs, was observed to cause the following
effects: in well-nourished animals in weight and nitrogen equilibrium
and fed continuously on a diet of constant composition, there was a
temporarily increased output of nitrogenous and sulphur-containing
products in the urine.” The same was observed by Bauer” and
Popiel.”* Jiirgensen,?’ however, in one of his experiments on a dog
from which the herhorrhage amounted to only 1 or 2 per cent of the
body weight, found a reduction of the N output, and Ascoli and
Draghi,”? who produced hemorrhages of from 220 to 475 c.c. in men,
witnessed a temporary reduction. Hence it is possible that small
blood losses up to something less than 3 per cent of the body weight
may cause reduction, while larger losses may produce an increased
nitrogen excretion.

It was impossible to collect the menstrual blood from the dog with
any degree of accuracy for analysis; consequently the loss of nitrogen
from this cause is unknown. It is not possible, however, that it
could have been equal to 3 per cent of the body weight (300 gm.).
Prussak 7° never found so much in the menstruation of women. It is
very probable, therefore, that the diminished output of nitrogen in
the urine of this dog during the three menstrual periods studied is
due, to some extent at least, to the small hemorrhage. At all events,
the influence of menstruation is exactly the reverse of what Bar
postulated.

The most pronounced effect is coincident with the most profuse
bleeding (Table III, week of April 17 to 23). The week immediately
preceding also shows the effect of approaching menstruation, and the
week following certainly was influenced, since bleeding did not cease
until the 27th of April. The week designated as “‘last of sexual rest”’

* HAwkK and Grrs: This journal, 1904, xi, p. 171.

* Bauer: Zeitschrift fiir Biologie, 1872, viii, p. 567.

26 PopreL: MAty’s Jahresbericht der Thierchemie, 1893, p. 505.

77 Cited by Ascorr and Dracut.

28 Ascori and Dracut: Berliner klinische Wochenschift, IQ00, XXXVii, Pp. 1055.

29 PrussAKk: Jahresbericht tiber Gynaekologie und Geburtshiilfe, 1899, p. 162;
see also HoprE-SEYLER: Zeitschrift fiir physiologische Chemie, 1904, xl, p. 545.

196 John R. Murlin.

exhibits a metabolism of about the same intensity as that of the
‘““second week of pregnancy.’’ From this experiment, then, we may
conclude that the protein metabolism immediately following concep-
tion is pitched at about the same level of intensity as i
preceding menstruation.

This is confirmed by the next experiment done with the same dog
six months later. With the thought that the minus nitrogen balance
observed in Experiments I and II might be due to the fact that the
protein of the beef-heart fed was too readily digested and absorbed,
that is, that the adult maternal tissues have no power to long retain
for the use of the embryo the nitrogenous materials, which ordinarily
are converted into urea, two feedings a day were given in this experi-
ment, and some milk protein was added in the form of a powder
known as “trumilk.”

Experiment IV. Dog A. Fourth pregnancy. Menstruation was expected
about the last week in October. Accordingly on October 1 the dog
was placed on a measured diet, which was given twice daily (at 9 A. M.
and 5 P.M.) until December 5. The diet follows:

too gm. beef-heart (3 %N) = 3.00 N and 120 cal.

5 cracker meal (2.08% N) = 1.14 N and 214.5

TO) caulk’? (6.4 %.N) = .64:N and. gz.

roe 4 Hard =o0.0 N and g4.

5 “ bone ash = 0.0 N and oo.

q. of csalt = 0.0 O.
Total = 4.78 N and 469.8 cal.

The daily intake therefore was 9.55 gm. N and 936.6 cal., of which
26 per cent was supplied by protein. On December 5 albumin appeared
in the urine, due to a cystitis induced by long-continued use of the
catheter, and it was thought best to discontinue the experiment for a
week. A change was made in the diet at the same time, to one very
nearly approaching that used in the previous experiments on this dog:
(see p. 186). The only difference was that the cracker meal used con-
tained more nitrogen. This change was made partly for the purpose
of giving a food whose ash was more strongly acid, so as to produce a
distinctly acid urine, thereby favoring the bladder condition, and partly
in order to compare the corresponding stages of different pregnan-

Metabolism of Development. 197

cies on the same diet. By December 11 the albumin had entirely
disappeared.

First menstrual blood was observed on November 2, and copulation
occurred on November 11, 12, and 13. Five puppies were born on
January 14, the sixty-fourth day.

TABLE IV.

EXPERIMENT IV. Doc A.t

BP eaci. in ee Ee rae rane awed bie
per week per week | per week | per week per day

1909
Oct. 3 to 9 56.812 5.44 62.25 +4.77 +0.68
Oct. 10 to 16 56.464 3.25 59.71 +7.21 +1.03
Oct. 17 to 23 58.62 4.10 62.72 +4.20 +0.60
Oct. 24 to 30 57.42 4.65 62.07 +4.85 +0.69
eee | 5487 3.29 58.16 48.76 +135
Een 50.02 4.44 54.43 +12.49 478
ea ea 58.31 2.88 61.19 ap 5073 +0.82
+a a to 27 59.16 3.43 62.59 +4,33 +0.62
ae. jack macud 66.53 3.94 70.47 —3.55 —0.51
ea mer 62.42 3.00(?)} 65.42 —0.392 —0.05
ee na en 58.02 S77 61.79 43.242 +0.46

1 The food contained each day 9.56 gm. N and about 940 calories (67 to 74 cal.
per kgm.).
* Food these two weeks contained only 9.29 gm. N per day.

The results given in Table IV present very much the same picture
as in the previous experiment. Up to the week of menstruation a
plus balance of about 0.6 gm. N daily prevailed. During the first
week of menstruation this figure was just about doubled, and during
the second week it was almost tripled, only to fall back again within
the next two weeks to the level which obtained before menstruation.
There can be no doubt of the influence of the menstrual processes.

The third week of the pregnancy, however, more nitrogen was

198 John R. Murlin.

found in the urine and feces than was contained in the food. It is
likely that this condition prevailed throughout the fourth week, for,
on the same diet, we find it persisting into the fifth. The sixth week
the retention of nitrogen had again set in, and without doubt would
have continued progressively to the end, had it been possible to follow
the experiment continuously. We shall return in the discussion later
to the matter of the negative balance.

Meantime the question which remains is whether the same level of
protein metabolism would be maintained after menstruation if the
dog were not copulated. If the prompt return to the same level as
before menstruation were due in any degree to conception, a duplicate
experiment in which copulation is not permitted ought to reveal a
difference. The literature does not contain any clear differentiation
of this kind. The experiment which follows was planned for this
purpose.

Experiment V. Dog A. Menstruation was expected early in May. Ac-
cordingly the dog was placed on a measured diet on the 23d of April.
With the exception of a slight difference in the N-content of the cracker
meal and of the milk powder used, the diet was precisely the same as in
Experiment IV. Just previous to this experiment the dog had been
used for a fasting experiment of eight days’ duration; hence the heavy
retention of nitrogen the first week. Possibly because of the previous
fast, the dog did not come into “heat” at the expected time, and to
hasten the onset she was liberated from the cage for a week and for
several days was taken for a walk of a mile or more in the hot sun. On
the 21st first signs of menstruation were observed, and the experiment
was resumed. The last trace of blood in the vagina which could be ob-
tained on a sponge of cotton was observed on the 7thofJune. Free
hemorrhage had ceased three days earlier.

The retention of nitrogen in this experiment is considerably greater
than in the previous one. This is doubtless attributable in part to
the previous fast (see protocol), which reduced the dog’s protein con-
dition considerably, and in part probably to the gradually rising tem-
perature of the springtime (the daily temperature record in the dog
room shows a gradual rise indoors as well as outdoors) as opposed to
the gradually falling temperature of the autumn under which the pre-
vious experiment was performed. Taking the third week as the base

Metabolism of Development. 199

level for the metabolism of sexual rest, the relative rise due to the
menstruation, however, is not quite so great as in Experiment IV on
the same diet, nor so great as in Experiment III done at the same time

« CABLE E MV
EXPERIMENT V. Doe A.!

N N Total N N N
in urine in feces in excreta} balance balance
per week. | per week. | per week. | per week. per day.

Week of
experiment.

48.30 3.58 51.89 +13.00 41.85
52.48 3.82 56.30 + 9.22 44.31
May 7 to 13 54.07 3.83 57.90 + 7.62 +1.09
ee se 44.53 2.91 47.44 +15.653 Be |
Pees wy fine 4 47.31 2.59 49,90 +15.62 +2.23
keto tt 47.71 2.63 50134 °| 1548 42.17

Wifiae 1) to 16 36.63 2.47 2010) 4) pee) tA

I
April 24 to 30
II

May 1 to May 6
iit

1 Food contained each day 9.27 gm. N and 940 cal. (70 to 85 cal. per kgm.).

2 Experiment was discontinued from May 14 to 21 (see protocol).

3 Through a mistake dog was given wrong food at one feeding. Nitrogen intake
was therefore 2.33 gm. less than it should have been.

of year but with a different diet. The return to normal, after men-
struation had entirely ceased, is also not so abrupt as in both the pre-
vious experiments. Not to lay too much stress on a single instance,
it nevertheless appears that the course of protein metabolism follow-
ing menstruation may be influenced to some extent by conception.

DISCUSSION.

We have seen that in the first experiment the dog lost nitrogen
from her body throughout the first four weeks of the pregnancy, and
in the second, throughout six of the nine weeks. It is now apparent
from Experiments III and IV that this condition in the first two
weeks of Experiment I, and probably for longer in Experiment II,
was due to the inadequate diet as regards energy supply. This, how-
ever, would not suffice to explain the negative balance of the third

200 John R. Murlin.

and fourth weeks, because, in the first place, in Experiment III the
same diet kept this dog in equilibrium under conditions which appar-
ently were identical in every respect except that she was in complete
sexual rest. In the second place, any deprivation of nitrogen due to
inadequate supply in the first two weeks would predispose to a nitro-
gen retention in the third and fourth. Again, in Experiment IV a
diet richer as regards both potential energy and nitrogen was not
sufficient to maintain equilibrium in the third week. These experi-
ments therefore fall into line with those of previous investigators in
exhibiting a preponderance of protein katabolism at a time which,
according to Bar, corresponds with the period of morning sickness in
pregnant women. To make this point clear all the experiments which
have been reported to date are brought together in the following table:

Author. | No. of Food cal. and | N balance by weeks of the pregnancy

Date. exp. em. N perkem. | 1. 2) 2 4

1891 | Hagemann I 821 1.01 — =—- = = = + f+ mt
rs Il 110 0.8 - — — = = + aborted

1903 | Jagerroos I 75 0.8 +. =) See eee
is II 106 1.84 |/+ -— =, + Fe ee eeeenee
il 75 131 |+ + A
i IV 70 0.2 ae ee

1907 | Bar I 112— | 0.64 4 ig

+ + + 4
|

1910 | Murlin I 70 MS |- - = =
.64
a5)

not continued

D IV 70 0.8 + + —- — — - not continued

1 These figures are based on the weight of the dog always at the beginning of the
experiment.

There seems to be no doubt, from this showing, that there is some-
thing characteristic about the occurrence of a negative balance in the

ES Py RE

Fk. ie ete a inte tee als Na in

Metabolism of Development. 201

first half of pregnancy in the dog. Even 106 calories of energy and
nearly 2.0 gm. N per kilogram per day were not sufficient in Jagerroos’
Experiment II to keep up a steady retention. What is the explanation
of this phenomenon? Hagemann believed that it is due to the inability
of the animal cell to transform one kind of protein into another without
loss of nitrogen. Jagerroos thought that some of these periods in his
experiments might be due to lack of appetite from the monotonous
character of the diet: Bar found actual sickness (vomiting) and
diarrhoea with loss of appetite in his Experiment III, but observed
nothing of the kind in any of the others, and in the experiments re-
ported in this paper the only instance of lack of appetite was in the
second week of the first and in the last week of the last experi-
ment where copulation did not occur. Jagerroos lays stress on the
fact that a “period of retention is always followed by a reaction,”
and Bar gives expression to the same thought in his “phase of satura-
tion’’ idea. He seems not altogether satisfied with this explanation,
however, for he mentions the possibility of proteolytic enzymes
(cytolysins) set free by the fetus and acting on the maternal proteins.
Inspection of the whole record as it now stands does not seem to favor
the ‘‘reaction”’ idea. In eight out of thirteen experiments tabulated
above a minus balance occurs in the fourth week of pregnancy.*°
-In seven certainly, and probably nine, of the thirteen a negative bal-
ance occurs in the fourth week. Two of the dogs in which it does not
occur were on very high protein, and should therefore, one would
think, show the “‘reaction”’ earlier, if that were the true explanation.

It is the writer’s belief that the presence of the embryos is in some
way responsible for the nitrogen loss. Whether this is due to enzyme
action it is at present impossible to say with certainty. Graefenburg *!
has recently found abundant evidence of the effects of proteolytic
enzymes produced by the ovum at the time of implantation. In fact,
there is no other rational way to account for the growth of the ovum
previous to the establishment of the placental circulation save by the
production of such enzymes. The placenta itself is now looked upon
as an organ which has a digestive function * and may be said to serve

30 Tt probably would have occurred in another — the III of the present series
. — making nine out of thirteen.

31 GRAEFENBURG: Zeitschrift fiir Geburtshiilfe und Gynaekologie, 1909, lxv,
pit;

202 John R. Murlin.

the purpose of restricting the action of enzymes to materials furnished
directly by the maternal blood instead of permitting them to act
indiscriminately. It appears from the investigations of Bonnet * on
the embryology of the dog that the placenta is fully matured at about
the time (thirtieth day) when rapid growth begins; or perhaps we
ought now to say that rapid development begins at this time because
the placenta has been fully established. It is probable, therefore,
that the period of greater nitrogen loss simply marks the culmination
of the more or less indiscriminate enzyme action, and that the recovery
therefrom marks the complete establishment of the placental func-
tions.** The plus balance occurring immediately after conception in
so many of the cases may mean only that the quantitative effect of
the enzymes is not yet sufficient to counterbalance the natural ten-
dency of the maternal tissues to maintain themselves in equilibrium.
The amount of nitrogen lost on this hypothesis might reasonably be
expected to bear some direct relation to the number of embryos
formed, and this, as we have seen on page 192, is the case.

Space forbids fuller discussion of this point at present, but it should
be noted that the biological reason for the production of enzymes by
the developing ovum is identical with the reason for the existence of
such enzymes in the adult stomach, namely, to enable the organism
to build foreign materials into its own type of protoplasm.

If, as is generally supposed, though by no means definitely proved,
the enzymes produced by the ovum can act on the maternal proteins
and the ovum can establish placental (enzyme) connection with the
uterine wall of its own physiological species only, the enzymes so
produced become an essential part of the mechanism of heredity.
We should then be obliged to say that the reason why dog ova grow
into young dogs is partly at least because they can grow at the expense
only of old dog proteins. At all events, the existence of a negative

8 See L. Zuntz: Ergebnisse der Physiologie, 1908, vii, p. 430.

88 BONNET: Anatomische Hefte, rer Abtheilung, 1902, xx, p. 477.

3 According to GROsSER (Entwickelungsgeschichte der Eihaiite und der
Placenta, Wien u. Leipzig, 1901, p. 118), two stages of placentation are to
be distinguished among the Placentalia including man, — an earlier in which the
nutrition is mainly embryotrophic, and a later in which the nutrition is mainly
by diffusion between the two circulations. In the woman it is believed that the
disappearance of morning sickness is coincident with the complete formation of
the placenta. See, e. g., Woop, Washington medical annals, July, 1908.

Metabolism of Development. 203

balance at a certain stage of the gestation in nearly every case of
pregnancy in the dog so far studied fits in well with the hypothesis
that such enzymes are at work, and that their influence may extend
into the whole maternal system, causing the destruction or ‘‘mobil-
ization” of more proteins than are needed by the embryos.

Explanation of the plus balance in the latter half of pregnancy is
a much simpler matter. This condition has been found in the last
three weeks at least of every case yet investigated which has reached
term. Bar was the first to show that the retention of nitrogen for
this period is parallel with the needs of the embryo, and Hoffstrém
found that the same was true in his single case also for the retention
of phosphorus, sulphur, calcium, and magnesium, although a large
quantity of each substance was not delivered at birth, but was re-
tained by the mother’s own body as a “reserve fund.”

It is to be supposed that, as growth proceeds and more and more
protein materials are diverted from the maternal system to the fetal

“* Tom 2PM. 12 12M SAM 3PM

Ficure 2. — The daily curves of nitrogen excretion in the urine of the Dog A, in the last
week of fourth pregnancy (unbroken line) and in sexual rest (dotted line). In each
case 8.9 gm. N were fed. The abscisse indicate tenth grams of nitrogen, and ordinates
are hours.

system, the maternal tissues react by holding back the nitrogenous
bodies which otherwise would find their exit from the body mainly
as urea. The question arises whether the rate of excretion of the total
nitrogen, that is, its daily curve, would be affected by the presence
of the embryo. To determine this point Dog A in the last week of
her fourth pregnancy (Experiment IV), and while on the same diet
as had been used up to the sixth week, was catheterized every three
hours for twelve hours immediately after feeding on both January 11

204 John R. Murlin.

and 12, 1910. On January 14 five puppies were born. The experi-
ment was repeated in every particular after the dog had been several
days on the diet again on March 24 and 25, two weeks after lactation
had ceased. The results are plotted in the two curves represented in
Fig. 2. .

It is seen that the curve for the pregnant condition runs below that
for the non-pregnant condition at a fairly uniform distance. In other
words, the amount of nitrogen diverted to the embryos from a given
diet is about the same from hour to hour.®

The curve is also significant in that it shows that the great bulk of
the daily intake of nitrogen is removed by the action of the liver
and kidneys within from nine to twelve hours, whether the animal
be pregnant or not. This result obviously favors the idea of frequent
small meals of protein for the pregnant organism rather than one
large meal.

SUMMARY AND CONCLUSIONS.

1. The nitrogen balance was followed through two complete periods
of gestation beginning with copulation, through parts of two others
beginning several weeks previous to menstruation, and through one
period of menstruation which was not followed by copulation.

2. In the first two experiments the diet contained 70 cal. and
0.65 to 0.75 gm. N per kilogram per day. From the first pregnancy
one puppy was born, and the result was a net gain of 8.69 gm. N to the
mother’s body. From the second four puppies were born, and the
result was a net loss of 55.6 gm. N from the mother’s body. Up to
the fourth week in these two experiments (on different dogs) the
amounts of nitrogen lost from the mother’s body were proportional
to the weights of puppies delivered.

3. The effect of the menstruation is to cause a retention of nitrogen,
which may be explained, in part at least, as a compensation for the
amount of blood lost.

4. The results of these experiments support the idea that nitrogen
loss from the mother’s body is characteristic of the first half of normal
pregnancy in the dog, particularly of the third and fourth weeks.

% See Proceedings of the Society for Experimental Biology and Medicine, 1910,
vii, p. 126.

Metabolism of Development. 205

This is probably due to the action of proteolytic enzymes produced
by the embryo and not yet limited by the placenta in their action to
the maternal blood.

5. Nitrogen retention has been found in these, as in all other ex-
periments, in the last half of the pregnancy. The curve of nitrogen
elimination in the urine shows that the retention in the last week of
pregnancy is fairly even from hour to hour.

ON THE NERVOUS MECHANISM OF THE RIGHTING
MOVEMENTS OF THE STARFISH.

By A. R. MOORE.
[From the Rudolph Spreckels Physiological Laboratory of the University of California.]

T was first noted by Romanes! that the detached arm of the star-
fish, when placed upon its aboral side, spontaneously righted
itself. Afterward Loeb? showed that if the oral nerve ring be cut in
two places, all the arms attach to the bottom, so that the animal is
not able to right itself. Loeb concludes from this experiment that the
righting of the normal starfish is possible because inhibiting impulses
set up by the arms which have the securest attachment pass through
the oral nerve ring and cause the attaching of the remaining arms to
be inhibited.

Opposed to Loeb’s analysis is the “‘trial and error” hypothesis ad-
vocated by Jennings and others. According to this hypothesis the
starfish possesses a sort of consciousness whereby the animal is enabled
to co-ordinate its movements so as to accomplish a desired act; all
movements not contributing toward the success of the animal’s pur-
pose are sooner or later abandoned as “errors,” thus allowing only
“successful” trials to persist. In support of such a view Jennings ®
has attempted to show experimentally that the starfish can learn by
experience to right itself in a certain way. The following quotation
from this writer indicates clearly the point of view taken by the “‘trial
and error” school of naturalists. Speaking of the efforts of an inverted
starfish to right itself, he says: ‘‘As soon as these one or two arms have
been successful the others cease their efforts; the attached arms turn
the body over. If all the arms attempted to turn the body over at the

1 Romanes: Jellyfish, starfish, and sea urchin, New York, 1808, p. 204.
2 LoreB: Comparative physiology of the brain, New York, 1900, p. 63.
3 JENNINGS: University of California publications, Zodlogy, November, 1907,

p. 156.

207

208 A. R. Moore.

same time, in other words, if there were no way of recognizing * suc-
cess in the trial, the animal could not right itself.’ ®

Georges Bohn,® however, has called attention to the significant fact
that old and mature starfish make a great many more “trials and
errors”’ in the act of righting than do the young starfish of the same
species; whereas if the ‘‘trial and error” hypothesis were true, facility
should be the product of experience, and the old starfish should be
more expert than their offspring in the common act of turning over.
The truth of Bohn’s observation is apparent to any one who has worked
with starfish of various sizes.

For the purpose of gaining more light on this question I have made
observations on a large number of Asterina miniata and Asterias
ochracea. I find that there is a common or normal way in which a
starfish rights itself.” For example, in Fig. 1, any two arms, say A and
B, face each other ventrally and attach; C and E may have obtained
a hold, but soon withdraw their tube feet, rise orally, pulling D with
them, and thus complete the somersault. During this process D may
either remain passive or hold to the bottom with its tube feet and so
retard the righting. Taking Loeb’s view that C and £E have their
attaching inhibited by impulses sent from A and B and move upward
at once, it is significant that D is affected by such impulses very
slightly if at all. Repeated observations of the passiveness of D led
me to think that the inhibiting impulses arising from A and B lose so
much in effectiveness, as they travel away from the point of origin,
that they affect only adjacent arms sufficiently to cause movement.
That is to say, an inhibiting impulse from A rouses only Z, and the
impulse from B causes only C to act. If such were the case, we should
have additional proof that the hypothesis of Loeb is correct and that
the righting of the starfish is accomplished by the simplest sort of
mechanism.

This rapid decrease in the effectiveness of a nerve impulse as it
travels away from its point of origin is a phenomenon familiar to physi-
ologists in the spreading of impulses, and was described by Pfliiger and

4 The italics are ours.

5 JENNINGS: Carnegie Institute publications, 1904, p. 245.

® Bown: Bulletin de l'Institut général psychologique, Paris, 1908, p. 95.
7 Moore: Biological bulletin, rg10, xix, p. 237. |

kighting Movements of the Starfish. 209

later by Sherrington ® with reference to the vertebrates. According
to these authors, ‘“‘the degree of reflex spinal intimacy between affer-
ent and efferent spinal roots varies directly as their segmental prox-
imity.” To illustrate, Lee,’ speaking of the shark, says: ‘‘The power
of the fins to make compensatory movements [due to the stimulation
of the eighth cranial nerve] diminishes in an antero-posterior direction,
the most delicately responsive [fins] being the two pectorals and the
anterior dorsal. Even these do not in these respects equal the
eyeballs.”

FicurE 1.—Showing oral nervous sys- FicurE 2.— Showing attempt at righting
tem of the starfish. after cut is made.

In order to test whether such were also true of the nervous system
of the starfish, I cut the oral nerve ring of one of these animals between
the arms B and C at M (Fig. 1). In the case of a starfish so treated, if
A and B face each other ventrally and attach, E is inhibited, but C
either remains passive or attaches and holds to the bottom until
pulled loose by HE and D. In addition to A and B attaching co-
ordinately, C and D may also attach and face each other ventrally. In
this case the starfish remains for some time in the position shown in
Fig. 2, until finally # pulls either A or D loose and thus decides the
course of the righting. The time required for the righting to take
place in this way is two or three times as great as that needed before
the cut was made. Of course if B and C should attach and inhibit A

8 SHERRINGTON: The integrative action of the nervous system, New York,
1906, p. 158.
® LEE: Journal of physiology, 1894, xv, p. 321.

210 A. R. Moore.

and D, then the righting would be accomplished as in the normal
starfish. Likewise, if A and £ or E and D attach co-ordinately, their
connection with the adjacent arms is unimpaired and the righting is
per‘ormed in the normal way. But when the nerve ring is cut at M,
if A and B attach co-ordinately, C is inhibited very slightly or not at
all and does not rise orally to assist in the righting as in the case of the
un njured animal; on the contrary, C acts as an independent arm and
renders no assistance by co-ordinated movement.

Numbers of control experiments have been performed in which the
cut between the arms was made from the outside so as not to injure
the oral nerve ring. In these cases the co-ordination of the arms was
not interfered with in the least, and the righting was accomplished in
normal time. This shows that the interference in co-ordination
caused by making one cut in the oral nerve ring is not due to general
injury to the animal.

The experiment was recently given to a class of eighteen students in
nerve physiology, to perform. The material consisted of small speci-
mens of Asterias ochracea. Although the students had no idea of the
result expected, their observations entirely harmonized with those
which I had previously made.

In making the one cut in the nerve ring of the starfish, none of the
arms of the animal are isolated nervously, all are connected; but in-
stead of the old circular connection there now obtains a linear system
comparable to the nerve cord of worms and crustacee. If conscious
judgment ever were present in the starfish, it scarcely could have been
destroyed by our operation, because each part of the animal still has
perfect nervous connection with every other part. The fact that the
arm C does not co-ordinate with the active arms proves that direct
nervous connection across the cut is necessary for co-ordination; that
C can only be caused to move “intelligently” by an impulse from an
adjacent arm. ‘This is different from saying that the co-ordinated
movements of the starfish are due to judgment exercised by a “‘psy-
choid” entity. Hence it is evident that there is no nervous centre in
the starfish, forany oneof the five arms may give rise to impulses which
strongly affect only adjacent arms. These impulses rapidly diminish
in strength as they travel away from their point of origin, and their
effect is usually imperceptible in the movements of other than adja-
cent arms. The simple nervous mechanism herein described is suffi-

Righting Movements of the Starfish. Art

cient to account for the co-ordinated righting movements of the star-
fish. The hypothesis which seeks to account for the co-ordinated
movements of the starfish by postulating judgment, decision, entelechy,
etc. for that animal is unnecessary, and since each arm is capable of
initiating impulses which bring about co-ordination, the hypothesis
of a psychoid unity or centre for the starfish is untenable.

In conclusion, I wish to thank Professor Maxwell and Dr.
Burnett for valuable criticisms and suggestions.

THE EFFECT OF LESIONS OF THE DORSAL NERVE
ROOTS ON THE REFLEX EXCITABICIDY OR) ta
SPINAL CORD.*

By CLYDE BROOKS.
[From the Hull Physiological Laboratory of the University of Chicago.]}

INTRODUCTION.

| Bes ae CARLSON, after some preliminary experiments in

which he observed, especially in the dog, temporary depression
of the crossed reflexes following transection of the dorsal nerve roots,
asked me to determine whether the changes in the reflex excitability
of the cord following lesions of the dorsal nerve roots were related in
any way to those following the transection of the cord itself.

After a somewhat prolonged investigation extending over most of
the species of animals commonly employed in laboratory experi-
ments and also including some that are not ordinarily used, the re-
sults indicate that there is a parallelism between the effect upon
reflex excitability following transection of the dorsal roots and that
fol’owing transection of the cord itself. In all those animals which
show depression of the reflex excitability of the spinal cord following
transection of the cord, the transection of the dorsal roots is followed
by a depression- comparable in amount and duration. Further, some
animals which are usually said to show no spinal shock do show it in
a slight degree when more delicate methods are employed for its de-
tection; and they also show a parallelism between the amount of
depression caused by section of the cord and that caused by section of
the dorsal roots.

These results are of interest because they bear upon the mechanism
of spinal shock; for if section of the dorsal roots alone may cause

* A preliminary report was published in Science,! and a brief report was also
made before the American Physiological Society at the Baltimore meeting in
December, 1908.

212

Effect of Lesions of the Dorsal Nerve Roots. 213

phenomena simulating those of spinal shock, doubt is cast upon the
hypotheses which assume that spinal shock is due only to the interrup-
tion of the reflex arcs themselves; section of the dorsal roots does not
transect the reflex arcs of the crossed reflexes, and yet apparently
produces shock.

STATEMENT OF PRESENT OPINIONS.

It is very well established that transection of the dorsal roots in
the frog,” dog,’ or,monkey ‘ is followed by a decrease in reflex irrita-
bility of the spinal cord and loss of tonus of the limbs. But depression
does not invariably follow section of sensory roots. For example,
Merzbacher ® has shown that there is good tonus and movement of
the tail of the dog after section of the sensory roots of the cauda
equina; and Trendelenberg ® has found that there is a slight increase
in tonus and reflex irritability of the pigeon’s wing after section of its
sensory nerve roots.

At a very early period spinal shock was known to occur in adult
animals‘ and to be absent in very young animals.* At present,
while there is very good agreement as to the occurrence of spinal
shock, there is a good deal of difference of opinion as to the mechanism
of it. The Goltz view ° seems to be largely superseded by the view
that spinal shock is the result of disruption of nervous connections
between the distal segment of the cord and the higher parts of the
central nervous system.’° As to the functions of these connections,
nothing is directly known. Some!” are inclined to what has been
called the ‘‘anatomical view,” ” which supposes that normally the
reflexes travel with less resistance through the upper part of the cord
and central nervous system, and that transection of the cord breaks
the normal reflex paths; while others ® take the ‘“‘functional view,” ”
which supposes that the severance of connections of the distal seg-
ment of the cord from its higher relations either leaves it bereft of
influences which normally tend to maintain the normal state of tone
or reflex irritability, or else that it initiates processes in the lower seg-
ments which cause depression of its tone and irritability.

More detailed explanations have been suggested for the functional
view. Moore and Oertel” regard spinal shock as the result of the
removal of higher regulatory influences which are normally exercised

214 Clyde Brooks.

upon the cord by the higher parts of the central nervous system.
Von Monakow “ has suggested that a “‘Diaschizis” takes place be-
tween his “‘Schaltzellen”’ and blocks the impulses. To Sherrington
spinal shock suggests a loosening of the nexus between links of the
mechanism composing an arc; a defect in transmission at the synapse.
According to Munk’s conception “ the irritability of the motor cells
is constantly kept up to the normal height by two influences: one,
the ‘“‘centrogenous,” is from the central nervous system; the other,
the ‘‘neurogenous,” is from the periphery va the sensory roots. Ac-
cording to this, transection of the cord cuts off the centrogenous in-
fluence, which results in loss of reflex irritability, or spinal shock.
Recovery occurs by the gradual restoration of reflex irritability
through the medium of the neurogenous influence.

Babak !”7 has made a systematic investigation of spinal shock in
the frog and has considered the problem from the view-point of on-
togeny and phylogeny. He found spinal shock absent in the tadpole
and developing gradually during and after metamorphosis until in the
adult frog it reaches the maximum. He also found that there is more
shock when the transection is made in the upper part of the cord,
especially at the level of the first or second vertebra, and less when
made in the lower part, especially the level of the fourth to the sixth
vertebra. These results are very difficult to harmonize with the ana-
tomical view. Pike 1* also has made studies from this view-point, and
has questioned Babak’s conclusions. Babak, however, considers that
Pike’s criticism is due to misunderstanding his position.”

EXPERIMENTAL METHODS.

In all the species examined, the same general plan of experimenta-
tion was followed. The animals were fixed or suspended in some suit-
able way while the reflexes were elicited by mechanical, thermal, or
electrical stimulation. In most of the animals fine wire electrodes
were fastened to the skin of one limb and connection made with an
induction machine so that electrical stimuli could be applied to the:
skin. After obtaining the normal reflex reactions the cord was tran-
sected high up and the reflexes measured from time to time to note
whether there was any change in excitability. After recovery from
the effects of transection of the cord the dorsal roots of the limb oppo-

Effect of Lesions of the Dorsal Nerve Roots. 215

site to the one to which the wire electrodes were attached were tran-
sected. If the electrodes were attached to the right hind leg, the
dorsal roots were transected on the left side of the lumbar enlarge-
ment of the cord. After section of the dorsal roots the reflex irrita-
bility was noted from time to time until recovery had occurred.
Modifications were made in order to adapt the plan to certain species.
In the frog and the turtle, which were the species most used, the graphic
method was also employed, and the stimuli were thrown in by an
automatic key connected with a metronome. The following are the
various species employed: dog, cat, rabbit, guinea pig, chicken,
pigeon, alligator, snapping turtle, terrapin, large bullfrog, small leop-
ard frog, and necturus. Some of these (cat and dog) were studied in
the very young as well as in the adult animal. Usually ten or twelve
animals of each species were used, but the number of snapping turtles
was about sixty, of frogs about thirty, and of pigeons about twenty.

EXPERIMENTAL RESULTS.

The dog. — As mentioned in the introduction, it was Professor
Carlson’s results especially on the dog that led to this research. Other
experiments have confirmed the first results showing that section of
the dorsal roots in the dog is followed by temporary depression of the
crossed reflexes, lasting for an hour or more. This depression is
comparable to that caused by transection of the spinal cord itself.

The dogs were anesthetized with ether and the spinal cord exposed
and transected in the lower cervical or the upper dorsal region. The
dorsal roots were exposed on one side of the lumbar enlargements of
the cord. The reflexes were tested at intervals before and after the
operation. After recovery from the preliminary operation the wound
was reopened and several dorsal roots cut, and the cross reflexes were
tested before and after the operation.

The following abbreviated protocol shows the result of such an
operation on an adult dog:

September 18, 1908. — 11.00 A.M. Small female dog in good flesh and

vigor. Reflexes normal. ;

11.10 A.M. Anesthetized with ether. Exposed the last thoracic and

first three lumbar dorsal nerve roots on the right side. Closed the
wound,

216 Clyde Brooks.

11.25 A.M. ‘Transected the cord in the upper dorsal region. Re-
moved the anesthetic and dressed the wounds.

11.35 A.M. No reflexes obtained by pinching or striking the hind
legs. Legs hang limp and feel soft and relaxed.

1.10 P.M. Cross reflexes on pinching toes or tail are good. Same
side reflexes also good. Placed dog in hospital.

September 19, 1908. — 10.45 A.M. Placed dog on its abdomen with its
hind legs hanging over the edge of the table. Removed dressing.
Wound in good condition. Opened wound. Exposed four dorsal
roots.

11.20 A.M. Reflexes good when tested by pinching or striking hind
legs or tail.

11.30 A.M. Cut four dorsal roots on right side (last thoracic and
first three lumbar).

11.40 A.M. No crossed reflexes; only local contraction of muscle
caused by direct stimulation with electricity or sharp blow. Muscles
appear to have lost tone.

12.15 P.M. Apparently a little more tone in muscles now. The left
hind leg contracts on striking or pinching it hard.

12.40 P.M. There is some slight movement in the right leg on strik-
ing or pinching the left sharply.

1.15 P.M. Good cross reflex (movement of right leg) on pinching and
on striking the left leg.

1.35 P.M. Reflexes still improving. Tonus returning.

In the experiments on the dog the depression of the cross reflexes
was much the same whether caused by section of the cord itself or
by section of the dorsal roots. ‘The results of similar experiments
upon the young puppy are given later.

The cat.— The methods used in the experiments upon the cat
were similar to those used upon the dog. The cats were etherized,
the spinal canal was opened, and the cord exposed. The cord was then
transected in the thoracic region and the wound closed. The reflexes
were observed from time to time. After recovery on the following
day, or later, the wound was reopened and the several dorsal roots
were transected, and the effect upon reflex excitability noted by pinch-

ing or striking the hind legs and tail, and also by stimulation with the —

induced interrupted electrical current. The following notes from one
of the protocols shows the chief feature of the results:

|
|

Effect of Lesions of the Dorsal Nerve Roots. 217

September 13, 1908. — Young adult female cat.

2.20P.M. Reflexes normal. Etherized.

2.25 P.M. Opened spinal canal. Exposed dorsal roots in lumbar
region.

2.27P.M. ‘Transected cord at the level of the last cervical or the
first thoracic vertebra.

2.30 P.M. Closed wound. No reflexes by pinching or striking.
Hind legs limp. ;

2.35 P.M. No reflexes. Legs limp.

3.05 P.M. Electrical stimulation of hind legs causes no movements.

3.10 P.M. Slight crossed reflex on electrical stimulation of the left
hind foot.

3.15 P.M. Crossed and homolateral reflexes good. Tonus in legs
improved. Stimulation of one hind leg causes toes of other hind leg
to clench.

7.00 P.M. Good reflexes by pinching, striking, or stimulating with
electricity. Legs in very poor tonus. Put animal in hospital in good
condition.

September 14. — 8.00 A.M. Cat in good condition; mews and purrs.
Pinching its tail causes lively kicking in both hind legs. Picking up
one leg causes kicking. Pinching toes causes several successive kicks
in that leg and one or two of the opposite hind leg. Crossed and
homolateral reflexes more readily elicited than normally.

September 15. 8.00 A.M. Condition of cat remains good.

September 16. — 2.30 P.M. Condition of cat good.

2.35 P.M. Removed dressing from lumbar wound. Wound clean.
Reopened wound and exposed dorsal roots.

3.15 P.M. Cut five dorsal roots on the right side at level of the lum-
bar enlargement.

3.20 P.M. No crossed reflexes; cat lies still, breathes regularly. No
crossed reflexes from pinching or striking the left hind leg.

4.05 P.M. Reflexes have returned. Pinching tail causes kicking of
both legs. Pinching left leg causes right to kick.

5.00 P.M. Crossed and same side reflexes are stronger.

6.00 P.M. Crossed and same side reflexes good.

From these experiments we find that in cats there is temporary
depression of the crossed reflexes, either by transection of the cord or
by transection of the dorsal roots, but the depression is not so long
continued as in the dog.

The results of observations on the young kitten are given later.

218 Clyde Brooks.

The frog. — A large number of experiments have been made upon
the frog, partly because it has long been the object upon which such
research has been made, and also because the frog shows more shock
than most of the other cold-blooded animals. In these experiments
very large Indiana bullfrogs were most frequently employed, but the
ordinary leopard frog was also used a good deal. The experiments
were conducted on the same plan as those described above for the dog
and the cat. The reflexes were tested from time to time during the
experiment by pinching with forceps or striking with some light in-
strument. Then transection of the cord was made. After recovery
from this operation the dorsal roots of one hind leg were transected.
It was found that transection of the cord in this species resulted in
depression of the reflexes lasting some minutes, also that section of
the dorsal roots from one hind leg was followed by a similar temporary
depression of reflex irritability. These results are in accord with those
of most of the writers mentioned above.

The following brief protocol shows the effect upon the cross reflexes
of section of the dorsal roots of the large Indiana bullfrog:

July 28, 1909. — Large specimen in excellent condition. Reflexes normal.

3.00 P.M. Opened spinal column. Exposed dorsal roots of left hind
leg. Reflexes still good.

3.07 P.M. Cut four dorsal roots on left side of lumbar enlargement.
Left leg hangs limp.

3.08 P.M. No cross reflexes.

3.13 P.M. Cross reflexes very slight.

3.19 P.M. Good cross reflexes.

3.27 P.M. Cross reflexes good.

This experiment shows depression of the cross reflexes lasting for
almost twelve minutes before complete recovery following transection
of the dorsal roots. Transection of the cord itself in other specimens
gave similar results. :

The alligator.— A number of observations were made upon alli-
gators which were about 75 cm. long. They were suspended in a:
sling and their reflexes tested by pinching or striking, and also by
applying fine wire electrodes to the skin of the thigh and stimulating
with interrupted induced current.

In normal intact alligators it was difficult to distinguish the exact

Effect of Lesions of the Dorsal Nerve Roots. 219

strength of stimulus just sufficient to give a distinct cross reflex, be-
cause the animal apparently inhibited all responses until a certain
strength of stimulus was reached, when it quickly began to flounder
and struggle vigorously. But after transection of the cord the cross
reflexes were more readily obtained. It was further observed that
these animals were more susceptible to hemorrhage or asphyxia than
were turtles or frogs. In some of the experiments when hemorrhage
was profuse, in the first ten or fifteen minutes there was a heighten-
ing of reflexes, probably due to anemia or asphyxia, followed by a
gradual persistent depression which ended in complete disappearance
of reflexes in about seventy to eighty minutes.

The following extracts from a protocol show the slight depression
caused by section of the cord or dorsal roots:

April 14, 1909. — Young alligator 70 cm. long. Lively and vigorous.
9.00 A.M. Applied the electrodes to the skin of the posterior part
of the thigh of the left hind leg.
9.31 A.M. Coil at 16.5 cm. No movement.
Coil at 14.5 cm. Flexion of same leg (left).
Coil at 8.0cm. Vigorous general movements.
9.50 A.M. Transected cord just below line drawn across the back
at the anterior border of scapula.
Coil at 12.9 cm. gave reflex on same leg.
Coil at 8.0 cm. gave reflex on same leg.
Coil at 7.0 cm. gave general movements.
9-55 A.M. Retransected cord at 2 cm. below first transection.
Coil at 12.4 cm. gave reflex same side.
Coil at 8.3 cm. gave cross reflex to right leg.
Coil at 6.8 cm. gave general movements.

After recovery, section of the dorsal roots was followed by a corre-
sponding temporary depression of the cross reflexes.

The chicken and the pigeon. — These two species may well be classed
together, as they gave very similar results. Experiments upon these
have been made by the simple method of pinching and striking the
toes or the abdomen and observing the resulting movements; then
transection of the cord under light ether anesthesia, or under local
' anesthesia, or in some cases (the chicken) the bird was quickly de-
capitated without any anesthesia and artificial respiration employed.

220 Clyde Brooks.

The result of such experiment has been that there is no apparent de-
pression of reflexes when tested by the method, whether from transec-
tion of the cord or by section of the dorsal roots. In fact, there often
seemed in the pigeon to be increase in irritability.

Necturus.— The experiments upon the necturus were made by
observing the reflexes obtained by pinching or striking or stimulating
by interrupted induction shocks the limbs, the body, or the tail of
the animal, and then making high section of the cord. The observa-
tions by these methods showed no change of reflex irritability so long
as the general condition of the animal remained good.

The kitten and the puppy.— As noted by many other observers
in very young animals, no depression of reflex irritability occurs upon
transection of the cord or dorsal roots, and indeed when ordinarily
tested this appears to be the case; but when measured stimuli are
used even in the very young puppy or kitten there is observed some
distinct though very slight temporary depression of the reflexes.

In these experiments the electrodes were applied to the skin of the
ball of the foot of one hind leg. After measuring the stimulus just
necessary to cause a distinctly perceptible reflex on the same side and
_also that just sufficient to produce a reflex to the opposite hind leg,
transection of the cord was performed. After allowing time for re-
covery the dorsal nerve roots were exposed and transected on the side
opposite the leg with the wire electrodes applied to the feet. As the
kitten and the puppy gave very similar results, they may be considered
together.

The following abbreviated protocol shows the results of such experi-
ments upon a kitten:

March 26, 1909. — Young maltese kitten about nine days old.

2.05 P.M. Intact animal; gave cross reflexes with secondary coil at
11.65 cm.; at 10 cm., strong crossed reflexes.

2.20P.M. Etherized kitten and exposed cord. After allowing thirty
minutes for recovery from the anesthetic the reflexes were again tested.

2.55 P.M. Coil at 11.5 cm. gave very slight cross reflex.

Coil at 11.0 cm. gave stronger cross reflex.

3.05 P.M. ‘Transection of the cord in the upper thoracic region.
Kitten kicks vigorously with both hind legs for a few seconds, then legs
hang limply down.

3.07 P.M. Coil at 9.5 cm. gave slight cross reflex to tail.

Effect of Lesions of the Dorsal Nerve Roots. 221

Coil at 9.0 cm. gave distinct cross reflex to right hind leg.

3.23 P.M. Coil at 10.0cm. Cross reflex.

3.48 P.M. The dorsal roots on the right side of the lumbar enlarge-
ment cut. As soon as possible after this the reflexes were tested as
before.

3.50 P.M. Coil at 9.0cm. Distinct though very slight cross reflex.

Coil at 8.0 cm. Strong cross reflex.

4.10 P.M. Not quite complete recovery.

The rabbit and the guinea pig.-— Since the results obtained on rab-
bits and guinea pigs are very similar, they may be reported together.
The cord was either transected under local anesthesia and afterwards
the dorsal roots sectioned, or else the animals were decerebrated under
ether and transection of the cord and of the dorsal roots made after
recovery from the first operation. The reflexes in these animals are
only slightly and temporarily depressed by section of the cord or dor-
sal roots. It was only by using delicately graduated strengths of
stimuli that the depression was noticeable.

May 15, 1909. — 3.00 P.M. Guinea pig of large size and in good condition.
3.25 P.M. Coil at 10.8 cm. Cross reflexes.
3-38 P.M. Transected cord. Ethyl chloride anesthesia. Appar-
ently good tonus in legs immediately after cutting cord.
3.39 P.M. Coil at 10.1 cm. Homolateral reflex.
3.45 P.M. Coil at 9.5 cm. Crossed reflex.

After recovery three dorsal roots on the right lumbar enlargement
were transected. This was followed by a similar temporary depression
of cross reflex irritability.

Experiments on the rabbit gave very similar results. There ap-
peared to be very slight loss of tonus as judged by the position and
resistance of the limbs in handling. There was a slight but distinct
temporary depression of irritability as judged by the strength of
interrupted induction current necessary to cause a cross reflex.

The turtle. —It is upon the turtle that the greatest number of
our experiments have been performed. More than sixty large lake
snapping turtles and a half dozen terrapin have been used. In the
snapping turtle, as in the kitten and puppy, the earlier experiments
which were made by pinching or striking the legs and observing the
resulting reflex movements, no depression was discernible after tran-

222 Clyde Brooks.

section of the cord or section of the dorsal roots. But later, by using
delicately graduated electrical stimuli, a distinct though slight fall in
cross reflex irritability was observed. These experiments were made
at room temperature. The stimuli were from an induction machine
with about fifteen to thirty interruptions per second. The stimuli
were made by a key operated by hand and were one second in dura-
tion, as measured by a second pendulum and a stop watch. The fol-
lowing abstract of one of the protocols illustrates the results:

March 17, 1909. — Large lively snapping turtle. Room temperature, 27° C.
8.39 A.M. Intact animal.
Coil at 9.0cm. Homolateral reflex.
Coil at 8.6 cm. Cross reflex.
8.42 A.M. Transected cord just below the medulla oblongata.
8.44 A.M. Coil at 8.0cm. No response.
8.45 A.M. Coil at 7.0 cm. Cross reflex.
8.48 A.M. Coil at 8.5 cm. Homolateral reflex.
9.16 A.M. Coil at 9.0cm. Homolateral reflex.
Coil at 7.0cm. Good cross reflex and general movements.
Placed in ice box, leaving electrodes attached to legs.
March 18. — 10.00 A.M. Coil at 8.6 cm. Homolateral reflex.
Coil at 8.0cm. Crossed reflex.
10.17 A.M. Cut seven dorsal roots to right side of lumbar
enlargement.
10.20 A.M. Coil at 7.0cm. No reflex movements.
10.23 A.M. Coil at1.ocm. No reflex movements.
* 10.26 A.M. Coil ato.o cm. Homolateral reflex.
10.41 A.M. Coil at 5.0cm. Homolateral reflex.
10.45 A.M. Coil at 8.5 cm. Homolateral reflex.
Coil at 8.2 cm. Crossed reflex.

A series of experiments was performed upon the turtle based upon
the above method, but registering the reflex movements graphically,
and also using the key connected with a metronome in order to obtain
more exact duration of the interrupted electrical current.

In beginning the experiment the normal turtle was placed upon the
stand and rigidly clamped in position. Each hind leg was attached
to its writing lever and its electrodes. The levers were adjusted to the
smoked drum so that the right leg traced about 3 cm. from the top of
the drum while the left leg traced about 3 cm. below that. Flexion of

Effect of Lesions of the Dorsal Nerve Roots. 223

the hind limbs caused a downward stroke of the writing points. When
both hind legs were completely relaxed and quiescent, a signal magnet
which indicates the period of stimulation was adjusted so that it was
on a level with the base line of the right leg, and a Jaquet instrument
marking time in seconds was adjusted so that it traced the base line
for the left leg. By this arrangement there were recorded on the
drum the homolateral and cross reflex movements and tonus changes
of the normal intact turtle. The tracings showing the effects of de-
capitation and of section of the dorsal roots were recorded and com-
pared with the normal. After decapitation a syringe bulb with tra-
cheal cannula and T-piece was connected with the trachea of the turtle
and artificial respiration was given. The spinal cord was cut across
with a narrow sharp chisel so that very little blood was lost. Later,
after the effects of transection of the cord were recorded on the drum,
the dorsal roots of the right hind leg were transected and the effect
upon the reflex irritability of. the cord recorded on the drum. The
results of this method agree with those obtained where the stimula-
tion was regulated by hand and the resulting reflex movement judged
by the eye.

These tracings also show the changes in tonus of the muscles of the
hind limbs that accompany the changes in irritability.

SUMMARY.

The above results as a whole show a parallelism between the changes
in the reflex irritability of the cord following transection of the spinal
cord itself, and those following transection of the sensory nerve roots
of the cord. The dog, cat, and frog show a marked depression fol-
lowing either of these operations. The alligator shows some depression.
The rabbit, guinea pig, pigeon, chicken, or the turtle, shows no depres-
sion when tested roughly by observing the reactions following pinch-
ing, or striking, or mechanical, electrical, or thermal stimuli. Neither
do very young kittens or puppies show any shock when tested by these
methods. But when the turtle, kitten, and puppy are tested with more
delicate methods, using carefully graduated stimuli and avoiding all
sources of error possible, they do show some slight but distinct depres-
sion following the transection of the spinal cord or section of the dorsal
roots. Very probably other species would show similar results.

224 Clyde Brooks. :

From the experiments so far performed the changes in tonus that
accompany the above findings are not clear. At present there is
some evidence that a change in tonus always accompanies the decrease
in irritability of the spinal cord, and that this change is usually a
decrease; but in some instances it is an increase of tonus.

These results tend to strengthen the viewof Von Monakow, Sherring-
ton, Moore and Oertel, Babak, Munk, and others, who hold that
spinal shock is due to the functional separation of the cord from the
higher structures; and to cast doubt upon the theory that spinal shock
is due solely to the anatomical severance of the reflex arcs directly
involved. 7

This influence of the higher parts of the central nervous system and
sensory roots upon the spinal cord at present appears to be most sat-
isfactorily designated as regulatory, that is, a sort of balancing of
excitatory and inhibitory influences. According to this view, in the
intact animal the cord is normally in a state of relative tonic equilib-
rium; when the cord is transected or when a sufficient number of dorsal
roots are cut, this state of relatively stable equilibrium is temporarily
disturbed, resulting in more or less depression of reflex irritability of
the spinal cord, that is, spinal shock.

BIBLIOGRAPHY.

1 BROOKS: Science, 1908, xxviii, p. 541.

* Harvess: Abhandlungen der kéniglich-bayerische Akademie der Wissen-
schaften, Miinchen, Mathematisch-physikalische Classe, 1858, xxxi. BRONDGEEST:
Archiv fiir Anatomie und Physiologie, 1860, p. 703. von Cyon, E.: Berichte
der kéniglich-sichsische Gesellschaft der Wissenschaften, Leipzig, Mathematisch-
physische Classe, 1865, p. 85. Brzorp und Uspensxy: Centralblatt fiir die
medicinischen Wissenschaften, 1867, v, p. 611. GuTTMANN: Centralblatt fiir die
medicinischen Wissenschaften, 1867, v, p. 689. von Cyon, E.: Centralblatt fiir
die medicinischen Wissenschaften, 1867, v, p. 643; Gesammelte physiologische
Arbeiten, 1888, p. 197. Marcaccr: Archivio per le scienze mediche, Torino,
1882, v, p. 283. Marcacct’s paper is correctly reported in an abstract in Horr-
MANN and SCHWALBE’S Jahresberichte, 1883, xi,, p. 36, and is incorrectly re-
ported in ScHArER’s Text-book of physiology, 1900, i, p. 799, and also in
VircHow and Hirscu’s Jahresberichte der gesammte Medicin, 1882, i, p. 218.
Herne, H. E.: Archiv fiir die gesammte Physiologie, 1893, liv, p. 614.

’ BELMONDO and Oppr: “ Interno all’ influenza cette radici spinale posteriori
sull eccitabilite della anteriori,’’ Laboratorio di fisiologia, Florence, 1890. SHER-
RINGTON: SCHAFER’S Text-book of physiology, 1900, p. 799.

Effect of Lesions of the Dorsal Nerve Roots. 225

4 Morr and SHERRINGTON: Proceedings of the Royal Society, London, 1894—
1895, lvii, p. 481.

5 MERZBACHER: Archiv fiir die gesammte iisetlene 192, KCl, Pp. S85.

6 TRENDELENBERG: Archiv fiir Physiologie, 1906, p. 1; [bid., 1906 (supple-
ment), p. 231; Zentralblatt fiir Physiologie, 1906, xix, p. 311.

7 HALt, MARSHALL: Synopsis of the diastaltic nervous system, London, 1850.
EckHArpD: Beitrige zur Anatomie und Physiologie, Giessen, 1881, ix, p. 29.
Wuvtt: The works of Robert Whytt, M.D., Edinburgh, 1768.

8 HALL, MarsHatt: Loc. cit. Wuytt: Loc. cit.

® Gortz: Archiv fiir die gesammte Physiologie, 1873, viii, p. 460; [bid., 1876,
xili, p. 39; Lbid., 1877, xiv, pi 44ne [bid, 1670, 2s, pa 1 Cle.

10 SHERRINGTON: Integrative action of the nervous system, 1906, Pp. 241.
PIKE: This journal, 1909, xxiv, p. 124.

1 Bastian: Medico-chirurgical transactions, London, 1890, Ixxili, p. 151.
RosENTHAL und MENDELSSOHN: Neurologisches Centralblatt, 1897, xvi, p. 978.
PIKE: Loc. cit.

12 Moore and OErret: This journal, 1899-1900, ili, p. 45.

18 Moore and OERTEL: Loc. cit. VON Monaxkow: Ergebnisse der Physiologie,
1902, I Jahrgang, II Abteilung, p. 563. SHERRINGTON: Integrative action of
the nervous system, p. 246. BaxpAx: Zentralblatt fiir Physiologie, 1907, xxi, p. O«
Munk: Sitzungsberichte der kéniglich-preussische Akademie der Wissenschaften,
Berlin, 1909, xliii, xliv, p. 1106.

144 von Monakow: Loc. cit.

15 SHERRINGTON: Integrative action of the nervous system, p. 246.

16 Munk: Loc. cit.

17 BapAxk: Archiv fiir die gesammte Physiologie, 1902, xciii, p. 134; Ibid.,
1905, cix, p. 78; Zentralblatt fiir Physiologie, 1907, xxi, p. 9; Lbid., 1907, xxi,
D.- 523;

18 PIKE: This journal, 1909, xxiv, p. 124.

19 BapAx: Zentralblatt fiir Physiologie, 1909, xxiii, p. 151; Biophysikalisches
Centralblatt, 1909, iv, p. 402.

A QUANTITATIVE STUDY OF FARADIC STIMULATION. —
V. THE INFLUENCE OF TISSUE. RESISTANCE AND
OF KATHODE SURFACE ON STIMULATING EFFEC-
TIVENESS.

By E. G. MARTIN.

[From the Laboratory of Physiology in the Harvard Medical Schooi.]

N the beginning of this investigation! the resistance of the sec-
ondary circuit was set down as one of the variable factors to be
considered in determining the stimulating values of induced currents.
At that time, however, the tentative assumption was made that
break induction shocks are not modified in efficiency by alterations
in secondary resistance. This assumption? was based upon cer-
tain experiments of Helmholtz,’ and has been shown by the work of
Hoorweg* and Gildemeister,® as well as by the experiments reported
in this paper, to be erroneous. For the earlier parts of the investiga-
tion it was immaterial whether or not the secondary resistance affects
the values of stimuli; therefore this phase of the problem was put
over till the other phases had been considered.

Inasmuch as the research of which this paper is a part is primarily
an attempt to increase the practical usefulness of faradic stimuli,
and concerns itself only incidentally with the theoretical considera-
tions which arise in connection with their study, the questions to be
asked with regard to the influence of the secondary resistance are:
(1) In what direction does this influence manifest itself, and is it
measurable? and (2) Is it so extensive that it must always or often
be taken into account? This latter question is one of great practical

1 Martin: This journal, 1908, xxii, p. 72.

2 See Martin: Loc. cit., p. 118.

° HELMHOLTZ: PocGcENDoRF’s Annalen der Physik und Chemie, 1851, Ixxxiii,
P. 536.

* Hoorwec: Zeitschrift fiir Elektrotherapie, 1899, i, p. 10r.

° GILDEMEISTER: Archiv fiir die gesammte Physiologie, 1910, Cxxxi, p. 610. —
226

A Quantitative Study of Faradic Stimulation. 227

bearing, for it must be admitted at once that if the influence of the
secondary resistance has to be taken into account in every quantita-
tive use of faradic stimuli the difficulty of such use becomes very
great, and will often be found prohibitive.

The subject matter of this paper falls naturally into two divisions:
the first dealing with the measurement of the influence of the second-
ary resistance, and the second with the question of how important it
is to take this influence into account, and in what classes of work, if
any, it may be disregarded.

THE MEASUREMENT OF THE INFLUENCE OF SECONDARY
RESISTANCE.

The relation of tissue resistance to secondary resistance as a whole. —
The secondary circuit usually has a comparatively high resistance.
Most inductoria used in physiological laboratories have secondary
coils with resistances mounting into hundreds of ohms, and the re-
sistances of the tissues undergoing stimulation are usually high like-
wise. In numerous determinations of the resistance of stimulated
tissues I have met with only one or two under 1000 ohms and have
found many exceeding 50,000 ohms.

Since it has been shown conclusively that the stimuli imparted by
faradic currents as well as those of galvanic origin arise from the
kathode,® and since the resistance of the physiological kathodes must
be small in comparison with that of the whole mass of tissue traversed
by the current, we are justified in considering tissue resistance as ex-
ternal to the actual seat of stimulation, and need make no distinction
between this and the other resistances that may be included in the
secondary circuit.

The method of experimentation. — In studying the influence of second-
ary resistance experimentally the usual procedure has been to in-
troduce known, non-inductive resistances into the secondary circuit
and to observe the effect of their introduction upon the stimu-
lating value of the shocks sent through the circuit. As a check upon
this method some experiments were performed in which different
amounts of tissue were included between the stimulating electrodes,

§ CHAUVEAU: Journal de la physiologie, 1859, ii, pp. 490, 553. See also
BIEDERMANN: Elektrophysiologie, Jena, 1895, ii, p. 622.

228 E. G. Martin.

and thus the resistance of the tissue itself was varied. This latter
method is of course less certain than the former, since the inclusion
of more or less tissue in the circuit may mean a variation in the number
and irritability of the physiological kathodes involved.

Tissue resistances were determined by means of an ordinary wheat-
stone bridge according to the Kohlrausch method, using an alternat-
ing current to avoid polarization, and a telephone in place of the
galvanometer. The average of three readings was always taken.
This procedure, in the hands of one experienced in its use, gives re-
sults accurate within 4 or 5 per cent, a degree of accuracy sufficient
for the purposes of this inquiry.

The measure of stimulating effectiveness was the same as in the
earlier parts of this research, namely, the stimulus required to pro-
duce a minimal contraction in a frog’s leg muscle, stimulated either
directly or indirectly. Break shocks were used throughout this part
of the work and the expression for the value of the stimulus is Z,
determined from the formula,’

Z= 7 Re GE)

The effect upon the stimulus of varying the secondary resistance. — The
effect upon the value of Z of varying the secondary resistance is shown
in two representative experiments cited in Table I. As appears from
this table, stronger stimuli are required to produce a given physio-
logical effect when the secondary resistance is high than when it is
low. That there is a definite mathematical relationship between the
effectiveness of the stimulus and the secondary resistance is shown
by plotting these values as a curve. Such a curve for the first ex-
periment of Table I is given in Fig. 1. It is virtually a straight line
having the general equation

i A
sen

in which Z is the intensity of the shock required at resistance R to
produce the desired effect, and 8 and A are constants. This formula
has been found to hold in every experiment, numbering more than
fifty, in which it has been applied. The value of the constant @ in

7 Martin: This journal, 19090, xxiv, p. 271.

A Quantitative Study of Faradic Stimulation.

2209

any given experiment can be determined geometrically by produc-
ing the curve to where it cuts the ordinate for zero resistance. Ac-

cording to Fig. 1, the value of 8
for the experiment of Table I
from which that curve is derived
is 3. Since this represents the
value of Z, whose effect at zero
resistance would equal that of the
various other values of Z at their
respective resistancés, it affords a
measure of the irritability of the
physiological kathode where the
stimu us actually arose, assuming
that the resistance of such kath-
ode is negligibly small. We have,
therefore, in @ an expression for
the value of any stimulus as it
affects the seat of actual stim-
ulation, namely, the physiolo-
gical kathode, irrespective of
the resistance of the secondary
circuit.

4000 8000 12000 16000 20000
Ficure 1. — Showing that the curve of in-
creasing stimulus against increasing fresist-
ance is a straight line. Ordinates represent
resistances in ohms; abscisse represent

values of Z.

TABLE I.

THE INFLUENCE OF SECONDARY RESISTANCE UPON THE STIMULATING
VALUES OF INDUCED CURRENTS.

Experiment of Dec. 15, 1909. ReEsISTANCE OF SECONDARY Cort = 1400 oHMs;

oF TissuE = 1700 onms.

TIssuUE = Froc’s GASTROCNEMIUS, UNCURARIZED.

Resistance in secondary circuit 3100

Value of Z 4.96

6100 10100 15100 18100

6.81 9.45 12.45 14.1

Experiment of March 1,1910. RerstsTANCE OF SECONDARY Cort = 1400-0HMs;

oF TISSUE =

16600. Tissu—E = Froc’s SARTORIUS, UNCURARIZED.

Resistance in secondary circuit

Value of Z

18000

28000 48000 68000

5.24 6.8 9

230 E. G. Martin.

By a slight transposition of equation (2) the equation for 8 be-

comes:
ZA

Pole ee (3)

and it is clear that if the value of Z for any secondary resistance is
known the actual or “specific” stimulus can be calculated from
equation (3), provided only the value of the other constant, A, is
known. For measuring stimuli with reference to the resistance
through which they are applied, therefore, there must be added to
the determinations previously required not only the secondary re-
sistance, but a constant A.

Kathode surface an important factor.— That the value of A depends
upon the surface offered by the physiological kathode or kathodes is
strongly indicated by the results of a series of thirteen experiments
upon frog’s leg muscles, gastrocnemius, triceps femoris and sartorius,
in which the surface of the physical kathode was accurately deter-
mined. The method was as follows: The kathode was a cylindrical
platinum wire of known diameter. This was thrust a measured dis-
tance into the muscle tissue. Thus the kathode surface could be
calculated. In the thirteen experiments under discussion there was
a direct proportionality between the kathode surfaces and the values
of the constant, A, as determined from the plotted curves of the
experiments. This proportionality is brought out in Table II. I do
not wish to urge that these experiments show anything more than a
dependence of the constant upon the kathode surface. They cannot
be looked upon as establishing a simple method of determining its
value. In fact in seven other similar experiments there was no marked
proportionality between kathode surface and the value of A. Nor
was I able in numerous experiments in which nerves were stimulated
instead of muscles to get evidence of any simple relation between
kathode surface and the value of A. Failure to find such propor-
tionality regularly does not, however, invalidate the idea that there
is a definite relationship of some sort, when we consider the many
factors which go to determine the actual physiological kathode, of
which the surface of the physical kathode is but one; and when we
recall that this idea is merely the application in special form of a fact

A Quantitative Study of Faradic Stimulation. 231

long recognized, namely, the influence of current density on stimula-

tion value.
TABLE II.

INDICATING A DEPENDENCE OF THE CONSTANT, A, UPON KATHODE SURFACE.

A Muscle

Kathode Sa
Value of A. Kath. surf. stimulated.

surface.

4500 800 gastrocnemius
9000 800

13500 800

10500 800
6000 800

4800 800
4800 300
5700 790
3700 800

6900 800 gastrocnemius
4200 810 #8

1000 830 sartorius
4800 800 gastrocnemius

I know of no reliable method of determining the value of the con-
stant, A, other than that used in this work, namely, to establish ex-
perimentally at least two values of Z for different secondary resist-
ances, and from these values compute the value of A. This can be
done by means of the equation

BEE eee
Ze — Zp

= 4)

in which Zz and Zz, are the stimuli required with resistances R and
!/ . =e e

R respectively to produce the minimal contractions used as the

index.

8 BIEDERMANN: Loc. cit., i, p. 185.

232 | E. G. Martin.

The factor of secondary resistance is thus, as we see, indissolubly
connected with another, depending on the kathode surface and re-
quiring the determination of a constant more difficult to obtain than
is the secondary resistance itself. |

Before imposing this additional burden upon ohysiolene ex-
perimentation we may well inquire how great errors are likely to
arise in comparing faradic stimuli if these two factors are completely
disregarded.

How EXTENSIVE IS THE INFLUENCE OF SECONDARY RESISTANCE
AND KATHODE SURFACE?

We must recognize at the outset of this part of our inquiry that if
comparisons are attempted between stimuli used under conditions of
widely varying secondary resistance and divergent kathode surface,
disregard of these two factors is sure to lead to erroneous conclusions;
but in a majority, probably, of physiological experiments the stimuli
to be compared are produced under conditions which tend to be closely
similar. It is with regard to such cases as these that we may properly
inquire whether the factors under consideration need be taken into
account.

Successive stimulation of the same tissue. — Probably the experiments
in which accurate comparisons of stimuli are most needed are those
in which a given tissue is to be stimulated successively. But in ex-
periments of this class neither the tissue resistance nor the electrode
surfaces undergo noteworthy variation during the course of the ex-
periment and so do not enter as modifying factors.

Stimulation of corresponding tissues in different animals. — Next in im-
portance are cases in which it is desired to impart comparable stimuli
to corresponding tissues through a series of experiments. Cases of
this sort arise very frequently in the course of physiological research,
and I have therefore given them special consideration.

Mr. E. L. Porter has been carrying on in this laboratory an in-
vestigation which involves, among other things, determining in a
series of cats the threshold stimulus for producing extension of the
wrist, the stimulus being applied to the deep branch of the radial
nerve below the elbow; and reflex flexion of the hind leg through
stimulation of the musculo-cutaneous branch of the peroneus. Here

A Quantitative Study of Faradic Stimulation. 233

was presented a typical example of the class of experiments described
in the paragraph heading, and I therefore secured Mr. Porter’s co-
operation in utilizing it in the study of my problem. At my request
he determined in several cases the threshold stimulus when the tissue
only was in the secondary circuit, and immediately afterward, the
threshold when an additional re-

sistance of 10,000 ohms had been ,.,,
introduced. I was thus able in ,.
these cases to compute the value of

ciecametant. A, and) from it, €0- EP *. 7:3 54 5 |6 “7B 9 34
obtain the solution of the equa- FicurE 2.— Illustrating the relatively

. ; hae cane slight departures of individual ratios
tion for “specific” irritability, of 6 to Z from the average. Ordinates
mee:

represent successive experiments; ab-
scisse represent ratios of 8 to Z. The
of this series,ten in all, the second- horizontal line is drawn at the level of
: the average ratio.
ary resistances ranged from 2800
ohms to 6000 ohms, averaging 3900 ohms. The values of A ranged
from 4300 to 14,000, averaging 7800. ‘The statistics for this series
are given in Table III.

Inspection of the table reveals a definite tendency of 8 to vary as
does Z. The closeness of this tendency is brought out more strik-
ingly, however, in Fig. 2, where the ratios of B to Z in successive ex-
periments are plotted. The horizontal line represents the average
ratio of 8 to Z as determined in these experiments; the variations
from this line of the different actual ratios are, as is seen, relatively
inconsiderable, the greatest being 18.5 per cent, the average of all
slightly under 11 per cent.

Assuming the data cited in Table III to be fairly representative
of the relations between 8 and Z that are likely to occur in experi-
ments of the sort under consideration, to what extent are we justified
in such experiments in making use of the values of Z for expressing
quantitative relationships?

The figures show clearly, I think, that all except the finest rela-
tionships are revealed with sufficient exactness by the values of Z.
While it cannot always be known certainly, in cases in which several
nearly equal values of Z are under comparison, which will give smaller
‘and which larger values of §; yet, if the experiments are carefully
performed, one can be practically certain whenever the values of Z

In the experiments

234 E. G. Martin.

differ by more than 15 or 18 per cent that the larger Z means also a
larger 8. In other words, the values of Z are accurate within about
15 to 18 per cent. With this degree of accuracy assured, probably
the demands of most researches of this class are fully met, and all

TABLE III.

ILLUSTRATING THE TENDENCY OF 8 AND Z TO VARY SIMILARLY IN DIRECTION AND EX-
TENT, Z REPRESENTS THE STIMULUS PRODUCING JUST PERCEPTIBLE WRIST EXTENSION
In Cat. STIMULUS APPLIED TO DEEP BRANCH OF RADIAL NERVE.

Secondary
SRSA Value of A. | Value of Z. | Value of f.

6000 7600 1.16 0.65
4400 5000 1.33 0.71
4800 8000 1.60 1.0
3400 7800 1.80 17S
3000 9800 2.3 1.76
4600 9600 2592 1.7
2800 4600 10.6 6.6

AS ABOVE EXCEPT THAT STIMULUS WAS APPLIED TO MuscuLo-CuTANEOUS BRANCH
OF PERONEUS, AND REFLEX FLEXION oF Hinp LEG wAs MovEMENT EVOKED.

july 4s le 3000 4300 1.70 59
Tuly 225.02" 4000 14000 2.75 76
July 20... 5.) 3000 7000 10.3 , 70

Average 65

such may safely disregard both the secondary resistance and the
kathode surfaces.

Differing from the series of experiments quoted above in that they
offer wider variations in both secondary resistance and electrode sur-
face, and therefore greater likelihood of error if these factors be dis-
regarded, is a series of observations on frog’s gastrocnemius muscle,
carried out by myself. The method of stimulation was that described

A Quantitative Study of Faradic Stimulation. 235

in an earlier paper of the series,° the electrodes being platinum needles
thrust directly into the muscle substance.

TABLE IV.

ILLUSTRATING TENDENCY OF Z AND 8 TO VARY TOGETHER. FRoG’s
GASTROCNEMIUS STIMULATED DIRECTLY.

Date. Secondary Value of A.| Value of Z. |} Value of 8. Ratio Bs
resistance. | HA

Feb. 24,1910 ... 5400 : | 59
Wer 28,1909 ... . 6500 : 46
Wee ts, 1909 3100 : .60
rr E910. se 8400 36
Hep, 1910. . 6800 ; 52
BED 24, 1910 ne .. «. ; A2
Hep18, 1910... : : 47
Nanssie1910; 2. . : 3 56
Mary, 1910 ... | : : A8
Benjte, 1910 . . . ; ; 52

Nov. 18,1909 ... E 34
Hea Agei910) 2.
Jan. 31, 1910

Die ie a! a!
Jan. 31, 1910

Bebe ts,1910 ....°.
webs if, 1910 . . .
Moy. 41909 -. .

Average

In the series of eighteen experiments cited in Table IV the sec-
ondary resistances ranged from 3100 to 13,000, and the values of A
‘from 2600 to 13,500. Yet, in spite of these wide ranges in the values

® Martin: This journal, 1908, xxii, p. 117.

236 E. G. Martin.

of the factors determining the relation of Z to 8, this latter relation
varies to a surprisingly moderate degree. The average ratio of @ to
Z is .49. The widest departures from this are ratios of .33 and .65,
amounting to 33 per cent in each case, while the average variation is
only 15 per cent. If the experiments of Table IV represent fairly the
variations in secondary resistance and kathode surface likely to be
met with in experiments on frog’s gastrocnemii, we can safely con-
clude that the values of Z, obtained in any such experiment, represent
the true relative values of the stimuli used within less than one third.
In a series of ten experiments on frog’s gastrocnemii stimulated
through the sciatic, with resistancés ranging from 6300 to 38,000 ohms,
and values of A from 6000 to 23,000, the ratio of 8 to Z averaged .48,
and the widest variation was a ratio of .28, amounting to 42 per cent,
the average variation being 20 per cent. In the experiments cited in
the two series above no attempt was made to keep conditions of
tissue: resistance and kathode surface approximately uniform. On
the contrary, these conditions were purposely made to vary widely
from one experiment to another. I feel, therefore, that they cover
the range of variation likely to occur in ordinary experimentation.

DISCUSSION.

The data thus far cited seem to me to show that in the hands of
a careful experimenter, who will take pains to keep his conditions of
stimulation as uniform as possible, quantitative results of great
value may be obtained without the labor involved in taking account
of secondary resistance and kathode surface. By the use of the
method outlined in previous papers of this series the strengths of
stimuli employed in any given case may be expressed in terms of
stimulation units, and if the conditions of experimentation, such as
nature of electrodes used, distance between them, and method of ap-
plying them, are carefully described, other experimenters can dupli-
cate the stimuli very closely. Certainly this method allows com-
parisons of much greater accuracy than can be made by the methods
of describing stimuli in vogue at the present time.

While it is true that in many investigations, in which accurate
knowledge of the strengths of stimuli used is of paramount impor-
tance, careful account will have to be taken of the two factors, sec-

A Quantitative Study of Faradic Stimulation. 237

ondary resistance and kathode surface, yet in a far greater number of
researches in which approximate knowledge of the stimuli used is all
that is desired, these factors may be wholly disregarded.

Use of high additional secondary resistance. — In connection with the
discussion of the influence of secondary resistance and kathode sur-
face it is important to emphasize the fact that kathode surface is
fully as influential in modifying the strength of stimulus as is sec-
ondary resistance. That this fact has not been appreciated hitherto
is shown by the frequent use of a device supposed to overcome any
inequality in stimulation strength due to differences in secondary re-
sistance, namely, the introduction of a very high additional resistance
into the secondary circuit, thereby making fluctuations in tissue re-
sistance relatively negligible. That this device is perfectly adequate
in experiments in which a single tissue of varying resistance is under
examination is of course obvious; there being under such circum-
stances no variations in kathode surface. But in experiments in which
different tissues are being compared the introduction of high addi-
tional resistance into the secondary circuit is more apt to be misleading
than otherwise because of the cumulative effect of variations in
kathode surface. The point can best be illustrated by a concrete
example:

Experiment of March 7, 1910. — Frog’s gastrocnemius muscle stimulated
directly. In the first test the kathode was in contact with the surface
of the muscle, but did not penetrate it. When the tissue only was in
the secondary circuit, the total secondary resistance was 17,000 ohms.
A minimal contraction was secured with a value of Z equal to 6.6.
When 70,000 ohms’ additional secondary resistance was introduced,
the value of Z was 16.8. By calculation the value of A was found to
be 24,000 and of § to be 4. In the second test the kathode was thrust
directly through the muscle tissue; the secondary resistance was 5400
ohms; the value of Z was 6.1. When 70,000 ohms’ additional resistance
was introduced, the value of Z was 40.5. The calculated value of A was
4800, and of 8 2.9. In this case the values of Z as determined with
the tissue only in the secondary circuit represent much more nearly
the true relationships between the stimuli than do the values as de-
termined with a large additional resistance in the circuit. In reality
the stimulus applied in the first test was stronger than in the second,
whereas, if reliance were placed upon the results given when the high
additional resistance was in circuit, it would appear that the second

238 E. G. Martin.

stimulus was more than twice as strong as the first. The subjoined
tabulation will serve to emphasize the error:

Z (tissue only). Z (70,000 ohms added). B.
1h 0 ee eet 6.6 16.8 4
OCOUP ia it, fey woh ve 6.1 40.5 - 2.9
Ratio of rstto 2d 1.08 0.41 1.38
SUMMARY.

1. This paper deals with two factors upon which the effectiveness
of any faradic stimulus depends: they are the resistance of the sec-
ondary circuit and the surface presented by the physiological kathode.
These are so closely interconnected that account cannot be taken of
either independently of the other.

2. If it be assumed that the resistance of the physiological kathode,
which is the point actually stimulated, is negligibly small, the value
of the stimulus applied to this point is expressed by the equation
B= eae ve in which f is the actual or ‘‘specific” stimulus, Z the
stimulus as determined regardless of the resistance of the secondary cir-
cuit, R this resistance, and A a constant depending upon the kathode
surface.

3. To determine the value of A, corresponding values of Z at
different secondary resistances must be found and A computed from
the formula
SO RUS ee
yee

in which Zp and Zz are the values of Z for the resistances R and R’
respectively.

4. By a study of series of experiments it is shown that the in-
fluence of the two factors under discussion upon stimulating effective-
ness is in many cases too slight to require attention. This is true:
(a) In experiments in which neither secondary resistance nor kathode
surface varies during the course of the experiment. (b) In those in
which certain nerves or muscles are to be stimulated in essentially
similar fashion in a succession of animals; in which case it is possible,
by care in maintaining uniform conditions, to reduce to a minimum

A Quantitative Study of Faradic Stimulation. 239

the variations in secondary resistance and in kathode surface. Ex-
periment shows that with due care the probable error need not exceed
18 per cent in these cases. (c) In experiments in which it is not so
important to know the actual value of the stimulus used as to describe
it so that other investigators can duplicate it. If the conditions of
stimulation are carefully described and the value of Z given, it is
possible in practically every such case to duplicate the stimulus within
one third or less by following closely the conditions given and employ-
ing the same value of Z.

5. It is necessary to take into account the factors of secondary re-
sistance and kathode surface in experiments whose prime object is
the comparison of stimuli applied to different tissues; and in all ex-
periments requiring a higher degree of accuracy than is indicated in
the paragraph above, in which either factor changes during the course
of the experiment.

6. On account of the close interdependence between secondary
resistance and kathode surface the introduction of a high additional
resistance into the secondary circuit for the purpose of rendering
negligible fluctuations in tissue resistance is permissible only in cases
where the kathode remains unaltered throughout; the method is
likely to give wholly misleading results if used in experiments which
involve changes in the position of the kathode, or shifting of the
stimulus from one tissue to another.

Nore. After this paper had been sent to press I discovered an error of cal-
culation in the determination of the mutual induction of the standard coil used
in calibrating my inductoria (see this journal, 1908, xxii, p. 121), which has led
me to use throughout the work a scale smaller than the true one. This error in
no wise affects the validity of the quantitative method developed in this series
of papers. On account of it, however, all figures for strength of stimulus reported
by me are too small. The correct values can be obtained in every case by
multiplying my figures by 2.4.

ON THE DYNAMICS OF CELL DIVISION. — II. CHANGES
IN PERMEABILITY OF DEVELOPING EGGS TO ELEC-
TROLYTES. )

By J. F. McCLENDON.

[From the Histological Laboratory of Cornell University Medical College, New York City,
and the Laboratories of the Carnegie Institution at Tortugas, Fla., and the U. S. Bureau
of Fisheries at Woods Hole, Mass.]

HE process of cell division may be divided into two distinct phe-
nomena, the division of the nucleus and of the cytoplasm. Al-
though these processes are closely interrelated, they can occur sepa-
rately. Karyokineses may occur without subsequent cytoplasmic
division, and cytoplasmic division may occur without the presence of
a nucleus or chromatin in any form.’ The division of the cytoplasm
in plant cells is accomplished by the formation of a division wall, but
in most animal cells by simple constriction.

The constriction of the cell seems to be a special case of these pro-
toplasmic movements that were shown by Quincke to resemble move-
ments accompanying surface tension changes.” If the constriction of
the cytoplasm were due to surface tension changes, we should expect —
a band of greater surface tension to include the cleavage furrow. In ~
this case there would be a flowing of the superficial cytoplasm from
the poles of the cell toward the cleavage furrow, and of the deeper
protoplasm in the opposite direction.

By a study of fixed material Nussbaum * showed movements of the
pigment granules in cells of frog’s embryos from the interior to the sur-
face and along the surface to the position of the future cleavage furrow.
On constriction of the cell the granules were massed in the form of a
plate in the cleavage plane. These movements indicate the surface
tension changes described above.

ieee, a eee

* McCLenpon: Archiv fiir Entwicklungsmechanik, 1908, xxvi, p. 662.
* Buetscuii: Archiv fiir Entwicklungsmechanik, 1900, x, p. 52.
® NusspAuM: Anatomische Anzeiger, 1893, viii, p. 666.

240

On the Dynamics of Cell Division. 241

Conklin ‘ found evidence for such movements in the changes in posi-
tion of certain structures in Crepidula eggs.

The first observation of this process in living cells was made by
Erlanger,” who saw movements of superficial granules toward the
cleavage furrow, and of internal granules toward the poles, of Nema-
tode eggs. |

Gardiner ° observed, in living eggs of Polychoerus caudatus, colored
granules move to the surface and then along it to the position in which
the cleavage furrow appeared immediately afterward. Fischel’s ob-
servations are considered below.

The fact that cells usually round up before cleavage, if not previ-
ously spherical, indicates a general increase in surface tension, and it
is only necessary to assume a greater increase to be localized along the ~
cleavage furrow to account for the constriction.

Robertson ” floated an olive oil drop on water and laid across it a
thread moistened with soap (or soap-forming) solution. After the
thread reached the edges of the drop the latter was torn in two. Since
soap decreases the surface tension between oil and water, he concluced
that cytoplasmic division is due to a decrease in surface tension along
the cleavage furrow. As this view has been accepted by Lillie ® and
Loeb,’ it seems worth whlie to point out Robertson’s error.!? In Rob-
ertson’s experiment three different surface tension films occur, between
air (A) and water (W), air and oil (O), and water and oil (Fig. 1), and
an equilibrium is established when the water-air surface tension equals
the horizontal components of the air-oil plus the oil-water surface
tensions. When the moistened thread is laid across the oil drop, two
more films are added, 7. ¢., air-soap solution (S) and oil-soap solution
(Fig. 2). At opposite edges of the drop where the thread touches the
water, the soap would decrease the water-air surface tension, and the
undiminished pull on the remainder of the edge of the drop would pull
it in two (Fig. 3).

4 CONKLIN: Biological lectures at Woods Hole, 1908, p. 60.

5 ERLANGER: Biologische Centralblatt, 1897, xvii, p. 152.

6 GARDINER: Journal of morphology, 1897, xi, p. 55.

7 Ropertson: Archiv fiir Entwicklungsmechanik, 1909, xxvii, p. 20.

8 LittiE: Biological bulletin, 1909, xvii, p. 203, footnote.

® Logs: Chemische Entwicklungserregung des tierischen Eies, p. 5.

” This was first pointed out by me before the American Society of Zodlogis's,

see Science, xxxi, p. 467. It was discussed by A. B. MACALLuM, Science, 1910,
XXxil, pp. 498-500.

242 J. F. McClendon.

I have repeated Robertson’s experiment and also modified it by
entirely submerging the oil drop. Enough alcohol was added to the
water to make the oil sink below the surface, and the soap solution
introduced through a capillary pipette, or a piece of solid soap held
near the oil drop, or a thread covered with solid soap was wrapped
around the oil drop. Very little movement of the oil occurred, but

Ficures 1 to 3.— A, air; O, oil; W, water; S, soap solution. The arrows show the
direction of the pull of surface tension. The dotted line in Fig. 3 bounds the soap
solution. Further explanation in text.

that which did occur was always a bulging toward the soap, and never
a constriction or receding from the soap. Similar experiments were
also tried on the under side of oil drops floating on water, and unless the
soap reached the water-air film, the oil advanced toward the soap and
no constriction occurred. We may conclude, then, that the cleavage
furrow is a region of increased surface tension, as shown by Biitschli
and others, and not of decreased surface tension, as Robertson, Lillie,
and Loeb maintain.

One may ask why such movements as seen by Eee Gardiner,
and others have not been observed in all dividing cells that have been
studied alive. The answer to this may be sought in the structure or
consistence of protoplasm. If the cytoplasm present an alveolar
structure, the spreading of the surface in regions of reduced surface
tension would be almost entirely confined to the individual alveoles,
and the general effect would be a slow stretching of the surface in

re rs.

eect im it ee Raa

On the Dynamics of Cell Division. 243

areas of less surface tension and a slow contraction of the surface in
areas of greater surface tension, as shown by Biitschli’s microscopic
oil foams. The constriction of the cleavage furrow would then take
place as though a rubber band around the cell contracted.

In the constriction of the egg of the sea urchin, Arbacia punctulata,
usually just such a number of chromatophores are carried into the
cleavage furrow that when the two daughter cells are formed the pig-
ment is evenly distributed over all parts of their surfaces. Under cer-
tain abnormal conditions in which the cleavage is more violent the
pigment is massed in the furrow. Fischel observed a massing of pig-
ment along the lines of the future cleavage furrows in the eggs of Ar-
bacia pustulosa. The reasons we get so little movement of pigment in
the egg of Arbacia punctulata probably are the alveolar structure and
the presence during cleavage of the “‘hyaline plasma layer,” or “Ver-
bindungsschicht,”’ to which the surface movements are chiefly confined.
The hyaline plasma layer is said to be formed by a recession of granules
toward the interior of the egg, leaving the superficial layer free from
granules and almost invisible. It is formed before the first cleavage
and becomes heaped up in the cleavage furrow." Whenever such an
outer layer occurs, as in eggs of sea urchins and ctenophores, it be-
comes heaped up in the cleavage furrow; this indicates increased sur-
face tension in this region (or decreased surface tension at the poles).

In the cutting off of the micromeres in the Arbacia egg the pigment
entirely disappears from the micromere pole, indicating spreading
movements due to the surface tension being less here than in the re-
gion of the future cleavage furrow. Similar movements of granules
have been observed in the cutting off of polar bodies in various eggs,
and it may be concluded that for the separation of a very small cell
from a large mass of protoplasm a very great difference in surface ten-
sion between the pole of the small cell and the cleavage furrow is
required.

Changes in surface tension may be the result of the presence of cer-
tain substances in one of the two fluids in contact, changes in tempera-
ture, or a difference in electric potential across the boundary. In
numerous instances electric changes have been found to accompany
vital movements. Hyde” detected electric changes accompanying

1 Goipscumipt and Poporr: Biologische Centralblatt, 1908, xxviii, p. 210.
® Hype: This journal, 1904, xii, p. 241.

244 J. F. McClendon.

cleavage, and the question whether the constriction of the cytoplasm
is due to electric changes seems worth investigation.

If an electric current were passed through a solution containing a
living cell, and if the cell surface offered more resistance to the passage
of ions than either the medium or the cell interior, a difference of po-
tential would be produced between the inner and outer sides of the
cell surface, and would be proportional to the angle that the surface
made with the current lines cutting it, 7. e., it would be greatest at the
point where the surface was at right angles to the current lines and
equal to zero at the point where the surface was parallel to the current
lines. The surface tension would be reduced at the poles (the points
nearest the electrodes), and the equator would lie in a region of rela-
tively greater surface tension. This would result in the protrusion of
the polar regions and constriction of the equator, thus producing the
form change of the first stage of cleavage.

When a current of a certain density was passed through an unfer-
tilized Arbacia egg, the surface nearest the anode showed spreading
movements and bulged out. We might conclude from this that this
egg was less permeable to anions than to cations. The confined anions
caused a difference of potential between the two sides of the surface
nearest the anode, thus decreasing the surface tension, and spreading
of the surface and bulging of the egg followed. At the surface nearest
the cathode, the anions that could not enter could pass around the
egg, and therefore the difference of potential was not so great as at the
opposite pole. If the egg were as poorly permeable to cations, we
should expect a reduction of surface tension at the pole nearest the
cathode.

It seemed to me that an analysis of artificial parthenogenesis might
throw light on the question of cell division, and in the summer of 1909
I began an attempt in this direction.

ARTIFICIAL PARTHENOGENESIS.

It is well known that the eggs of different individuals of the same
species vary in response to stimuli. Investigators usually suppose that
they have normal material when a large per cent of the eggs develop

in a control to which sperm is added. In order to test this method of

RE se 5 lpm —

SETS NS ai aap

On the Dynamics of Cell Division. 245

controlling experiments and be sure no unknown factor vitiated the
results, I made a series of experiments by fertilizing eggs of the same
mother with sperm of different fathers, and eggs of different mothers
with sperm of the same father, with the following results, giving the
percentage of eggs developing:

Mothers.

Fathers.

It may be seen by inspecting the table that development depends more
on the eggs than on the sperm, so the practice of keeping a fertilized
control seems to be a good one. Fertilized and unfertilized controls
were kept to all of my experiments, and the experiment thrown out if
fertilization did not occur.

The question which concerns us first is, in what ways have cells
been caused to divide, and maturation (in most cases), as well as seg-
mentation, is cell division. We may summarize the methods used as
follows:

Hypotonic solutions (distilled water, Schiicking).

Nearly isotonic solutions made by adding to sea water or to distilled
water the following substances:

Acids (Delage, Fischer, Herbst, Lefevre, Loeb, Lyon, Neilson,
Schiicking, Tennent).

Alkalis (Delage, Loeb, Schiicking).

Neutral salts (Delage, Lillie).

Hypertonic solutions:

Acids (Delage, Loeb).

Alkalis (Delage, Loeb).

Neutral salts (Bataillon, Bullot, Delage, Fischer, Hunter, Kosta-
necke, Loeb, Lyon, Mead, Scott, Treadwell, Wilson).

246 J. F. McClendon.

Non-electrolytes (Delage, Loeb).

Mechanical shock (Delage, Fischer, Mathews, Scott).

Thermal changes (Bataillon, Delage, Greeley, Lillie, Loeb, Schiicking).

Electric changes (Delage, Schiicking).

KCN or lack of oxygen (Loeb, Lyon, Mathews).

Fat solvents (Loeb, Mathews).

Alkaloids and glucosides (Hertwig, Loeb, Schiicking, Wassilieff).

Blood sera (Bataillon, Loeb).

Soap and bile salts were found effective by Loeb when followed by
hypertonic solutions.

As all of these agents were not tried and found effective on eggs of
the same species, their differences might be thought to correspond to
differences in the eggs, and therefore I thought it worth while to try a
large number of them on the egg of Arbacia punctulata at Woods
Hole, Mass. Segmentation was produced by the following methods:
the time indicated is the optimum duration found after a series of ex-
periments, the tables being omitted:

Hypotonic solutions (70 c.c. sea water + 30 c.c. tap water or dis-
tilled water, one to one and a half hours).

Nearly isotonic solutions:

Acids (so c.c. sea water and 3 c.c. 1/10 normal acetic, fifteen to sixty
seconds. Or sea water charged with CO, in a “‘sparklet fountain,”
five to ten minutes).

Alkalis (1.2 c.c. 1/10 normal NH,OH or NaOH + 50 c.c. sea
water, twenty to sixty minutes).

Salts (5/8 normal NaCl, one-half to two hours, this is very slightly
hypertonic).

Hypertonic solutions (t00 c.c. sea water + 15 c.c. 2% normal NaCl,
one hour. Or sea water boiled down to .76 of its volume, one hour.)

The eggs were also made to segment by placing them in sea water
brought from Boca Grande Key, twelve miles west of Key West,
Florida, in steamed out, glass-stoppered and paraffine-sealed bottles.
The eggs were allowed to lie twenty minutes in Woods Hole sea water,
then placed for two and a half hours or four hours in Boca Grande sea
water and returned to Woods Hole sea water, or allowed to remain in-
definitely in Boca Grande sea water. At the end of nine hours seg-
mentation had occurred in all three lots, about 10 per cent were seg-
mented in that left four hours in Boca Grande sea water.

On the Dynamics of Cell Division. 247

Woods Hole sea water has a A = 1.818, specific gravity = 1.024
' (Garrey). Boca Grande sea water has a A = 2.05, specific gravity =
1.0248. The alkalescence of the Boca Grande sea water is greater
than that of Woods Hole. The hypertonicity and greater alkalinity
may have both aided in producing the segmentation. This experi-
ment producing such poor results (10 per cent), is given so much
space merely because the effects were produced by natural sea water.
However, this does not prove that Arbacia grown at Boca Grande
would be naturally parthenogenetic.

Mechanical shock (shaking in vial, by hand, five minutes. Or pour-
ing from one dish to another every ten minutes for three hours. One
effect of agitation is the removal of the jelly-like covering from the
eggs, after which, perhaps, the mere contact of the egg with another
surface will start development. Mathews supposes the effect of shak-
ing is rupture of the nuclear wall, at least in the starfish egg).

Thermal changes (keeping at 32° C., four minutes; keeping at 1° C.,
one to eleven hours; or at 10° C., one to twenty hours. By the end of
twenty hours some were already segmented).

Electric changes (several entire ovaries were placed in longitudinal
series in a glass tube, and an alternating current from a small induc-
tion coil was passed through it for two hours. To avoid error from
polarization at the electrodes, only the eggs from the central ovary
were observed).

KCN or lack of oxygen (a stream of hydrogen eight to twenty-two
hours, with or without previous boiling of the sea water. Or 1/500 nor-
mal KCN, seventeen to thirty-two hours. Or 1/1000 normal KCN
thirty-two hours).

Fat solvents (14 saturated solution of ether in sea water, ten
minutes).

Certain combinations Such as carbonic or fatty acid followed by
hypertonic sea water, or tannic acid + an excess of NH,OH,* or tannic

13 In making his ““ammonium tannate” solution, DELAGE considered tannin a
hexivalent acid, but I can find no confirmation of this view in chemical handbooks.
On adding this solution to sea water, a slight precipitate, probably calcium or
magnesium tannate, forms, and the fact that DELAGE obtained as good results
_ by adding the “ammonium tannate”’ to a sugar solution does not prove that the
precipitation of some salt in the sea water solution was not in this case a factor in
the production of parthenogenesis.

248 | J. F. McClendon.

or acetic acid followed by NH.OH or NaOH, seemed to produce better
results than the single treatments.

The following results were obtained on other species:

First, at Woods Hole, eggs of Cumingea and Mytilus were caused
to maturate by treatment with hyperalkaline sea water.

Second, at Tortugas, Fla., where I found the A = 2.03, specific
gravity = 1.0246, and alkalescence greater than at Woods Hole.

A few of the immature eggs of Ophiocoma rizii maturated when
left twelve hours in 50 c.c. sea water + one drop of dilute ammonia,
whereas none maturated in the control.

Eggs of Toxopneustes (Lytechinus) variegatus and Tripneustes
(Hipponée) esculentus were made to segment by placing a test
tube of sea water containing them for one minute in water at
38°-44° C., then pouring the eggs into a dish of sea water at the
normal temperature.

After treatment with sea water carbonated in a “sparklet syphon,”’
followed by hypertonic sea water, development went farther in both
species than after any of the single treatments tried. By increasing
the duration of the treatment the rate of development was increased
(approached or equalled that of fertilized eggs), but the percentage of
resulting larve decreased, indicating injury to the eggs.

The optimum for Toxopneustes was: carbonated sea water one
and one-half to five minutes, followed by roo c.c. sea water + 16 c.c.
2% normal NaCl, thirty to forty-five minutes. If eggs had remained
a long time in sea water before the beginning of the experiment, a
shorter stay in carbonated sea water was required than if they had
been just taken from the ovaries. In this connection it is to be re-
marked that carbonated sea water hastens the solution of the jelly-
like coverings, which takes place more slowly in natural sea water.

The optimum for Tripneustes was carbonated sea water ten minutes,
followed by hypertonic sea water one hour. If we look for something in
common in all of these methods of artificial parthenogenesis, we meet
with many difficulties. I concluded that all of the methods of artificial
parthenogenesis could not directly initiate any one single chemical
reaction in the egg, but must have their first common effect in some
physical or physico-chemical change.

The osmotic methods have this in common, that in all there is an
increase in osmotic pressure. If the eggs are ‘placed in sufficiently

On the Dynamics of Cell Division. 249

concentrated sea water or other hypertonic solution, some may seg-
ment while remaining in the hypertonic solution, but if placed in dis-
tilled water (Schiicking) or diluted sea water (McClendon) they
segment only after removal to natural sea water, which means an
increase in osmotic pressure of the medium. I do not, however, con-
clude from this that in the latter instance it is the return to sea water
rather than the sojourn in the hypertonic solution that starts the de-
velopment of the egg. _

Traube ™ showed that “fertilization membrane-forming” substances
are effective in greater dilution the more they lower the surface tension
of water. If the egg surface contain lipoids, such substances will be
adsorbed or absorbed by the lipoids in the ratio that they lower the
surface tension of water. But in what way can an absorption or ad-
sorption by the lipoids of the cell, of a host of different substances,
cause the development of the egg, and how can we explain those
methods in which no lipoid soluble substance is used?

Loeb had shown the similarity between methods of artificial mem-
brane formation and hemolysis, and many eggs segment after artificial
membrane formation. But in my opinion Loeb has not given a satis-
factory explanation of the mechanism of hemolysis. It seems to me
that the ‘““membrane theory” of hemolysis, so admirably presented
by Stewart,” is a satisfactory explanation.

Lillie * advanced the view that the essential element in artificial
parthenogenesis is the increase in permeability of the plasma mem-
brane to COs, allowing the chief end product of oxidation to escape
and the rate of oxidation to increase, the more rapid oxidation causing
development. But Lillie has never published any determinations of
changes in permeability of the egg to anything except pigment, and
there is no certain proof that the escape of pigment is due to increased
permeability of the plasma membrane, as the pigment must first be
liberated from the chromatophores, in which it is held physically or
chemically.

I know of no method of determining changes in permeability of the
egg to CO:, but have thought of five methods for detecting changes in

14 TRAUBE: Biochemische Zeitschrift, 1909, xvi, p. 182.

19 STEWART: Journal of pharmacology and experimental therapeutics, 1909,
i, p. 40.

#6 LILLIE: Biological bulletin, 1909, xvii, p. 188.

250 J. F. McClendon.

the permeability of the egg to electrolytes in general: 1. Electric con-
ductivity of masses of eggs; 2. Electric conductivity of individual
eggs as determined by destructive effects of the electric current on the
cell; 3. Plasmolysis; 4. Chemical analysis of masses of eggs; 5. Mi-
crochemical analysis of single eggs. These will be considered in the
order given.

Brown!’ concluded that the membrane of the Fundulus egg is prac-
tically impermeable to salts and water during the first eight hours,
and becomes most permeable after eighteen to twenty hours. Ap-
parently he refers to the thick membrane which is pushed out after
oviposition, but Sollmann '§ observed that the ‘‘yolk” swells in dis-
tilled water or one-fourth molecular cane sugar, obliterating the peri-
vitelline space, and thus indicating that the plasma membrane is less
permeable to salts than is the thick egg membrane or chorion (vitel-
line membrane of Sollmann).

Biataszewitz !® found that the absorption of water by the unfer-
tilized frog’s egg increased five times for every rise of 10° C., and con-
cluded from this that heat increased the permeability of the plasma
membrane to water.

Tue Erectric ConpuctTiviry OF MASSES OF EGGS.

This work was done at the Tortugas Laboratory of the Carnegie
Institution. The experiments were made on board the yacht “Phy-
salia’’ anchored off Boca Grande Key, twelve miles west of Key West,
Fla. Iam indebted to Dr. W. R. Warren of Key West for the use of a
centrifuge, as the one provided was left behind. My thanks are also
due Dr. Alfred G. Mayer for the unusual facilities at my disposal.

The determinations were made by Kohlrausch’s method. A re-
sistance box of 15,000 ohms and a metre sliding resistance were used.
A number of difficulties arose in eliminating possible sources of error,
and these will be considered in order: 7

First. The procuring of sufficient quantities of suitable eggs to

17 Brown: This journal, 1905, xiv, p. 354.

18 SOLLMANN: This journal, 1904, xii, p. 112.

19 BIATASZEWITz: Bulletin de l’Académie des Sciences de Cracovie, Math. —
Nat., October, 1908.

On the Dynamics of Cell Division. 251

make accurate determinations. This was met by going to Boca Grande
Key, where the sea urchins, Toxopneustes variegatus and the Trip-
neustes esculentus, could be picked up by the ton in shallow water.
As the ripe eggs of the former species were more abundant, they were
used exclusively.

Second. The handling of the eggs with sufficient rapidity to insure
their being in normal condition. Washing the eggs repeatedly by
allowing them to settle in sea water requires much time, though it will
be shown later that for the first washing this is an advantage.” But the
time required for them to settle into a compact mass for the conduc-
tivity determination must be shortened as much as possible, as the
continued crowding of the eggs might produce abnormal effects. No
suitable conductivity vessel on the market could be placed directly
in the centrifuge.

I made a conductivity vessel at the Tortugas Laboratory. It con-
sisted of two glass tubes, the inner one fitting nicely into the outer
one. Owing to a very slight curvature of the tubes, a rotation of one
in the other would clamp them tightly together. The inner tube was
131 mm. long, 10 mm. inside diameter, and sealed at the lower end.
Two small glass tubes were placed longitudinally within the outer tube
and sealed into its upper end. One of these projected 25 mm. below
the other. Platinum electrodes 6 X 9 mm. were sealed in the lower
ends of the two smaller tubes. The electrodes were “‘platinized”’ with
platinic chloride solution containing a little lead acetate.

The advantages of this conductivity vessel were the comparatively
large surface area of the electrodes, 108 sq. mm. each; the distance
between them, 25 mm.; the small volume of eggs, less than 3 c.c.,
required to cover the electrodes, and the short time required for its
contents to reach the temperature of the thermostat. The electrodes
were plane and vertical, and hence could be pressed down into a mass
of eggs with the least possible disturbance of them. The inner tube,
containing the eggs, could be placed directly in the centrifuge, and
thus the eggs could be washed with sea water or other solutions without
removal from the conductivity vessel. By inserting the lower end of
the outer tube into a short, closely fitting test tube, the electrode could
be protected from drying while the inner tube was in the centrifuge.

© The eggs of Toxopneustes are of but very little greater specific gravity than
sea water and hence settle much slower than those of Tripneustes or Arbacia.

252 J. F. MeClendon.

Third. The temperature of the thermostat, if constant, would on
some days be very far from that of the sea surface, which varied
greatly (from about 25°-30° C.), and the sudden change of the eggs
from one to the other might affect them in some undesirable way, as
they could be caused to segment by a change in temperature; also the
time required for the eggs to reach this temperature would be greater.
This was obviated by making the temperature of the thermostat about
one degree higher than that of the sea before each set of determinations
and then keeping it constant, within one tenth of a degree. The ther-
mostat held 20 litres of water, was closed at the top and well insulated
at the sides, stirred with a paddle attached to a rod going through a
small hole in the cover, and regulated by hand with a minute flame
beneath.

Fourth. The spaces between the eggs might vary. This was ob-
viated by centrifuging the eggs in the conductivity vessel and marking
their upper limit accurately in indelible ink with the finest drawing
pen, and before each reading centrifuging them again to the same.
line.

The egg as it leaves the ovary is surrounded by an invisible gelati-
nous covering, which I have called the “‘jelly,” since it shows similarity
to the jelly-like coat of the frog’s egg, a mucin which yields galacto-
samin (Schulz and Ditthorn). Loeb applies the name chorion to it.
This jelly seems to be a mucin which in sea water slowly dissolves.
The solution of the jelly is aided by weak or strong alkalis and very
weak acids, though it is coagulated by tannin and basic dyes, and ©
seems to contract (coagulate?) on addition of strong mineral acids.
Delage says it is lifted from the egg before the formation of the fertil-
ization membrane. It is possible that this appearance was due to
contraction caused by the dye used to make it visible.

If eggs bearing this jelly are put in the conductivity vessel and
centrifuged, and later mixed with sea water and centrifuged again,
much less force is required to precipitate them to the same level, and
they will slowly settle below this level by gravity while the eggs are
being brought to the temperature of the thermostat. This is due to
the washing away of some of the jelly. To prevent this occurrence,
the jelly had to be entirely washed off of them before they were first
placed in the conductivity vessel, a process accomplished by stirring
in large quantities of sea water and repeated centrifuging. The

ead

On the Dynamics of Cell Division. 253

jelly could be removed completely from Toxopneustes eggs and with
more difficulty from Tripneustes eggs. Care was taken that no treat-
ment was used that would cytolyze even a few eggs, as cytolysis causes
swelling or disintegration which would affect the volume of the eggs
and therefore the spaces between them when they were precipitated
to a certain level.

The pushing out of fertilization membranes might affect the shape
of the spaces between the eggs and thus change the free paths of the
ions in the sea water filling those spaces. This was obviated by wash-
ing the eggs so long in sea water that no fertilization membranes
were pushed out when they were fertilized or treated with the solu-
tions used. This was tested by microscopic examination of control
eggs and of the eggs taken from the conductivity vessel after each
experiment. Such eggs develop.

It has been objected that a membrane was formed and not pushed
out. I found two methods of detecting membranes that lie so close to
the egg as not to be distinguishable with the microscope. If the egg
be plasmolyzed with a molecular solution of cane sugar, such mem-
branes are often if not always lifted from the egg. Harvey says the
fertilization membrane is relatively impermeable to sugar, and one
would suppose that sugar would push the membrane closer to the egg.
From many of my experiments it is evident that sugar will go through
the fertilization membrane, and Harvey has not stated the degree of
impermeability to sugar that he observed. The second and more
certain method is as follows: If an electric current of sufficient density
be passed through, the membrane will be lifted from the cathode end
of the egg.

It may be objected that the sugar solution and electric current
caused the membranes to be formed, and that they were lifted from the
egg in the usual manner. I can only answer to this that the same
sugar solution and electric current were tried on normal unfertilized
eggs and no membranes appeared.

Under the same conditions as those of the later conductivity ex-
periments, eggs did not form membranes that could be detected by
either of the above methods, and similar eggs developed. Harvey *
says that after an egg stands in sea water twenty-eight hours and is
then fertilized, it becomes surrounded by a thick adhering membrane

*1 HARVEY: Journal of experimental zodlogy, 1910, viii, p. 365.

254 J. F. McClendon.

which, on cleavage of the egg, surrounds each blastomere. He calls
this a fertilization membrane, and maintains that fertilization mem-
branes are formed on all developing sea urchin eggs. Evidently in
the above case and in that of eggs placed in hypertonic or calcium-
free sea water he mistook the so-called “hyaline plasma layer” or
“‘Verbindungsschicht”’ for the fertilization membrane.” Harvey states
that he saw membranes on Hipponoe eggs on slightly “high focus.”
A membrane on the surface of a sphere could only be determined
positively with so high a power that the optical] section was extremely
thin in comparison to the diameter of the sphere, and passed exactly
through the point of contact, with the sphere, of a tangent drawn
from the eye to the sphere. Hence there is only one focus, and “high”
or “low” focus means out of focus. One need only try this “high
focus” on an air bubble or oil drop to see what appearances of ““mem-
branes” are thus obtained.

But it is impossible for me to see how a change in a surface film,
of immeasurable thickness, of the egg, thus forming a “‘membrane”
in close contact with the egg, can cause such a change in the conduc-
tivity of the inter-egg spaces as to account for the great differences in
the conductivity of unfertilized and fertilized eggs which I obtained.
I account for the change in conductivity by a change in the surface
film of the egg, allowing ions to pass through more easily. Probably
such a change would be accompanied by a visible change if micro-
scopic technique were sufficiently developed to detect it, but I would
not call this the formation of a new membrane; it is a change in the
‘“‘tlasma membrane,” a condensed surface tension film or haptogen
membrane.

The conductivity of the spermatic fluid and of the acidulated sea
water were slightly less than that of natural sea water, so that if they
replaced natural sea water between the eggs, they would cause a
slight decrease in the conductivity reading, but they were thoroughly
washed out with natural sea water before the readings were taken.

Fifth. The electric current passed through the eggs might alter
their conductivity. As no measurement of conductivity could be taken
before the current began to pass through, I cannot determine this
point. But the first reading and later readings taken at short intervals
were always the same provided the eggs had been centrifuged down

2 See GotpscumipT and Poporr, Biologische Centralblatt, 1908, xxvili, p. 210.

On the Dynamics of Cell Division. 256

so compactly that they did not settle further by gravity, and the
content of the conductivity vessel was at the same temperature as the
thermostat. By using a special induction coil with a rheostat, an
alternating current of such high frequency and such low amperage
was obtained that when passed through my finger from electrodes
wet with sea water, I could not feel it, yet this was the current used
in the experiments, and variations in it could easily be detected with
the telephone used in the experiments.

Sixth. The increased elimination of carbon dioxide by fertilized
eggs might cause an increase in conductivity of the sea water between
the eggs. To test this, the conductivity of a sample of sea water was
determined. It was then charged with CO, in a “‘sparklet syphon,”’
and no increase in conductivity could be detected with the electrodes
used in the experiments with eggs. Perhaps with electrodes specially
adapted to good conductors like sea water, a change could be detected,
but it would evidently be extremely small and incapable of account-
ing for the large differences observed in the experiments with eggs.

THE CONDUCTIVITY DETERMINATIONS.

The conductivity of one sample of sea water at 30° C. was found to
be .o61, while that of unfertilized eggs washed in the same water and
precipitated by gravity was .04655 at the same temperature. The
conductivity of the same water at 32° C. was .o624 and of the same
eggs at 32° C., .o469. At 26° the conductivity of another sample of
sea water was .05535, of spermatic fluid direct from testes, .o439, and
of unfertilized eggs precipitated with the centrifuge, .co2404. In this
case the conductivity of the eggs was about one twentieth that of the
sea water, although they still contained some of the latter between
them. The conductivity of the eggs would have fallen still lower if
all of the sea water had been pressed out from between them.

First lot of experiments. — In all of these preliminary experiments
the thermostat was kept at 32° C., which was at times very near,
but at other times very much above, that of the sea. The eggs
were precipitated by gravity. The supernatant sea water in the con-
ductivity vessel was pipetted off, and the vessel shaken, then read-
ings taken, until successive identical results showed that the eggs
were of the temperature of the thermostat, after which less than a

256 J. F. McClendon.

drop of sperma was added, the vessel shaken again, and a second
series of readings taken. The observed conductivities are given in
the following table:

1. Unfertilized eggs 05980

. Fertilized eggs .16950
2. Unfertilized eggs 04480
Fertilized eggs 04835

3. Unfertilized eggs .04690
Momentarily heated .05600 2

The above figures show a great increase in conductivity on fertili-
zation, or momentary heating to a point that will cause segmentation.

Second lot of experiments. — In this and all subsequent lots of ex-
periments the thermostat was brought before each set of readings
to about sea temperature, the eggs were precipitated in the conduc-
tivity vessel to such a degree that they were not further precipitated
by gravity.

1. Unfertilized eggs at 27.5° C. 01524

Fertilized exes 9)" eg 01627
2.-“Untertilized €ges.4** 287°C; .01076
Fertilized eggs “ iG .01266

Third lot of experiments. — As it was feared that the admixture of
but a fraction of a drop of spermatic fluid might raise the conduc-
tivity independent of fertilization, and as by this method only a small
per cent of the eggs were fertilized, in the third and fourth lots of ex-
periments the eggs were centrifuged in the conductivity vessel and
their upper level marked accurately, and the conductivity determined,
then they were mixed with sea water containing sperm or acetic acid
in the conductivity vessel, and washed by repeated precipitations and
precipitated down to the same level, before determining the conductiv-
ity of the developing eggs. By this method almost 100 per cent of the
eggs could be caused to begin development.

1. Unfertilized at 26° C. .002404
Fertilized ‘‘ sid .004320

23 The vessel was set a few moments in water at 45°, then returned to the ther-
mostat; this was shown by control to cause segmentation.

~
~~

the Dynamics of Cell Division.

Unfertilized at 25.75° C.
Fertilized ee Me

. Unfertilized at 25° C.

Fertilized ‘“ %:
Unfertilized at 26.25° C.
Fertilized “ _
Unfertilized at 28° C.
Fertilized ‘“‘ je
Unfertilized at 28.17° C.

Fertilized ey er

Unfertilized at 26° C.

After .006 normal acetic
acid in sea water 114
mim, 20° ©;

004445
.006340
.006523
009544
.004900
.006390
.008230
.009220
.006876
007298
005620

006000

25/7

Fourth lot of experiments, in which all precautions were taken. — In
these experiments the eggs were washed so long in sea water that

no fertilization membranes could be caused to push out.

.01182
01537
OL TSS

Lie

Z-

Unfertilized at 29.5° C.

Fertilized. ‘ #

Unfertilized at 30° C.

After .006 normal acetic
acid in sea water 114
min. 30° C.

Unfertilized at 30° C.

After acid sea water 1144
min.g0° €:

4. Unfertilized at 29.5° C.

After acid sea water 114

min, 29.5° C,

0127
00877

.00965
005135

.00839

From the above experiments I have concluded that there is an in-
crease in electric conductivity of the sea urchin’s egg at the beginning
of development. The question now arises whether the resistance to
the movement of ions through the mass of eggs is the impermeability
of the plasma membrane or the presence of fat globules or proteid
granules within the egg, or the combination of the egg electrolytes
with colloids, forming poorly dissociated or poorly diffusible com-
pounds. I found by centrifuging them that there was as great a vol-
ume of fat globules and proteid granules in the sea urchin’s egg im-

mediately after, as there was immediately before, fertilization.

258 J. F. McClendon.

ON THE INTERNAL CONDUCTIVITY OF THE CELL.

The majority of my experiments on this subject were made at Cor-
nell Medical College.

Hober ™ has devised a method by which the electric condaee ae of
the cell interior may be measured without breaking the cell wall. The
determinations cannot be made with great accuracy, but the results
on blood corpuscles clearly demonstrate that the conductivity of the
interior is many times greater than that of the corpuscle as a whole,
indicating that the greatest resistance to the current lies in the plasma
membrane.

Stewart made certain determinations which I take to indicate
that hen’s egg yolk is a very much poorer conductor than is a solution
of its salts made up to the same volume.

The yolk of the hen’s egg before it leaves the ovary forms the bulk
of a single cell. The yolk of an egg is often considered as ‘‘dead”
material, but in conductivity experiments we cannot separate “liv-
ing’? and “‘dead”’ portions of the cell. There is a small amount of
white yolk, but the major volume is yellow yolk.

The yellow yolk under the microscope presents a fluid matrix con-
taining large globules of another fluid of almost the same specific
gravity, viscosity, and refracting index as the matrix, the boundary
between the two fluids being the seat of little surface tension. Both
matrix and globules contain numerous fine granules. On the addition
of alcohol, under the microscope, a substance (or substances) in both |
fluids disintegrates, setting free lipoids which appear as droplets, which
are blackened by osmic acid and colored by Sudan III. If one of the
large globules is watched closely as the alcohol is applied, fine, lipoid
droplets appear and grow and fuse to form larger drops, and some of
them may then migrate toward the periphery and fuse to form a lipoid
envelope surrounding the globule. The lipoid droplets appearing in
the matrix grow and fuse to form larger ones.

The yellow yolk darkens after the addition of osmic acid, the glob-
ules becoming darker than the matrix, but if alcohol and then osmic
be applied, the lipoid droplets thus formed, quickly become an in-

*4 HOspeR: Archiv fiir das gesammte Physiologie, 1910, Cxxxill, p. 237.
25 STEWART: Journal of experimental medicine, 1902, vi, p. 257.

On the Dynamics of Cell Division. 259

tense black. Only the lipoid droplets are colored by Sudan III in 80
per cent alcohol.

The white yolk resembles the yellow yolk that has been treated
with alcohol in that it contains lipoid droplets. But as there is very
little white yolk in the hen’s egg before incubation, there are very few
lipoid droplets to impede the electric current. The lipoids in the yellow
yolk are probably bound up with proteids, forming combinations
which are disintegrated by alcohol, as indicated by the above
observations. ;

In order to determine to what extent the granules impede the electric
current, I precipitated them with the centrifuge. The large globules
cannot be separated from the matrix by this means. With small quan-
tities I was able to obtain the fluid entirely free from granules. It
forms 11/17 of the total volume, dissolves in dilute alkalis, and with
slight milkiness in dilute acids, and when shaken with water the
insoluble portion forms an emulsion or a coagulum resembling yeast
plants.

With large enough quantities to fill the smallest suitable conduc-
tivity vessel at hand, the precipitation was so slow that I feared de-
composition might commence before a granule-free fluid was obtained,
so I contented myself with the comparison of a portion containing a
very small per cent of granules with a portion containing a very large
per cent of granules.

At 25° C. the conductivity of the granule-poor layer was .00302
and that of the granule-rich layer .co278, showing that the granules
impede the current to a great extent.

Since dilution with water breaks up many ion-colloid compounds, I
used this method to determine whether the electrolytes in the yolk
were bound up with colloids. The conductivity determinations at
25° C. are given in the table below:

Breen er (amare Ct sok Buicihaiie ENO) Che 0)
Granule-poor .00302 .00268 .00162 .00096
Granule-rich .00278 .00278 .00200 .00125

The above table shows that whereas the granule-poor layer de-
’ creases in conductivity on dilution with distilled water, at first slowly
and later slightly more rapidly (which may be partially accounted
for by the more rapid increase in ionization of inorganic salts at the

260 J. F. McClendon.

beginning than at the end of the series) the conductivity of the granule-
rich layer is not reduced at all by a dilution with one volume of H.O.
It may be said that this is due to the separation of the granules, thus
widening the conducting paths, and I have demonstrated that such
might occur in an emulsion of oil in soap solution, as shown by the
following table of conductivities at 25° C.:

Matsa, aged) Ch LILO) (+ 5 vols THO)
Soap solution .002490 .001460 .000903
Emulsion of oil, containing
17 per cent soap solution .000434 000434 .000335

But how are we to explain the fact that on dilution with one or more

volumes of water the conductivity of the granule-rich portion of yolk
is greater than that of the granule-poor layer; although the former
contains less of the fluid portion of the yolk? Evidently (since there
are no inorganic crystals in the yolk) some of the electrolytes must
have been bound up in the granules (either by adsorption or chemi-
cal combination) and liberated on dilution. Since the fluid portion of
the yolk does not entirely dissolve in water, the undissolved portion
may impede the current, but this would occur in the granule-rich as
well as in the granule-poor portion.

I doubt that Héber’s method of measuring the internal electric con-
ductivity of cells is sensitive enough to determine whether the increase
' in conductivity of the egg is due to liberation of electrolytes in the
interior or to increased permeability of the plasma membrane of the
egg, but it shows, by exclusion, in case of the cells on which it was used,
that by far the greatest resistance to the current lies in the plasma
membrane.

Swelling (first stage of cytolysis) of sea urchin eggs causes a decrease
in the conductivity of the mass of eggs, as shown by the following de-
-terminations of the conductivity:

1. Unfertilized eggs of Toxopneustes at 27.25° C, .01354
After addition of nicotine at 27.25° C. .01318
2. Unfertilized eggs at 32° C. .04850
After momentary elevation to about 50° C. at 32° .04780
3. Unfertilized eggs at 27.5° C. .01645
After momentary elevation to about 50° C. at 27.5° .01626
4, Unfertilized eggs of Tripneustes at 32° C. .04730
After shaking with fraction of a drop of chloroform at 32° C. .02286

Eas:

On the Dynamics of Cell Division. 261

Microscopic examination showed that the addition of nicotine or
chloroform, or momentary elevation of temperature in the above ex-
periments, caused the eggs to swell. This could only take place by
the absorption of one or more constituents of the sea water between
the eggs. If the salts of the sea water did not go into the eggs, the ab-
straction of H,O from the sea water would increase the concentration
of salts in the sea water remaining between the eggs, and might cause
a liberation of electrolytes (by dilution) within the eggs, in the latter
case causing increased conductivity of the egg interior without a cor-
responding decrease in conductivity of the inter-egg spaces. If the
membrane became freely permeable to salts, the swelling of the eggs
might increase, but should not diminish, the conductivity of the mass.
The fact that the conductivity decreased can only be explained by
assuming that the salts of the sea water entered the eggs and were ad-
sorbed to or combined with colloids, or that the membrane was very
poorly permeable to salts (though perhaps more permeable than the
normal egg) and the narrowing of the inter-egg spaces caused the de-
crease in conductivity. In fact, I think this can be taken as an indi-
cation that the egg is a poor conductor not so much because of the low
concentration of free electrolytes within it, but chiefly because the
electrolytes cannot easily pass the plasma membrane.

As no dilution of the contents (swelling) of the sea urchin’s egg
occurs at the beginning of development, it is improbable that a libera-
tion of the electrolytes within it, sufficient to account for the increased
conductivity, occurs. The only alternative is that the increase in con-
ductivity is due to an increase in permeability of the plasma mem-
brane to electrolytes.

‘THE ELECTRIC CONDUCTIVITY OF INDIVIDUAL EGGS.

It is well known that cells may be killed or injured by the passage
of electric currents through the media containing them. The current
might affect them by raising the temperature, by the passage of ions
into or out of the cells, by the accumulation of ions of one sign that
are stopped by parts of the cell.

In which of these ways does the current affect the cell most de-
structively? The heating effect may be practically eliminated. If
electrolytes are transported into or out of the cell by the current, they

262 J. F. McClendon.

could also diffuse in the absence of a current, but the accumulation of
ions of one sign would not occur by free diffusion. Therefore I have
regarded the accumulation of ions impeded by the cell structures as
explanation of the destructive effects, and the destructive effects as
an indicator of the resistance to the passage of ions.

The experiments were made at the United States Bureau of Fish-
eries at Woods Hole, Mass. The 110-volt direct current from the
light circuit was used. Cylindrical non-polarizable electrodes of cop-
per in one-half molecular copper sulphate were plugged at their free
ends with absorbent cotton and connected to rubber tubes of 4 mm.
internal diameter and about one foot each in length, filled with sea
water. The free ends of the rubber tubes were plugged with absorbent
cotton, which was allowed to protrude sufficiently to conduct the cur-
rent to the sea water containing the eggs under a cover glass on a slide
on the microscope stage. The current was reduced by passage through
a 16-candle power light and further regulated by turning the screw of
a pinch-cock which was clamped on one of the rubber tubes. The

,copper sulphate diffusing into the sea water in the rubber tubes re-
acted with the calcium carbonate, copper hydrate and calcium sulphate
being precipitated and carbonic acid being liberated. The copper
sulphate solution and sea water were renewed before each experiment.
Usually a piece of ash-free filter paper was cut the size of the cover
glass and a hole cut in its middle. This filter paper ring was placed on
the slide, and sea water containing eggs placed in the hole in the ring,
so that when the cover glass was placed on the preparation, the eggs
were contained in a cell which was freely permeable to ions at the
sides. At other times the eggs were mixed with sea water containing
enough cotton fibres to support the cover glass. Eggs of Arbacia punc-
tulata were used.

When the current is passed through the egg, the latter is affected
at that surface nearest the anode, as observed by Brown, who placed
the eggs in a molecular solution of urea. Changes in surface tension
are indicated by bulging or amceboid movements. The pigment sud-
denly leaves each of the chromatophores in turn and diffuses into the
cytoplasm in this region of the egg, which is turned a red or orange
hue (it is a deeper red if the chromatophores have been stained with
neutral red), showing that the reaction is not alkaline. The anodal

** The pigment extracted from the eggs is red or orange in acid according to
dilution; it is violet or green and precipitates in alkali. If the eggs, or especially

On the Dynamics of Cell Division. 263

end of the cell absorbs water and swells, often a blister is formed and
masses of granular cytoplasm pass into the blister fluid and dissolve.
Gradually these changes extend from the anode end to the cathode
end of the egg, the egg swells enormously and may burst.

Very probably this disintegration commencing at the anode end of
the egg is due to the accumulation of anions which cannot pass the
plasma membrane. If the plasma membrane is poorly permeable to
anions in one direction, it is probably so in the other, and it may be
asked why they do not accumulate outside the cell at its cathode end.
The anions which are unable to enter the egg at its cathode end are
free to move around the egg and hence do not accumulate to form
as great a concentration as at the anode end.

Since no destruction of the egg of Arbacia, beginning at the cathode
end, was observed, we may conclude that the plasma membrane is more
permeable to cations than to anions. This is not true of all eggs, as I
observed the eggs of Hydractinea begin to disintegrate at the cathode
as soon as at the anode end. However, it is true of a number of living
cells.?”

If fertilized and unfertilized Arbacia eggs are placed in an isotonic
sugar solution containing little sea water, through which a current of
gradually increasing density is passed, the unfertilized eggs begin to
disintegrate, at their anode ends, sooner than the fertilized eggs do.
We may interpret this as indicating that the fertilized eggs are more
permeable to anions, which therefore accumulate in them to a less
extent, or the fertilized eggs are more permeable to electrolytes, which
therefore have passed out into the sugar solution to a greater extent,
and therefore the current passes through them less, than in case of the
unfertilized eggs.

I did not obtain the same results on eggs in sea water, but the un-
certainty of the material toward the end of the season prevented the
determination of the mode of action of the sugar solution. Possibly
the heating effect of the current in sea water increased the permeabil-
ity of the unfertilized eggs. Sea water is so much better a conductor
than the eggs that only a small per cent of the current passes through
the latter, and in order to produce visible effects on the eggs an enor-

the perivisceral fluid cells containing much pigment, are killed, the nuclei and
some other parts absorb the pigment and turn brownish purple.
*7 See VERWORN’S Physiology.

264 J. F. McClendon.

mous current must be passed through the sea water. It is known that
sugar solutions produce abnormal conditions in eggs, but these ex-
periments were made quickly after placing the eggs in the sugar solu-
tions. The nucleus does not begin to disintegrate as soon as the cyto-
plasm; this is in harmony with McCallum’s view that the nucleus
contains no free salts. The nucleus as a whole or the contained nucleo-
proteids migrate toward the anode.

PLASMOLYSIS WITH NON-ELECTROLYTES.

Osterhout has obtained shrinkage of marine cells in distilled water,
and thinks the action of sugar similar; 7. e., first the membrane is made
permeable and then the salts diffuse out and the cell contracts by
some non-osmotic force. But in the only animal cell in which he has
obtained this result there is first a swelling, with formation of blisters,
and later shrinkage, with the nucleus becoming homogeneous and
distinct, which, I think, denotes death and perhaps coagulation. Since
the Arbacia egg in an isotonic sugar solution does not swell first and
then shrink, I think this objection may not apply to my experiments.

The following tables show that fertilized (Arbacia) eggs shrink
more rapidly than unfertilized eggs in molecular sugar solutions,
which are calculated to have only slightly greater osmotic pressure
than the sea water at Woods Hole, where the experiments were made.
It appears that the plasma membranes of the fertilized eggs are more
permeable, allowing the salts to diffuse out of the eggs more rapidly,
thus lowering the internal osmotic pressure to a greater extent than
is the case with unfertilized eggs. Sollmann ?* observed Arbacia eggs
contract in hypertonic, and swell in hypotonic, salt solutions.

In normal sea water fertilized are not smaller than unfertilized
eggs.” Before the first cleavage the hyaline plasma layer forms, thus
taking material away from the more opaque portion of the egg, and
it might be supposed that the failure to include this layer in taking
the measurements caused the appearance of shrinkage, but such would
be the case also in the control in normal sea water, and furthermore
the measurements were taken before the hyaline layer was formed or
had reached visible thickness.

28 SOLLMANN: This journal, 1904, xii. p. 111.
*9 McCLENDON: Science, 1910, xxxii, p. 318.

On the Dynamics of Cell Division. 265

Fertilized and unfertilized eggs in a molecular solution of dextrose
were placed under the same cover glass, which was supported to pre-
vent compression of the eggs, and sealed to prevent evaporation. The
fertilized were distinguished from the unfertilized eggs by the presence
of the fertilization membrane. The eggs were observed in the order
in which they appeared in the field as the slide was moved so as to
observe the whole area under the cover glass once and once only. The
diameter of each egg in turn was drawn with the camera lucida, and
the drawings were measured later with a rule. In case an egg was
irregular, approximately its mean diameter was drawn.

The results of two series of measurements are recorded on page

266. In the first column of figures the diameter of the egg in the
unit used for all the measurements is represented. In the second and
third columns of figures the frequencies of the occurrence of fertilized
and unfertilized eggs of the diameters given in the same horizontal
line are represented. The fourth and fifth columns of figures repre-
sent a second series of measurements in the same manner.
- The table shows that there is considerable variation in the
size of the eggs, but that the mean (and also the mode if the curve
were plotted) of the diameters of the fertilized eggs is less than the
mean of the unfertilized eggs. I did not succeed in making measure-
ments fast enough to determine the rate of plasmolysis.

CHEMICAL ANALYSIS OF MASSES OF CELLS.

Fertilized and unfertilized eggs may be placed in solutions differ-
ing from sea water, and the passage of substances into or out of them
detected by analysis of masses of the eggs. There are three sources
of error: 1. The presence of the jelly-like coverings on the eggs;
2. The fluid in spaces between the eggs; and, 3. The large surfacé for
adsorption.

I tried some preliminary experiments on yeast cells at a time when
suitable eggs could not be oktained. I found yeast and dextrose
placed in .3 molecular MgCl, eliminated CO, more rapidly than in .5
molecular NaCl or .325 molecular CaCls, all of which are calculated to
. have approximately the same osmotic pressure. Also the CO, elim-
ination was more rapid in the magnesium solution than in a solution
- of the same concentration of magnesium chloride with either of the

266

Diameter.

Mean diameter

J. F. McClendon.

Frequency.

Fertilized. Unfertilized.

Fertilized.

Unfertilized.

rwMWnND NFR NM CO WNW

e e
Senet

7
6
+
+
1
3
0
2
0
0
1

Oe ORO OO rai es NS 103i 1Go) (Got ON St) (Con Ort! “ICN ON ON Cs SOT Cola Ee ais

FS (Ol) OS: Eas) Gomer None ONG) ac S Co) HES PRBCOIOU IEE arae ene

RPNON NY PWWHN UN PNY WWNH HY WWNH ON OCF:

On the Dynamics of Cell Division. 267

other salts in addition, or in a solution containing NaCl and CaCl, in
the same concentration as in the respective pure solutions, or in a solu-
tion containing all three salts, or in tap or distilled water.

The magnesium must have entered the cell or altered the permea-
bility of the plasma membrane to COs, sugar, alcohol, the enzyme, or
some other substance. In order to determine whether the magnesium
entered the cells, I took two blocks of compressed yeast of the same
volumes and weights and mixed one with H.O and the other with a
molecular solution of MgCl, for five hours, then washed each by rapid
precipitation in renewed H;O several times with the centrifuge. The
two lots were ashed and weighed with the results: control, ash = .0466
gm.; ash from Mg culture = .048 gm. Evidently the magnesium
did not enter the yeast to any great extent and probably acted on the
surface, increasing the permeability to some other substance.

Lyon and Shackell have analyzed fertilized and unfertilized eggs
placed in salt solutions, and obtained some results indicating that the
salts enter and leave the fertilized more easily than the unfertilized
eggs. They found an exception in the case of iodine. Iodine (in po-
tassium iodide solution) is absorbed by the unfertilized more quickly
than by the fertilized eggs.*°

I had intended to work along this line, but was forced to postpone
it until another season.

MIcROCHEMICAL ANALYSIS OF INDIVIDUAL EGGs.

Lyon and Shackell *° and Harvey have concluded that certain dyes
enter fertilized more easily than unfertilized eggs. Loeb supposes that
the dye is chemically combined in the fertilized egg and merely in solu-
tion in the unfertilized egg. Unfortunately these dyes belong to the
class of substances which Overton found to most easily penetrate plant
cells, so that a demonstration that they more easily enter the fertil-
ized than the unfertilized egg does not necessarily indicate that the
same is true for electrolytes in general.

Harvey * found that eggs became more permeable to NaOH after
being fertilized or treated with cytolytic agents.

30 Lyon and SHACKELL: Science, 1910, xxxii, p. 250.
Sl TIARVEY: ‘Science, Lolo, XxXxIl,.pi 565.

268 J. F. McClendon.

THe MIGRATION OF THE CHROMATOPHORES.

The chromatophores of the egg of Arbacia contain a red substance
which I found to have an absorption spectrum similar to McMunn’s
echinochrome, at least in certain solvents. I have crystallized two
derivatives of the Arbacia pigment and perhaps the pigment itself,
and a chemical study of it is being attempted.

These chromatophores or pigment plastids show similarities to the
chloroplasts of some green plants. Similar plastids occur in the peri-
visceral fluid cells of Arbacia, where they are so closely packed to-
gether in the cytoplasm as to be separately distinguishable only on
careful observation. In some of the cells the plastids contain pigment
and in others they are colorless. |

McMunn, finding that the spectrum of echinochrome in certain
solvents was changed by strong reducing agents, concluded that it was
respiratory in function. Griffiths ® briefly states that on boiling with
mineral acids echinochrome is transformed into hemochromogen,
hematoporphyrin, and sulphuric acid, indicating a relation to
hemoglobin.

I separated the cells from about 50 c.c. of the perivisceral fluid of
Arbacia and mixed them with sea water to form 50 c.c. This suspen-
sion of cells, and 50 c.c. of sea water as a control, were exhausted under
an air pump for six hours, during the last half hour at practically
water vapor tension. While in the vacuum, the cells must have exerted
a reducing action on the pigment if it can be reduced. Each was then
shaken with air for thirty minutes in closed apparatus. The suspension
of cells had absorbed 1.25 c.c. of air and the control only 0.8 c.c., at
atmospheric pressure. The volume of oxygen used in oxidations in
the cells during the shaking was probably partly replaced by CO:
given off by them, but the difference of about half a cubic centimetre
does not demonstrate conclusively that the pigment combined with
oxygen. Under somewhat similar conditions dogfish blood absorbed
many times as much air as the perivesceral fluid of Arbacia.

The migration of the chromatophores in the egg is evidently not
always in the direction of greater oxygen concentration, but whether
it is ever a chemotropism toward oxygen I was unable to determine.

8 GRIFFITHS: Comptes rendus, 1892, cxiii, p. 419.

On the Dynamics of Cell Division. 269

In 1908 I observed movements of the chromatophores in the eggs
of Arbacia punctulata. As Roux had caused a whitening of the catho-
dal pole of the frog’s egg by passing an electric current through it, I
tried in 1909 and again in 1910 to move the chromatophores of the
Arbacia egg with the electric current. I observed that in the unfer-
tilized egg the chromatophores are distributed throughout the cyto-
plasm, but after the egg is fertilized or stimulated artificially the
chromatophores migrate to the surface.®

Harvey * says that the pigment comes to the surface within ten
minutes after fertilization, but I found that this process sometimes
required half an hour, by which time the cleavage spindle had formed.
At each cleavage chromatophores sink into the cleavage furrows of
the blastomeres. Just before the micromeres are formed the chroma-
tophores move along the surface of the blastomeres, away from the
micromere pole of the egg, so that after the resulting cleavage the
micromeres are practically free from pigment. Under abnormal con-
ditions there is a great massing of pigment in the cleavage furrow or
other regions of the surface or in the interior of the egg. The sinking
of pigment into the cleavage furrows and its retreat from the micro-
mere pole are probably due to surface tension changes as discussed
above, and perhaps the abnormal massing of pigment at one portion
of the surface is due to a local increase in surface tension.

““Membrane-forming”’ and parthenogenetic agents, even in concen-
trations too low to produce membranes or segmentation, cause the
pigment to come to the surface. If a few normal unfertilized eggs are
kept in a relatively large amount of sea water protected from evap-
oration, and oxygen is very abundant, it appears that there is more
pigment at the surface after twelve or more hours than at the begin-
ning of the experiment, but disintegration commences before all the
pigment has reached the surface. In an oxygen vacuum this did not
seem to occur. The pigment may all come to the surface in a stream
of washed hydrogen, but this may be caused by some impurity.

Fischel,® observing similar movements of pigment in the egg of
Arbacia pustulosa, concluded that the pigment was repelled by the
asters according to the forces described by Rhumbler as moving

33 McCLENDON: Science, 1909, XXX, Pp. 454.
34 HARVEY: Journal of experimental zodlogy, 1910, vili, p. 355.
3° FISCHEL: Archiv fiir Entwicklungsmechanik, 1906, xxii, pp. 526-541.

270 J. F. McClendon.

granules toward or away from asters in the cytoplasm.*® Biitschli and
Rhumbler have shown how the contraction of an area in a foam struc-
ture causes aster-like radiations around it, and Rhumbler has shown
that such radiations to a limited extent may occur around a rigid
sphere inserted into a foam or alveolar structure. Rhumbler assumes
that the concentration of the alveolar wall substance would increase
its surface tension, and that this increase toward the centre of the
aster would reduce the thickness of those alveolar walls perpendicular
to the astral rays, both of which assumptions have no facts of which I
am aware to support them,. On them rests Rhumbler’s explanation
of the movement of granules away from asters.

However, if those bodies which seem to be repelled by asters (chro-
matophores of Arbacia eggs, yolk platelets of frog’s eggs) lie within
or are larger than the largest alveoles, as I have observed to be the
case, aster formation might explain their repulsion. Rhumbler’s
theoretical asters were made of a central body and of alveoles of a
uniform size. If the alveoles were of different sizes, the largest ones
would seek the periphery of the aster.

Isectioned eggs that had been so treated artificially that all of the
pigment came to the surface but no segmentation occurred, and found
no asters, though perhaps asters had formed and disappeared.

After the passage of an electric current of a certain density and
duration through unfertilized eggs, some of them have their pigment
more abundant toward their cathodal surfaces. If the current exceed
a certain density, one by one the chromatophores toward the anodal
surface of the egg lose their pigment suddenly. When the current
was slowly and carefully increased just to the density required to
change the distribution of the pigment, no loss of pigment by the
chromatophores toward the anode could be observed, but it is mechan-
ically impossible to watch every chromatophore in the anodal region of
one egg. I found it possible to observe a single chromatophore for a
long time, and attempted, by noting its distance from the anodal sur-
face of the egg, to record its movements. Each time this observation
was attempted the chromatophore appeared to move, but its move-
ment was not constant in direction, and a considerable migration in
any one direction was not observed, except rarely in case the chro-

6 RuUMBLER: Archiv fiir Entwicklungsmechanik, 1896, iii, p. 527; 1899, ix,
pp. 32 and 63.

On the Dynamics of Cell Division. 271

matophore was very near the surface. In this exceptional case the
chromatophore moved along the surface, toward the cathode, which
movement was probably due to surface tension changes. In the egg
just taken from the ovary the chromatophores are slightly more nu-
merous near the surface than in the interior, and when the current is
passed, this difference is increased. The passage of the current
causes the anodal surface of the egg to spread (the increased differ-
ence of potential between the two sides of the surface reducing the
surface tension), sometimes carrying the more superficial chromato-
phores along the surface toward the cathode. This is not a cataphore-
sis of the chromatophores, since they do not go in the direction of the
current, but is due to surface tension changes, and is therefore a
secondary effect of the current.

Fearing that the high viscosity of the cytoplasm might interfere
with the movement of the chromatophores by electric convection, I
centrifuged both fertilized and unfertilized eggs until the pigment
was massed at one pole of each, and passed the current through solu-
tions containing them. No orientation of the eggs to the potential
gradient occurred. I then tried to move the perivisceral fluid cells,
which are practically masses of chromatophores, by means of the
electric current, but my apparatus did not exclude all sources of error,
and this experiment was reserved for another season. The pigment
may be caused to leave the chromatophores in these cells by the elec-
tric current or by chemicals, to which agents these cells are much less
sensitive than are the eggs.

We have, then, no evidence that the chromatophores are electrically
charged.

Harvey *’ attempted to explain my observation that the chroma-
tophores come to the surface at the beginning of the development of the
egg, by assuming that there is a positive charge over the surface of
the egg until the commencement of development, when the surface
becoming permeable to anions causes a potential gradient between
the surface and centre of the egg. He further assumed that the chro-
matophores are charged negatively and migrate in the potential
gradient.

His evidence for the existence of the positive charge over the surface
of the unfertilized egg of Arbacia punctulata is the fact that it is not

37 Harvey: Science, 1909, XXX, p. 694.

272 J. F. McClendon.

always spherical when it leaves the ovary. His evidence for the loss
of the charge is the fact that this egg rounds up more rapidly when it
is fertilized than when it is left in sea water without sperm. His evi-
dence for the negative charge on the chromatophores is the fact that
they come to the surface after development commences.

My observations indicate that the plasma membrane of the unfer-
tilized egg is less permeable to anions than to cations, which would
cause the appearance of the positive charge over the surface provided
some electrolyte whose undissociated molecules could not easily pass
the membrane was more concentrated in the egg than in the sea water,
or was produced with sufficient rapidity within the egg. Carbon diox-
ide might be this substance. However, my observations seem to indi-
cate that the permeability of the egg is increased suddenly (in less
than five minutes) on fertilization, in which case the positive charge
over the surface would be lost suddenly, and if the ions within the egg
were free to move, the potential gradient would be of momentary
duration, whereas the chromatophores require from ten to thirty
minutes to come to the surface. Before Harvey made this hypothesis
I had attempted, as described above, to move the chromatophores by
inducing a potential gradient, in order to determine whether they
were electrically charged. Harvey has yet to prove that they are
charged, and furthermore that they are negative, and that the poten-
tial gradient is of sufficient intensity and duration to move them to the
surface. I do not wish to be considered an opponent of his hypothesis,
but am merely searching for facts. Garbowski observed chroma-
tophores move toward the centrosomes.

ON THE CONTENTS OF THE ‘‘PERIVITELLINE”’ SPACE.

The assumption has been made by several observers that there
exists a colloid between the fertilization membrane and the egg. Here
the question arises, what is meant by the surface of the egg? The “hy-
aline plasma layer,” or ‘‘Verbindungsschicht,’’ which forms before
the first cleavage, is considered by some as part of the egg and by
others as a “‘membrane”’ outside of the egg. In this section I will not
include the hyaline layer in speaking of the egg, as under these experi-
mental conditions the surface of the hyaline layer (if such had formed)

On the Dynanucs of Cell Division. 273

could not usually be distinguished, 7. e., the presence of this layer
could not be ascertained.

When an electric current is passed through the egg of Arbacia punc-
tulata having a ‘‘pushed-out”’ fertilization membrane, the latter is
bulged out toward the cathode, and the egg moved in the opposite
direction and pressed against the anodal portion of this membrane.
When the current ceases, the egg returns to the centre of the “‘peri-
vitelline space.”’ I first thought that this was caused by anodal elec-
tric convection of the egg, due to confined anions, but sometimes the
fertilization membrane bursts at its anodal pole and the egg passes
out, and should on this hypothesis continue its migration toward the
anode. But as soon as the egg is free from the fertilization membrane
it stops its migration, even though floating in a fluid of equal specific
gravity. Perhaps an invisible colloid having a positive charge fills the
perivitelline space, and its migration toward the cathode pushes the
egg in the opposite direction.

Loeb postulated a colloid in the perivitelline space as exerting an
osmotic pressure which pushed out the fertilization membrane. This
may be true, but the membrane must harden in the expanded condi-
tion, for if it is burst by passage of the electric current or other means
it does not collapse, but remains spherical unless distorted by violence.

When the electric current causes bulging of the fertilization mem-
brane, the perivitelline space exhibits fine striations radially to the egg
or parallel to the current lines. Schiicking, Goldschmidt and Popoff,
Herbst, and others have described striations or fibres in the peri-
vitelline space, or around the fertilized egg, including the spaces
between the early blastomeres, usually under abnormal conditions.
These striations are probably due to tension of the colloid filling
the perivitelline space (including the hyaline plasma layer or
“‘Verbindungsschicht”’).

THE ACTION OF PARTHENOGENETIC AGENTS ON THE PLASMA
MEMBRANE.

Salts, acids, alkalis, shaking and thermal or electric changes might
alter the aggregation state of the colloids of the plasma membrane.
Fat solvents, alkaloids, glucosides, blood sera, soap and bile salts

274 J. F. McClendon.

might alter the aggregation state of the colloids, especially lipoids of
the plasma membrane.

Lillie ** found that pure solutions of sodium salts were effective as
parthenogenetic agents in the following series arranged according to
the anions: Cl< Br<NO,<CNS<I. This order of anions is reversed
in the precipitation of lecithin, and a somewhat similar reversed
order of anions occurs in the salting-out of proteids.*? Hence it
appears probable that these salts act by virtue of their dissolving
power on the colloids of the plasma membrane.

Alkalis may act not only on the membrane, but by slow diffusion
into the egg favor oxidations, as an alkaline reaction favors oxidation
in general, and Loeb has shown that an alkaline reaction of the medium
is necessary for the normal oxidations in the sea urchin’s egg.

KCN or an oxygen vacuum may act by suppressing oxidation until
enough of the confined CO, can escape to raise the alkalinity within
the egg to such a point that when oxygen is readmitted oxidation may
proceed with sufficient rapidity to allow the development of the egg.

It seems probable that the undissociated molecules of carbonic acid
or CO, can diffuse out of the egg at all times. How then could the
resistance of the plasma membrane to one or both or its ions so reduce
oxidation within the egg as to prevent its development? Perhaps the
per cent of CO, within the unfertilized egg is sufficient to lower the
alkalinity to such a degree that oxidation cannot proceed with suffi-
cient rapidity to allow development.

Loeb *° finds that an oxygen vacuum or KCN reduces the toxicity
of certain poisons to unfertilized and to a greater extent to fertilized
eggs (poisons that affect fertilized in less concentration than unfer-
tilized eggs). He concludes that this cannot be explained on the per-
meability hypothesis (as the mere absence of oxygen would probably
not affect the permeability?). It may be that these poisons are toxic
because they increase the permeability of both fertilized and unfer-
tilized eggs, but since the fertilized eggs are more permeable before the
action of the poison, the additional increase in permeability is fatal
because oxidation is abnormally increased. Hence KCN or an oxygen
vacuum would be antitoxic.

88 Littie: This journal, 1910, xxvi, p. 106.
*° HoEBer: Zeitschrift fiir Allgemeine Physiologie, 1910, x, B, p. 178.
*0 LoEeB: Biochemische Zeitschrift, 1910, xxvi, p. 288.

On the Dynamics of Cell Division. 275

Loeb *! finds that after membrane formation a saccharose solution
of much lower osmotic pressure will cause the egg to develop than a
pure NaCl solution, which, he says, proves that the membrane is per-
meable to salts or ions, for the explanation requires that the egg salts
diffuse into the sugar in the former case or the NaCl diffuses into the
egg in the latter. This shows a certain degree of permeability of the
egg to electrolytes after ‘‘membrane formation,” but proves noth-
ing as regards the normal unfertilized egg. It is not necessary to
postulate absolute impermeability, even of the unfertilized egg, to
electrolytes, in order to account for development by increased perme-
ability, and I have shown that an increase in permeability follows
the action of membrane-forming substances. Furthermore it has not
been demonstrated to my satisfaction that the action of the hyper-
tonic solution after membrane formation is purely osmotic.

Lyon has shown that the CO, and catalase elimination by the sea
urchin’s egg increases after fertilization, and Warburg, Mathews,
and Loeb and Wasteneys have shown that the oxygen absorption in-
creases. These changes might be due to an increase in permeability.

It is not supposed that an increase in permeability will cause any
cell to divide; growth is prerequisite to division. However, permea-
bility might influence growth. Growth is supposed to cause division
only when it affects the volume of the cytoplasm more than that of
the nucleus. The ratio of the cytoplasm to the nucleus in the egg
may be considered sufficient for a number of successive divisions, or
the “true”? cytoplasm may grow at the expense of the yolk after
each division.

“ LoresB: University of California publications, Physiology, 1908, iii, No. 11,
p. 81.

THE RELATION OF AFFERENT IMPULSES TO THE
VASOMOTOR CENTRES.

By W.-T; PORTER
(WitH THE CoLLABORATION OF R. RICHARDSON anp F. H. PRATT).1

[From the Laboratory of Comparative Physiology in the Harvard Medical School.]

HE successful observations in physiology — the observations
that largely influence opinion— are commonly attended by
parasitic hypotheses. Useful as these may be, they have in them-
selves no demonstrated truth, but derive their apparent credibility
from their associations. Bathed in the radiance of the Great Fact,
they seem to shine with their own light, and, as years pass, their
doubtful origin is almost or quite forgotten. |
It is known that the stimulation of the central end of many nerves
causes reflexly a rise or fall in the blood pressure, and that the central
mechanism for this purpose lies in the bulb near the calamus scrip-
torius, since the destruction of this region puts an end to vasomotor
reflexes. It is known, further, that the moderate constriction or tonus
of the blood vessels is maintained by impulses that stream continu-
ously from this same region, for the blood vessels dilate when the
nerves connecting them with the bulb are severed or when the vaso-
motor region is destroyed. This reflex and this tonus are fundamental
truths in the physiology of the circulation.

1 The stimulations of the sciatic nerve the data of which are used in this paper
and the original stimulations of the depressor nerve were made in 1908 (W. T.
PoRTER and R. RicHarpson: This journal, 1909, xxiii, p. xxxiv; W. T. PORTER
and F. H. Pratt: Jbid., p. xxxv). These studies were repeated by the writer some
months later, and the depressor curves used here are from this second investiga-
tion. The writer desires to thank Mr. RICHARDSON and Dr. Pratt, and to express
the hope that they will not be held responsible for the views set forth in the present

communication, for which he alone is answerable.
276

Relation of Afferent Impulses to Vasomotor Centres. 277

With these fundamental truths are associated several hypotheses.
Thus the bulbar cells are believed to transmute by an unknown
alchemy the afferent into efferent impulses. These same cells are
believed to produce the vasomotor tonus as a product of cell life. Not
a few physiologists contend that the tonus itself is a reflex, incited by
an unceasing flow of impulses from the periphery, but governed by
the personal activity of the bulbar centre. Under all these ideas lies
the primary hypothesis that both tonus and reflex are controlled
by the same masterful syndicate. The importance of this hypothesis
is obvious. For if the same apparatus control both tonus and reflexes,
the measurement of the vasomotor reflexes will reveal the condition
of the apparatus for the maintenance of arterial tone. In other words,
the mechanism for arterial tone cannot be impaired so long as the
vasomotor reflexes are normal.? The present communication brings
forward two methods by which this far-reaching assumption may be
tested.

The first method to be presented in this communication rests upon
interesting propositions. To begin with, it has long been known that
muscle will not change its form unless stimulated with a certain
intensity and speed. When this threshold stimulus is reached, the
muscle will shorten, and, as the successive stimuli increase in force,
the shortening is greater and greater until the muscle cannot shorten
more, however intense the stimulus. When the stimuli to the gastroc-
nemius are set down on an abscissa and the changes in the length of
the muscle are taken as ordinates, a characteristic curve is produced ®
(Fig. 1, G).

The tonus contractions of the heart muscle rise in a similar character-
istic curve as the stimulus increases from a minimal to a maximal
value (Fig. 1, H, and Fig. 2).

Since the heart is a modified blood vessel, it might be expected that

* It need hardly be pointed out that a weak or failing reflex does not neces-
sarily indicate depression in the bulbar cells; the difficulty may lie in the afferent
path or in the efferent path between the bulb and the blood vessel.

_ * Ticerstept, R., and A. WittHarD: Mittheilungen vom physiologischen
Laboratorium des Carolinischen medico-chirurgischen Instituts in Stockholm,
1883, i, Heft 3, pp. 1-20, Plate I, Fig. 1.

278 W. T. Porter.

the muscles of the blood vessels would also reply in this characteristic
form, and the present investigation will show that this is indeed the

Change
in Blood
Pressure
mm. Hg
or Per
cent.

0 500 1000 1500 . 2000
3000 4000 5000 6000 7000
60 70 80 90
Units of Stimulation 500 1000 1500 2000

Ficure 1.— Stimuli progressively increasing from threshold to maximal value cause
progressively increasing changes in the skeletal muscles, the tonus substance of the
heart, and the muscles of the blood vessels, and these changes are expressed quan-
titatively in curves of identical form. A S, the absolute rise, and P S, the percen-
tile rise, in blood pressure following the stimulation of the central end of the sciatic
nerve in the cat with Kronecker units from 250 to 2000. A D, the absolute, and
P D, the percentile, fall in blood pressure following the stimulation of the depressor
nerve in the rabbit (units 500 to 2000). H, therise in the tonus contractions of the
tortoise auricle stimulated directly with 3000 to 7000 units. G, Tigerstedt’s curve
of the contractions of the frog’s gastrocnemius stimulated indirectly through its
motor nerve (rheochord 60 to 100).

case* (Fig. 1). The natural load of the muscles of the blood vessels
is the blood pressure: when these muscles contract, the blood pressure

4 The fact that a strong stimulus produces a greater reflex rise of blood pres-
sure than a weak stimulus was first noted by Dirrmar: Arbeiten aus der phy-
siologischen Anstalt zu Leipzig, 1870, pp. 25 et seq.

Relation of Afferent Impulses to Vasomotor

rises; and when they relax, the blood
pressure falls. The extent of the rise
or fall in blood pressure is at least an
approximate index of the extent of
their contraction and relaxation.

The measurement of the contrac-
tions between the threshold and the
maximal stimulus is naturally much
simpler in the cardiac and skeletal
muscle than in the muscle of the blood
vessels. In the heart the stimulus is
applied directly to the muscle itself; in
the contractions expressed in Fig. 1, G,
the gastrocnemius muscle was stimu-
lated through its nerve; but in the
vasomotor reflex the stimulus to reach
the blood vessels must pass not only
through nerve fibres but through many
nerve cells. It has just been shown
that the form of the minimal maximal
curve is not distorted by the interven-
tion of nerve fibres. These act like the
indifferent conductors that they are.
But what of the nerve cells, which, in
the physiological tradition, have so
long had a prescriptive right to molest
the passing impulse? And, especially,

what of the bulbar cells, whose auto- |

cratic control of the vasomotor reflexes
was at first a wild surmise, then a
dignified hypothesis, and, latterly, a
dogma?

An answer to this question is not
altogether hopeless, as the following
reflections will show.

Stimuli desired to be of uniform

strength applied to nerves containing

vasomotor fibres by no means always
cause uniform changes in blood pres-
sure. The application of the stimu-
lus and the measurement of the

Centres.

The stimuli were the groups

The contraction is proportional to the intensity of the stimulus.

Tonus curves from the auricle of the tortoise.

of break currents indicated on the middle line.

FiGURE

279

(This journal, 1905, xv, p. 8.)

For the first stimulus the secondary coil was at 3000 of the Kronecker scale, for the second at 4000,
The lowest line records every fifth second.

and for the third at 5000. The fourth tonus curve was spontaneous.

280 W. T. Porter.

resulting changes in blood pressure are subject to a host of small
errors inseparable from any similar operation. But these errors are
“accidental.”

Statisticians are familiar with the doctrine of “accidental” and
“constant” errors. A marksman fires at a distant target on a calm
day. Small deviations over which he has no control and which are .
therefore accidental cause the bullets to strike above, below, to the
left, and to the right of the exact centre. If the number of shots be
large, the target could be painted out and the centre would still be
evident from the position of the bullet marks. They would be found
almost or quite symmetrically on all sides of the centre. Those to
the right would balance those to the left, and their distribution, if
plotted on coérdinates, would produce a symmetrical curve, the
well-known curve of error.

Far different would be the result if an unsuspected wind blew
across the line of fire. Then a constant error would operate. The
flying bullets would be carried down the wind toward one side of the
target. The curve of their distribution would be distorted.

A mathematical curve declares its origin; it does more than picture
its constituent data—it is those data self-revealed. Hence, if a
curve built up from individual observations of the same entity is
symmetrical, the errors involved are accidental; in other words, their
distribution is such that the errors on one side of the true value bal-
ance those on the other.® But if the curve is distorted or ‘‘skewed,”’
as statisticians say, a constant error is at work.

Accidental errors in measurement would not therefore affect the
form of the minimal maximal curve as an expression of the reflex
contractions of the blood vessels, provided only that the number of
observations be large enough to give the law of compensation free
play.® If, on the other hand, the bulbar cells have power to change
the vasomotor impulses, the exercise of this prerogative will operate
as a “‘constant’’ error. In short, if the vasomotor cells are conductors
merely, the form of the minimal maximal curve of blood pressures

® The number of observations must be large enough to make compensation
possible.

6 Evidently the number of individual observations necessary for satisfactory
compensation must be far greater in the case of the complicated vasomotor pro-
cedure than with the simpler muscle and nerve preparation.

Relation of Afferent Impulses to Vasomotor Centres. 281

obtained by stimulating the sciatic and depressor nerves will closely
correspond with the curve obtained from a muscle outside the body,
but if the vasomotor cells are the masters of the reflex, the curve will
not so correspond, but will show a disturbing force, just as the curves
at the rifle range showed that the wind blew on one day and not on
the next.

I propose therefore to measure the average change in blood pres-
sure on stimulation of the sciatic and depressor nerves by induction
currents increasing from the threshold to the maximal value, and
with these fixed points to determine the form of the curve that should
express the entire range between the minimal and maximal reflex.
But first, a few words must be given to the operative procedure.

The vasomotor reflex from stimulation of the sciatic nerve was
measured in the cat; for the depressor reflex, rabbits were used.

The cats were etherized, tracheotomized, and cannulas were placed
in the carotid artery and in the external jugular vein. The artery
was connected to a mercury manometer which recorded the blood
pressure upon a kymograph. Curare was injected into the vein. The
drug was dissolved in warm normal saline solution and the injection
was made very slowly. Artificial respiration was now begun. The
sciatic nerve was prepared. It was ligated and severed between knee
and pelvis. The peripheral portion was used in testing the effect of
the curare. The central portion was freed for a distance of about an
inch and three quarters. For stimulation the nerve was lifted into
the air and the electrodes were applied at least half an inch from the
ligature. The ligature was kept as dry as possible, and during stimu-
lation it was not allowed to touch the surrounding tissues. Between
stimulations the nerve was replaced, covered with the neighboring
tissues, and the outer wound closed with a pad of cotton wet in nor-
mal saline solution. The cotton did not touch the nerve. A very
small amount of bleeding was allowed at the bottom of the wound,
so that the nerve between stimulations might be bathed in an isotonic
solution. The greatest pains were taken not to stretch the nerve,
either during its preparation or afterward.

In the depressor observations curare was not employed; the can-
nula in the jugular vein was omitted and artificial respiration
was also avoided.

Extreme care must be given to the ether. The least inattention

282 W. T. Porter.

will impair the reflex. A few whiffs should be administered and the
ether should then be removed. Stimulation should not be made
during or very near the periods of ether inhalation.

A standard Kronecker inductorium was employed. It was grad-
uated in Kronecker units. The primary coil was supplied from a
Daniell cell of 1.1 voltage. The large smooth zinc was cleaned
and amalgamated after each experiment. Except during the experi-
ment itself the porous cup was kept constantly immersed in 20 per
cent sulphuric acid.

“1000

Tot tate fee be bet bine besten ofoatealie

10 A. M. 2p. M. 5.40 P. M,

FicurEe 3.— Original size. The fall in the carotid blood pressure of a rabbit on stimula-
tion of the depressor nerve with 1000 units at 10 A.m., 2 p.m., and 5.40 p.m. The
blood pressure fell 35, 41, and 42 mm. Hg respectively. To show that the method
can be safely employed for quantitative measurements. February 9, 1909.

A quantitative study of the vasomotor reflexes would have little
value if the circulation and the reflex nerve apparatus were im-
paired by the operative procedures or the length of the experiment.
It is essential that the physiological status be preserved. Fig. 3 is
evidence that this condition was satisfied in the present experiments.
In the rabbit from which the curves in Fig. 3 were taken, the first
reflex was recorded about ten o’clock in the morning. Observations
were repeated at frequent intervals for eight hours, but at six in the
evening the depressor reflex was still unimpaired and the observations
could probably have been continued some hours longer.

Since the correct form of the curve obtained depends on the accurate

Relation of Afferent Impulses to Vasomotor Centres.

placing of the points which
it connects, it is desirable
to inquire whether the
number of observations fix-
ing each point was large
enough to permit acci-
dental errors to be com-
pensated. It may at once
be said that the number
of observations in this in-
vestigation, though fairly
large for the usual research,
is regrettably small for
statistical purposes. In-
deed, one reason for pub-
lishing these results at
this time is to urge the
collection of such data by
many investigators.
Nevertheless, the present
results appear to be sub-
stantially accurate. The
following illustration will
bersoh “mterest, Im - the
experiments of January 21,
January 30, February 2,
and February g, the depres-
sor nerve was stimulated
eighty times with induc-
tion currents of a strength
between 400 and 799 Kro-
necker units. The average
absolute fall in the blood
pressure was 24 mm. Hg,
around which central value
the individual observa-
tions were symmetrically
grouped.

mm.Hg

The stimulation of the depressor nerve with 100, 200, 400, 600, 800, and 1000 units causes

original size.

Ficure 4.— The

the carotid blood pressure of the rabbit to fall, respectively, 11, 17, 20, 25, 28, and 31mm. Hg as registered by a membrane

Experiment of February 9, 1909,

manometer.

284 Wek Porcer.

The blood pressure fell not more than . 14 mm. Hg - 5 times.
The blood pressure fell between. 15 andig “ “ 14 “
The blood pressure fell between . 20 “ 24 “ “ 24 “
The blood pressure fell between. 25 “ 29 “ © Sagas
The blood pressure fell between. 30 “ 34 “ “ x2 “
The blood pressure fell more than . . . 35 “ “ ay

It is also desirable to inquire regarding the individual experiments.
The averages here offered might be supposed a mere statistical
accident. It is not to be denied that the average, or arithmetical

TABLE I.

Tur PERCENTILE FALL IN BLOOD PRESSURE ON STIMULATION OF THE DEPRESSOR NERVE
WITH INDUCTION CURRENTS INCREASING FROM THE THRESHOLD TO THE MAXIMAL
VALUE.

Kronecker Number of
Ree, Jan. 21. | Jan. 30. Feb..2. Feb. 9. |ohservations.| “Verase:

per cent. per cent. per cent, per cent. per cent.

15.3 14.0 15.3 13

19.7

mean, gives no information as to the character and distribution of
the individual data. The average height of buildings is very likely the
‘same in Berlin and New York. In New York this average conceals
the presence, and in Berlin it conceals the absence, of the tallest
buildings in the world. Examples of the relation of individual ob-

Relation of Afferent Impulses to Vasomotor Centres. 285

servations to the averages in this investigation are shown in Table I,
in which are presented the depressor reflexes in four animals. As
might be expected, not all these rabbits give exactly the same data,
but the divergence strengthens rather than weakens the use made of
them in this research.

Ficure 5.— The original size. The stimulation of the sciatic nerve
with 250, 500, and 1000 units causes the carotid blood pressure
of the cat to rise, respectively, 12, 26 and 54mm. Hg. Experi-
ments of March, 1908.

With these facts in mind the reader is invited to return to Fig. 1,
in which are placed the absolute and percentile changes in blood pres-
sure following the stimulation of the sciatic and depressor nerves.
Examples of the observations upon which these curves of Fig. 1 were
constructed are given in Figs. 4 and 5. It is clear that the minimal
maximal curve is the same in the tonus muscle of the heart, the
gastrocnemius stimulated through the sciatic nerve, and in the reflex
constriction and dilatation of the blood vessels. If the central nervous
system autocratically alters the impulse, the resulting contractions
in muscles under its control should differ from the contractions ob-
tained from muscles not under its control.

ETE.

The relation of afferent impulses to the vasomotor cells is certainly
a subject of great difficulty. However clear the results of the method
just presented may appear, the prudent investigator will wish to sup-
port them by a second method; and he will be glad if he can make
this second method totally unlike the first. This welcome contrast
has been secured in the procedure now to be described.

The problem is whether the vasomotor cells arbitrarily control the
vasomotor impulse. If they do so act upon the impulse, their action
should differ as their irritability differs. An impulse arriving in cells
that are in a state of augmented irritability should emerge with a force
greater than that imparted to the same impulse by cells in a state

286 Wot Porter:

of lowered irritability.
But to make this method
practicable, it is neces-
sary to secure wide
variations in irrita-
bility. The necessary
condition is obtained in
the Traube-Hering phe-
nomenon. In Figs. 6,
and 7 are shown Traube-
Hering waves of about
twelve seconds’ duration
and an amplitude of
about 40 mm. Hg. They
were observed in dogs
that had been given 5 c.c.
of a 3 percent solution of
morphine and in whoma
small quantity of curare
had been injected into
the crural vein. Artifi-
cial respiration was of
course employed.

The central end of the
sciatic nerve was stimu-
lated with submaximal
induction currents of
uniform intensity in the
trough and on the crest
of the waves and at vari-
ous points between. If
the vasomotor cells con-
trol the reflex, the rise
in blood pressure should
have been greater when
the stimulus passed
through the cells at the
height of the wave. As
Figs. 6 and 7 testify,
this was not the case.

A submaximal stimulus applied to the sciatic

As in the other figures, the lowest tracing records every fifth second and the signal magnet marks the stimulus and also writes the

The carotid blood pressure, recorded by a membrane manometer.
nerve at the crest of the Traube-Hering wave causes no greater rise than one applied in the trough or half-way down the slope of the wave.

From an experiment upon a dog, December 23, 1909.

atmospheric pressure.

FicurE 6. — Original size.

put

Relation of Afferent Impulses to Vasomotor Centres. 287

EV.

When two wholly different trains of thought bear the investigator to
the same point, the fact of arrival is not to be disputed, at least not
by the investigator himself. When the scientist espouses a method, it
should be for better or worse. Nevertheless, the writer reserves his

Hh,

FicurE 7.— Original size. A submaximal stimulus applied to the sciatic nerve on the
downward slope of the Traube-Hering wave causes no greater rise in carotid blood
pressure than one applied in the trough of the wave. The difference between the
blood pressure at trough and crest is about 40 mm. Hg. There are about five waves
per minute. From an experiment upon a dog, December 9, 1909.

judgment. In the present instance the problems involved are very
complicated, and they are moreover of grave importance in practical
medicine. The hypothesis to which this investigation directs atten-
tion has long ruled a state of thought and as a reigning theory is
ex officio entitled to allegiance until a successor is enthroned. I would
not hastily cast aside so old and so important a tool as the control
of the vasomotor reflexes by the bulbar cells. It is enough if this
investigation shall unmistakably mark the speculative character of
that dictum and the need of further research to justify its present
acceptance.’

7 It is perhaps advisable to point out that the two methods here described may
be usefully applied to other problems in physiology.

fee, PHYSIOLOGY OF CELL DIVISION. —TIL. THE ACTION
OF CALCIUM SALTS IN PREVENTING THE INITIATION
OF CELL DIVISION IN UNFERTILIZED EGGS THROUGH
ISOTONIC SOLUTIONS OF SODIUM SALTS.

By R. S. LILLIE.

[From the Marine Biological Laboratory, Woods Hole, and the Physiological Laboratory,
Department of Zotlogy, University of Pennsylvania]

Le a recent paper! I described the results of experiments in which
the unfertilized eggs of Asterias forbesii and Arbacia punctulata
were exposed for brief periods to pure isotonic solutions of neutral
salts of alkali metals and were then returned to sea water. Salts of
strong acids with monovalent anions were used, so chosen with the
aim of avoiding the possible influence of variations in hydrolytic dis-
sociation, valence of the anion, or calcium-precipitating power, while
giving a series whose successive members should exhibit increasing
action in promoting colloidal dispersion, namely, chloride, bromide,
chlorate,” nitrate, iodide, sulphocyanate. It was found that mem-
brane formation and parthenogenetic development could be induced
in both eggs by this treatment, but that the two species showed a
decided difference in readiness of response: thus with Asterias the
more weakly acting salts, chloride and bromide, produced a large
proportion of membranes and cleavages, while with Arbacia these
salts were almost ineffective, and only those with strongly acting
anions, iodide and sulphocyanate and to a less degree nitrate, showed
well-marked action. JI have since found that chlorate resembles
chloride and bromide, though somewhat more active than either,
producing membranes readily in starfish eggs, but in only a small
proportion of sea urchin eggs.
I suggested that this contrast in responsiveness to the above

_ treatment is to be correlated with the more general physiological con-

1 LitwiE: This journal, rg1o, xxvi, p. 106.

2 Used for the first time last summer (1910).
289

290 RAS. Lathe,

trast shown by these eggs; parthenogenetic development is much
more readily induced in Asterias than in Arbacia, and the former eggs
are more easily injured, and hence less uniform in their behavior,
and die sooner after maturation unless fertilized. On the view that
the critical change in fertilization, as well as in many forms of toxic
action and in natural death, is an increase in the permeability of the
plasma membrane — only temporary and reversible in the former case
— the contrast indicates the existence of a thinner or less resistant
plasma membrane in the starfish than in the sea urchin egg. With
the latter eggs, having a relatively resistant membrane, only the more
active salts are effective in producing the initial rapid increase of
permeability with consequent membrane formation followed by
cleavage; and the comparative efficacy of the salts in the above series
shows an unmistakable parallelism with their comparative power (a
function of the nature of the anion) of increasing colloidal dispersion.
In Asterias, however, with an easily modified plasma membrane, all
of the salts, including chloride, have well-marked and rapid action.

The composition and physical consistency of the plasma membrane
are thus to be regarded as important factors in determining what
response a particular species of egg will make to treatment by a given
parthenogenetic method. It is to be expected that any egg exhibiting
unusual powers of resistance to unfavorable conditions, or living rel-
_atively long in the unfertilized condition between its maturation and
its natural death — peculiarities indicating a plasma membrane of
more than average impermeability — will require relatively vigorous
treatment to call forth the desired response. This may well be the
reason why methods that are perfectly effective with sea urchin eggs
produce no impression on (e. g.) frogs’ eggs. On such an hypothesis,
at least, the choice of an appropriate method for a given egg ceases
to be a purely empirical matter. Such considerations may suggest
appropriate forms of treatment for eggs that have hitherto proved
refractory.

The experiments about to be described were performed at Woods
Hole during the past summer and form a continuation of those de-
scribed in the preceding paper. The relation of the increase in per-
meability induced by the pure salt solution to the initiation of cell
division has been again the chief question under consideration. I
have therefore compared the action of the pure salt solution with

The Physiology of Cell Division. 291

that of the same solution plus a small quantity of calcium or magne-
sium chloride. The addition of this substance prevents or retards
the increase in permeability resulting from the action of the pure
salt solution; this, there is good reason to believe, is the basis of the
well-known ‘‘antitoxic”’ action of salts.’ If the initiation of cleavage
is due to a similar increase in permeability, the addition of any salt
with typical antitoxic action should prevent this result. This has been
found to be the case. Experiments were conducted with both Asterias
and Arbacia. In the present paper the conditions in Asterias will be
more particularly considered. The experiments with Arbacia will
be described in more detail in a later paper.

I shall first consider the nature of the changes produced by pro-
longed exposure to the pure salt solution. Unfertilized mature eggs
of Asterias washed in sea water and then transferred to a pure isotonic
solution of one of the above salts of (e. g.) sodium undergo a progres-
sive series of changes of a very definite kind. Within a few minutes
after transfer to the pure salt solution (e. g., 55 m. NaBr) the eggs
show a decided tendency to form stringy masses or to agglutinate;
the protoplasm appears under the microscope at first unaltered, that
is, clear and translucent, but gradually becomes coarser and darker,
and within a certain time, varying with the salt, temperature, and
condition of the egg, it acquires an opaque and coagulated appear-
ance. The time required for complete coagulation is variable; but
typically at normal summer temperature all the eggs in a pure .55 m.
NaCl solution are completely opaque and dead within three hours;
the onset of death coagulation is more rapid in .55 m. NaBr and NaClO;
solutions and most rapid in .55 NaI and NaCNS. Ii left still longer
in the solutions, the coagulated eggs swell — most slowly in .55 m.
NaCl and most rapidly in .55 m. NaCNS and .55 m. Nal — and
eventually disintegrate. The above-described changes are of the
nature of cytolysis; the coagulation which precedes the swelling is,
however, highly distinctive. It invariably takes place both when the
egys are killed by poisons (as saponin) and when they die naturally,
_ as unfertilized mature starfish eggs left in sea water usually do within
eighteen hours after removal from the animal. Loeb has shown that

3 Cf. my experiments on Arenicola larve, this journal, 1900, xxiv, pp. 14, 450;
cf. especially pp. 24, 25. Other kinds of antitoxic action have probably an essen-
tially similar basis.

292 Re So Lilhie:

in the latter case the change is independent of bacterial action or other
external injury. It is, in brief, a natural process in which oxidations
are fundamentally concerned, since eggs kept in sea water freed from
oxygen or containing cyanide may live some days without dying or
coagulating.* One effect of normal or parthenogenetic fertilization
is thus to prolong the life of the egg. The above results show that the
coagulative change is greatly expedited by the action of pure salt
solutions. To what extent oxidative processes are concerned in this
salt action is not yet known; the addition of a little calclum or mag-
nesium chloride to the pure solution checks both the agglutinative
and coagulative action of the pure solution, so that the essential change
appears to consist simply in a marked increase of permeability. In
Arbacia eggs, which contain a red pigment, the increase of permeability
in the pure solution is rendered more directly evident by the outward
diffusion of this substance. A coagulative change also occurs in these
eggs under the above conditions, but more slowly;° both changes are
checked by the addition of a little calcium.

It should be noted that this change, coagulation at death or under
the influence of toxic or other unfavorable action, is not peculiar to
egg cells, but is shown by all forms of protoplasm; it is, in fact, a
general characteristic of living cells.6 It is associated with an in-

4 J. Lors: Archiv fiir die gesammte Physiologie, 1902, xciii, p. 50.

* Lors (Biochemische Zeitschrift, 1906, ii, p. 81) finds that in Strongylocen-
trotus, a very resistant egg as shown by its living unfertilized for forty-eight
hours in pure isotonic NaCl solution, the toxicity of pure or slightly alkaline salt
solutions is greatly diminished by removal of oxygen or addition of cyanide. He
concludes that increased or misdirected oxidations are a main source of the toxic
action of salt solutions on these eggs. The conditions in Asterias suggest rather,
in my opinion, that oxidations are in themselves a direct cause of rapid increase
in the permeability of the plasma membranes, and that the injury is to be as-
cribed primarily to this latter change. Of course, if the rate of oxidations is de-
pendent on the permeability of the plasma membrane, as I have maintained
elsewhere, increase in permeability may in itself result indirectly in an injurious
increase of misdirected oxidations; but the latter change would be a secondary
one, like the increased oxidations in the early stages of death rigor. There is
undoubtedly an intimate connection between the rate of oxidations and the con-
dition of the plasma membrane (cf. Warsurc: Zeitschrift fiir physiologische ~
Chemie, 1910, lxvi, p. 305).

° Cf. my paper, this journal, 1908, xxii, p. 75, especially pp. 81-82. I have
discussed briefly the essential nature of the connection between increase of sur-
face permeability and death coagulation in this paper.

Sy a oe caine amos

The Physiology of Cell Division. 2093

crease of surface permeability; the diffusion of substances into and
out of cells that have undergone death coagulation occurs with great
readiness; pigment-containing cells, like sea urchin eggs, lose pig-
ment, and in all cases the protoplasm becomes readily penetrable
to dyes and to other substances formerly unable to pass the surface
film (thus plasmolysis is no longer possible). The increase of per-
meability is undoubtedly the primary change, and coagulation of the
protoplasmic colloids is one of its secondary consequences. The asso-
ciation of protoplasmic coagulation with the death process is thus
simply another indication that the latter is associated with an increase
of surface permeability sufficient to destroy the normal or physiological
semi-permeability. Since such a change destroys the normal struc-
ture of the cell by altering the physiological state of aggregation of its
colloids, and profoundly alters its chemical constitution by allowing
free diffusion of crystalloidal compounds through the plasma mem-
brane, the properties of the cell as a living system are soon irrecoverably
lost.

To explain this connection between increase of surface permeability
and coagulation of the protoplasmic colloids, I advanced some time
ago the following hypothesis, which it seems desirable to restate here
in somewhat amplified form. The clear translucent appearance of
living protoplasm indicates an extremely fine dispersion of the cyto-
plasmic colloids. The electrical charge of the colloidal particles is
undoubtedly negative; and the fineness of the subdivision indicates
the existence of a considerable potential difference across the surface
of the particles, corresponding to the low surface tension which such
fine subdivision implies. It is assumed that this condition is promoted
by the electrical polarization of the plasma membrane which allows
certain cations (supposedly H-ions) to pass, but not anions. The
influence of the cations on the colloids within the membrane is thus
reduced, while that of the anions remains unaltered. Now, if the
potential difference across the interface of each colloid particle with

the medium be determined (1) by the specific ion-liberating properties
of the colloid and (2) by the difference between the respective ad-
sorptions” (or other form of combination) of anions and cations in

7 For evidence that the hypothetical ion-protein compounds of LorB and
PauLi are of the nature of adsorption complexes, cf. PauLi: Beitriéige zur chemi-
schen Physiologie und Pathologie, 1908, xi, p. 415.

204. R.S., Lathe:

the medium, it is clear that conditions tending to diminish adsorption
of cations will increase the negative charge on the colloid. These con-
ditions exist in the cell because of the ability of certain cations to pene-
trate the plasma membrane, producing the characteristic polarized
condition of the plasma membrane, whose outer surface appears to
be about one tenth of a volt more positive than its inner. The conse-
quence of this will be that the portion of the adsorption potential §
at the surface of the colloidal particles due to adsorption of cations
by the particles will be correspondingly decreased. The negative
charge on the colloid will thus be increased by the physiological
polarization of the plasma membrane, and a low surface tension (pos-
sibly negative in value) of the colloid particles, with consequent fine
subdivision, will result. This is the condition in the resting cell during
life. Decrease in the polarization of the plasma membrane following
increased permeability will increase the adsorption of cations by the
colloidal particles, decrease their negative charge, and hence increase
their surface tension.® Coagulation will result if this increase in sur-
face tension is sufficient and lasts sufficiently long.

The immediate action of the pure salt solution and presumably of
other membrane-forming agencies on unfertilized eggs is thus to
produce an increase in the permeability of the plasma membrane.
Such a change in a system like the living cell involves far-reaching
consequences. Increased rapidity of interchange across the plasma
membrane will follow, and probably an increased rate of oxidations.
It is known that cells when dying evolve carbon dioxide in increased
quantity: thus the carbon dioxide output of muscle increases mark-
edly at the onset of rigor; the increase in oxidation following fertiliza-
tion '° is thus, in all probability, to be referred, at least partly, to the
increase in permeability. On the present view these and other changes
in the egg system, conditional on temporarily increased permeability,
initiate, in some manner as yet mainly obscure, the complex series of

8 The potential due to unequal adsorption of anions and cations at the surface.
Cf. Micuaéris: Dynamik der Oberflachen, Dresden, 1909.
® Stimulation of a contractile tissue probably involves this kind of change,
the surface tension of the contractile elements undergoing increase with increase
in the permeability of the plasma membrane.
10 Lyon: Science, 1904, N. S. xix, p. 350; WARBuRG: Zeitschrift fiir physio-
logische Chemie, 1908, lvii, p. 1.

The Physiology of Cell Division. 205

chemical and physical changes of which cleavage and development
are the normal expression.

It should be emphasized that any marked increase of permeability
under physiological conditions can be only temporary and not too
prolonged, otherwise injurious alteration of cell organization must
follow. It has, in fact, usually been found to be essential that the eggs
should be exposed to the membrane-forming conditions for only a
brief period, a few minutes at most, and should then be returned to
normal sea water. In the case of sea urchin eggs, subsequent treat-
ment with hypertonic sea water, cold, or lack of oxygen greatly in-
creases the proportion of developing eggs; and we may infer from this,
in conformity with the above hypothesis, that the essential effect of
such after-treatment is to restore the normal permeability. The whole
procedure would then produce, first, a well-marked and rapid increase
of permeability; this condition would then be partly reversed by
transfer to sea water and completely by further treatment with hy-
pertonic sea water or cyanide. Presumably a rhythm of increased
followed by decreased permeability is thus started; such an alternate
irfcrease and decrease of permeability is, on the present view, anessential
condition of cell division."

If an initial rapid increase in permeability is the first critical change
in the initiation of cleavage, it is clear that any conditions that check
or inhibit this process should also prevent fertilization, whether
normal or artificial. The addition of small quantities of calcium
chloride to pure sodium salt solutions has a marked action in prevent-
ing the increase of permeability normally induced by the pure solu-

1 Tt seems probable from Lorp’s researches with ELDER (Chemische Entwick-
lungserregung, p. 249) that in normal fertilization the superficial action of the
spermatozoon, which apparently increases permeability, must be followed by a
second action, exercised after the sperm has penetrated the egg, and which re-
stores the latter to a state favorable for normal development. I make the sug-
gestion, in conformity with the above, that this second action consists essentially
in restoring the normal permeability. The spermatozoon thus first by its contact
increases, and then by its penetration decreases, the permeability of the egg.
This is probably not by any means the whole matter; but the continued life of
the fertilized egg demonstrates that its permeability is normal, while the fact
' that most eggs die soon unless penetration occurs indicates that abnormal per-
meability persists under these conditions. This hypothesis should be susceptible
of experimental confirmation or disproof.

296 Rf). sale:

tion. I have found that it also checks or prevents membrane forma-
tion and initiation of cell division by the above salts.”

The addition of small quantities of calcium and potassium salts to
pure solutions of sodium salts has long been known, since the researches
of Sidney Ringer, to diminish the toxicity of such solutions. The
researches of J. Loeb and his pupils have since shown that a large
proportion of plurivalent cations have this ‘‘antitoxic’’ power — so
called by Loeb to draw attention to the resemblance to the toxin-
antitoxin relation. Now the toxic action of pure solution of various
sodium and potassium salts on sea urchin eggs runs parallel with their
laking or permeability-increasing action, and is probably directly
dependent on the latter, increase of permeability beyond a critical
degree being in itself destructive to the cell’ This increase in per-
meability, as shown in sea urchin eggs by the diffusion of the char-
acteristic red pigment into the solution, may be checked or prevented
by the addition of salts of calcium or other appropriate metal to the
solution. The following experiment will illustrate. Equal quantities
of Arbacia eggs were placed in a series of twelve test tubes to the
depth of ca. one-half inch in each; the sea water was removed as far
as possible and the following solutions were added: (A) pure .55 m.
NaCl, NaBr, NaClO3, NaNO;, NaI, NaCNS, and (B) the same series
plus 1 c.c. m/2 CaCl, to each 20 of solution. Within five minutes
the eggs in the pure solutions .55 m. Nal and .55 m. NaCNS showed
exit of pigment; .55 m. NaNO; showed perceptible extraction within
ten minutes, and the others not till considerably later. But all the
calcium-containing solutions remained perfectly colorless for many
hours; next day, after eighteen hours, all were colorless except those
of iodide and sulphocyanate, the most toxic salts. The eggs were
then transferred to sea water and sperm added; of those left in the
pure solutions none developed, while those from the calcium-containing
solutions gave larve in all cases — though few from the iodide solu-
tion — except the sulphocyanate. Antitoxic action and prevention

12 NEWMAN, working with MAtTuHews, found calcium salts particularly effective
in preventing the normal fertilization of Fundulus eggs by spermatozoa. The
effect was reversible. This action is similar to the above; the calcium salt pre-
vents the initial increase of permeability which the sperm or the salt would other-
wise produce. Cf. NEWMAN: Biological bulletin, 1905, ix, p. 378.

18 R. Lituie: This journal, rg10, xxvi, pp. 116 ef seq.

The Physiology of Cell Division. 207

of increase of permeability thus run parallel; the antitoxic action is
less complete with the more toxic salts iodide and sulphocyanate than
with the others. In all cases, however, the increase of permeability,
with the associated toxic action, is greatly checked.“

Calcium similarly antagonizes the membrane-producing and
cleavage-initiating action of the pure solution. In the following series
of experiments the eggs were allowed to mature and were then exposed
for brief periods — five and ten minutes — (a) to the pure solution
and (6) to the same solution plus a small quantity of m/2 CaCh;
they were then returned to sea water. The result has always been
that eggs thus exposed temporarily to the pure solution showed a large
proportion of membranes and form changes and developed in a con-
siderable proportion of cases to blastule; while those similarly ex-
posed to the calcium-containing solutions remained — with a few
exceptions, especially with iodide and sulphocyanate solutions —
essentially unaltered. The presence of calcium, in other words, pre-
vents the initiation of cell division through the action of the salt.

*In each series of experiments some eggs were allowed to remain in
the pure solution and in the calcium-containing solution; the former
always underwent rapid agglutination followed by coagulation, while

4 The action of calcium salts in checking hemolysis by various toxic sub-
stances (saponin, digitalin, quillain) is an instance of the same kind (cf. J. B.
MacCattum: University of California publications, 1905, ii, pp. 87, 93). So
long as a cell retains the normal impermeability of its plasma membrane to diffu-
sible substances, the latter, if toxic, can have little injurious action. It seems,
however, to be a general characteristic of toxic substances, of whatever nature,
to increase the permeability of cells. Their action is thus twofold: first, they
break down the barrier normally existing between the cell and its environment,
a change which, besides destroying the osmotic equilibrium, permits abnormal
loss of diffusible constituents from the protoplasm and entrance of abnormal
substances from outside; and, second, they may then enter the cell and exert a
destructive action on the protoplasm itself. The presence of any substance which
opposes this increase of permeability has thus necessarily an antitoxic action.
In former papers I attempted thus to explain the antitoxic action of salts; cf.
This journal, 1909, xxiv, pp. 14, 459; such an explanation, if valid for salts, is
almost undoubtedly valid for antitoxic action in the more general sense. Anti-
toxins would thus protect cells against the permeability-increasing action of toxins.
This seems unquestionably true in the case of hemolysins and anti-hemolysins.
If agglutination is also a consequence of increased permeability, as I have else-

where maintained (Loc. cit., p. 21), the anti-bodies act by preventing increase of
permeability also in this case.

298 R. S. Lillie.

the latter remained normal in behavior and in appearance, with clear
protoplasm and capable of fertilization for several hours, showing,
in fact, little difference from eggs kept in normal sea water. Eventually
such eggs die and coagulate, as do also eggs in normal sea water, but
the contrast to those left in the pure solution is always marked, espe-
cially with the less toxic salts, chloride, bromide, nitrate, and chlorate;
in the case of iodide and sulphocyanate the toxic action is less com-
pletely checked by the calcium.

The series on page 299 is typical.

A marked contrast is thus shown between the effects of the pure
and of the calcium-containing solutions, the latter having little or
no action in increasing the proportion of parthenogenetically develop-
ing eggs. It should be added that the eggs used in this series were
not entirelynormal and showed rather more than the average tendency
to spontaneous membrane formation and change of form; this ten-
dency was undoubtedly spontaneous, since (1) the animals were all
thoroughly washed in fresh water before being opened, and (2) what
cleavage appeared was much slower and less regular than in the fer-
tilized control, and no eggs developed to the blastula stage. It should
also be noted that the calcium-containing solutions are not absolutely
indifferent in their action on starfish eggs; they usually show some
slight action in increasing the proportion of spontaneously cleaving
eggs; this is especially likely to be the case with the more strongly
acting salts sulphocyanate and iodide; in the above series the calcium-.
containing nitrate solution showed noticeably greater action of this
kind than the other calcium-containing solutions. In view of the
susceptibility of starfish eggs to the action of membrane-forming
agencies this is not surprising, since a solution of sodium chloride con-
taining calcium is far from being precisely equivalent to sea water.

® This tendency to spontaneous membrane formation and form change is
very frequently seen in starfish eggs that have lain some hours in sea water. It
is not to be regarded as evidence of the accidental presence of spermatozoa, since
the succeeding cleavage is imperfect and development rarely goes far, though
occasionally blastule are formed. The starfish egg is on the verge of partheno-
genesis, and cases of natural parthenogenesis are known (GREEF and others).
Eggs exhibiting such spontaneous development are always few and often absent.
It is evident, however, that in parthenogenetic experiments with starfish com-
parison of the experimentally treated eggs with untreated controls is especially
necessary. Such spontaneous development is never seen in sea urchin eggs.

_— |

The Physiology of Cell Division. 299

TABLE I.

June 18, 1910. — The starfish used were washed in tap-water to destroy any adhering
spermatozoa, and the eggs from several animals were used. Most of these eggs under-
went apparently normal maturation. A few after some hours showed membranes and
irregular form change.’ The eggs were exposed to the following solutions for five and
ten minutes in each case with the following results:

Solution. Time of exposure and result.
A. .55 m. NaCl. ieee Good proportion of membranes. A few larve.
_ 2. 10m. Considerable number of larve (< A1).
Aca. 55m. NaCl (250 c.c.). 3. 5m. Little change in three and one-half hours.

+ m/2 CaCl» (12.5 c.c.). Next day a few membranes; all dead and

~ coagulated. No larve.

Like A;. (One feeble abnormal larva found.)

5m. Fair proportion of membranes and irregular

forms after three and one-half hours. A
few larve next day.

2.-10m. Like B,, but larve more numerous and
more active.

i
=
oO
}

B. .55 m. NaBr.

Bca. .55 m. NaBr @50iceya Ss: ‘5m. A few membranes, imperfectly separated;
-+- Ca€l], (12.5 c.c. m/2). practically like unfertilized control. No
larvee.
4, 10m. Similar to Bs, but one larva found.
C. .55 m. NaClOs. 1. 5m. After three and one-half hours a somewhat

small proportion of membranes and
irregular forms. A fair proportion of
larve next day.

2. 10m. More favorable than (C,; considerable
number of larve.

Cca. 55 m. NaClO; (250 c.c.).3. 5m. Little change in three and one-half hours; a

+ CaCl, (12.5 c.c. m/2). few membranes and irregular forms. No
larve next day.
4. 10m. Like C,. Very few form membranes. No
larve.

D. .55 m. NaNO3. 1. 5m. Membrane formation in a somewhat small
proportion. A fair proportion of larve,
irregular and feeble.

2. 10m. More favorable than D,; _ considerable
number of larve, many active and vigorous.

D ca. .55 m. NaNO; (250 c.c.). 5m. Little change, but a few form membranes

Gaels (12:5 ce: m/2)- and cleave; one or two feeble distorted
larvee found.
4. 10m. Practically like D,; one or two feeble larve.

Control of unfertilized eggs. — All are dead and coagulated next day. Almost all of the
coagulated eggs are compact and without separated membranes; a few, however,
eer widely separated membranes; such eggs show more complete disintegration.

o larve.

Control of sperm-fertilized eggs. — Membrane formation is defective in many eggs and
the majority die before reaching a larval stage. Many form larve, largely irregular
or thick-walled.

Eggs remaining in the pure solutions (A, B, C, D). — After three hours all are coagulated
and opaque. A few have membranes, most not. The tendency of the eggs to cohere
or agglutinate in the pure solution increases in the order NaCl < NaBr < NaClO;
< NaNO; (condition after seven minutes in the solutions). The eggs were returned
to sea water after three hours and sperm was added; none developed.

_ Eggs remaining in the Ca-containing solutions (A ca, etc.). — After three hours all have

clear normal-looking protoplasm as in the unfertilized control. The tendency to
agglutinate is absent, there being only slight coherence as in normal sea water. On
transfer to sea water and fertilization with sperm after three hours in the solutions,
each lot yields a considerable number of blastule.

300 it. Se alia:

The further addition of magnesium and potassium salts, as in van’t
Hoff’s solution, is necessary to produce a practically indifferent
medium. The contrast between the action of the calcium-containing
solution and that of the pure salt is, however, invariably a striking
one. .

Two exact repetitions of the above series gave the same result.
Experiments with sodium and potassium iodides and sodium sulpho-
cyanate showed a less decided difference between the pure and the
calcium-containing solutions. The cytolytic and coagulative action
of the pure solution is markedly diminished by adding calcium, but
the prevention of membrane formation and development is less com-
plete with these salts than with those having less strongly acting
anions; with sodium sulphocyanate, in fact, comparatively little
difference was found between the pure and the calcium-containing
solution.! }

In one series magnesium chloride, as well as calcium chloride, was
used to offset the permeability-increasing action of sodium chloride.
Eggs were exposed for five and ten minutes as above to (1) pure .55 m.
NaCl, (2) a mixture of 250 c.c. .55 m. NaCl and 15 c.c. m/2 CaCls, and
(3) 250c.c..55 NaCl plus 18 c.c. m/2 MgCls. Magnesium showed the
same action as calcium, but was somewhat less effective in preventing
cytolysis; in correspondence with this difference more eggs were
found to form membranes and to cleave after treatment with the
magnesium-containing than with the calcium-containing solution.
Only a small proportion of eggs, however, were thus affected as com-
pared with those treated by the pure solution which formed mem-
branes on the majority of eggs. Doubtless salts of other bivalent
metals would be found to act similarly; experiments with these have
not yet been tried.

Nature of the membrane-forming action. — The separation of a sharply
defined thin film or membrane from the surface of the egg immediately
after fertilization is highly characteristic of echinoderm eggs but not
of eggs in general; it must therefore be regarded as dependent on special
conditions peculiar to these and a few other eggs (certain Mollusca,
Amphioxus). What appears to be fundamental and universal in the

16 This is in agreement with the general experience as to the relative difficulty

of counteracting the action of these salts by the addition of others. Cf. e. g. my —

results with cilia, this journal, 1906, xvii, pp. 104 ef seq.

SL le Sl ee Eg

The Physiology of Cell Division. 301

fertilization process is a temporary initial increase in surface per-
meability, and the separation of the surface film in certain eggs is, I
believe, to be regarded asalargely incidental consequence of this primary
change. Since the volume of fluid enclosed by the membrane — which
at first is indistinguishable from the general cell surface — increases
immediately after fertilization, an increase in the effective osmotic
pressure of the cell contents against the membrane is indicated. This
points to a disturbance of osmotic equilibrium as the essential con-
dition; the observed effect would result from an increase in the per-
EADity of a detachable surface film sufficient to allow ready passage
of salts but insufficient for the passage of the more complex diffusible
substances and colloids in the surface layers of the protoplasm; under
these conditions the outwardly directed osmotic pressure of the latter
substances would no longer be equilibrated by the salts of the ex-
ternal medium and the volume of fluid enclosed by the membrane
* would increase. This is what actually happens.

It is characteristic of echinoderm eggs that the surface of the egg
beneath the detached membrane remains sharply defined, while the
volume of the egg remains practically unaltered. The space between
the egg surface and the separated membrane is occupied by a clear
fluid which apparently contains colloidal material derived from the
egg.’ Loeb has suggested that on fertilization a colloidal substance
is set free in the superficial layer of protoplasm by a cytolytic action,
and absorbs water or swells, pushing a modified portion of the sur-
face film away from the egg surface. I believe that in its main features
this explanation is essentially correct, but suggest further that what
conditions this swelling is simply an increase in the permeability of
the surface film, so that the osmotic pressure or ‘‘Quellungsdruck” 18

17 Herest: Biologisches Centralblatt, 1893, xiii, p. 14; J. Lors: Archiv fiir
Entwicklungsmechanik, 1908, xxvi, p. 82.

18 Tt is not clear to me that any fundamental distinction has been established
between ‘these two phenomena. As I have already urged (This journal, 1907,
XX, p. 127; cf. pp. 133, 140), the absorption of water by a liquid colloidal system
separated from the solvent by a membrane —a process usually ascribed to
Osmotic pressure — appears to be essentially the same phenomenon as the ab-
sorption of water by a solid colloid — in this case called swelling. Pavii and

' Hanpowsky have maintained that the osmotic pressure of protein solutions is
a phenomenon of swelling or hydration different from true osmotic pressure
(Biochemische Zeitschrift, 1909, xviii, p. 340); and Hanpowsky states that there

302 Ro Ss bale.

of this colloid substance (which need not necessarily be set free aé
the time of fertilization) is no longer equilibrated by that of the ex-
ternal salts. The film is then separated from the egg surface as the
colloid absorbs water.!? This view implies the essential identity of
the fertilization membrane with the plasma membrane. So far as can
be distinguished by observation with the water-immersion lens, the
most external layer of the unfertilized egg of Arbacia and the fertiliza-
tion membrane are identical in thickness and optical properties. They
are probably therefore composed of the same material. The separated
membrane, however, has different osmotic and probably other proper-
ties from the surface film of the egg; thus it is demonstrably freely
permeable to salts, though impermeable to colloids like serum albumin
or egg albumin” and difficultly permeable to sugar; *! whereas the
plasma membrane of the unfertilized egg is practically impermeable
to salts. This is shown by the fact that the whole egg including the
visible surface film shrinks when placed in hypertonic sea water; if
this film — which is afterwards, on the present view, separated as the
fertilization membrane — were, like the latter, freely permeable to
salts, the egg ought te shrink away from the membrane.” The separa-
tion of the two cannot, however, be accomplished until after normal
or parthenogenetic fertilization; while immediately after this event

is no corresponding depression of the freezing point in such solutions (Kolloid-
Zeitschrift, 1910, vii, p. 193). This appears to me thermodynamically impossible;
a solution exhibiting a “‘swelling pressure,’’ as well as one showing ‘‘true osmotic
pressure,’ must have a lower vapor tension than the pure solvent in proportion
to the height to which the pressure can raise the level of the solution above that
of the pure solvent. If its vapor tension is lower, its freezing point, 7. e., the
temperature at which contiguous solid and liquid phases are in equilibrium, must
also be proportionately lower. The explanation of the failure to detect any de-
pression of the freezing point in protein solutions exerting an osmotic or swelling
pressure simply lies, in my opinion, in the fact that such determinations become
uncertain when the osmotic pressure is very low. )

19 ““A small concentration of sugar or protein, even though its osmotic pressure
were far less than that of sea water, would be capable of absorbing sea water
through a membrane perfectly permeable to sea water.”” E. N. Harvey, Journal
of experimental zoélogy, 1910, viii, p. 363.

20 ‘Loren: Loc.ch,

21 HARVEY: Loc. cit.

22 Unless, indeed, there were close adhesion between the two — a supposition
which seems inconsistent with the ease of separation of the membrane.

the film separates from the egg surface and proves itself, on examina-
tion, to be freely permeable to salts but impermeable to colloids. The
facts thus seem clearly to indicate that the surface film of unfertilized
eggs, like plasma membranes in general, is semi-permeable in relation
to the salts of the medium, and loses this semi-permeability temporarily
in consequence of fertilization; it is then separated from the surface
as a result of absorption of water by the superficial colloid substance
while it is yet impermeable to the latter. This altered surface film
(which probably represents only the most external layer of the true
plasma membrane} since it is unaffected by lipoid solvents) is the
fertilization membrane. A new plasma membrane with a surface
film similar to the original is normally soon afterward 1 re-formed by
the protoplasm (see below, p. 305, footnote).
Further evidence that the egg surface becomes more permeable to
dissolved substances immediately after fertilization has during the
* past summer been brought forward by McClendon,” Harvey,* Lyon
and Shackell” and by Loeb” himself. These investigators agree in
finding that the rate of diffusion of substances into or out of eggs un-
dergoes marked increase shortly after fertilization. In harmony with
these observations I may cite another of apparently different nature
which I have frequently made with Asterias eggs: when treated
with isotonic sodium chloride for five minutes or warmed to 35° for
thirty seconds, not all of the mature eggs form membranes, and of
“those which do only a small proportion undergo favorable develop-
ment; the remainder, usually after undergoing irregular cleavage or
change of form, die and disintegrate. When the eggs are examined
after eighteen hours, a striking contrast is invariably shown between
the eggs with membranes and those without; the latter, while dead
and coagulated, are only slightly swollen and present a compact ap-

‘The Physiology of Cell Division. 303

%8 McCLENDON: Science, 1910, xxxii, pp. 122, 317.

4 HARVEY: Science, 1910, xxxii, p. 565.

* Lyon and SHACKELL: Science, 1910, xxxii, p. 2409.

6 J. Lors: Science, 1910, xxxii, p. 411. Lors had formerly (cf. Biochemische
Zeitschrift, 1906, ii, p. 87) considered the possibility that the greater toxicity of
pure NaCl solutions on the fertilized as compared with the unfertilized ege was
due to the greater ionic permeability of the former; in the paper cited he rejects
this explanation in favor of one based on differences in the oxidative metabolism
of the two kinds of eggs. There is, however, no conflict between these two
possibilities, since changes in permeability and in rate of oxidation appear to go
hand in hand.

304 KR. S: Tealhe.

pearance and definite outline; while the eggs with membranes are
always found to have undergone extensive disintegration and are
usually converted into masses of loose detritus filling the whole space
enclosed by the distended fertilization membranes.?’ It is plain
that in these eggs the resistance to swelling or disintegration is slight
compared to that of the eggs without membranes. This can only
mean that the surfaces of eggs with separated membranes have an
unusually high permeability; that an effective surface of separation
between egg and medium has in fact practically ceased to exist. While
the phenomena of dead eggs can be regarded as throwing only a partial
light on the conditions during life, the above contrast is so constant
and striking as to leave no doubt of a marked difference in permeability
between the two classes of eggs.

Eggs like those just described appear to have undergone the typical
coagulative disintegration or cytolysis as a result of the destruction
of the semi-permeable properties of their plasma membranes. Loss
of semi-permeability, if the present point of view is valid, is a neces-
sary accompaniment or consequence of the change leading to mem-
brane formation; hence such a change, if not reversed, must eventu-
ally lead to the dissolution of the egg protoplasm. This, I believe,
is why the majority of eggs (particularly sea urchin eggs) subjected
to a simple membrane-forming treatment die without development.
In order that favorable development shall follow, the initial increase |
of permeability must be succeeded, after an appropriate interval,
by a decrease, that is, by a return to or toward the original condition.
To produce this effect some further treatment is necessary. This in-
ference is a simple corollary of the membrane theory as applied to
the case of the dividing cell. For the concrete evidence that the
initial increase of permeability is followed by a decrease, I may point
to the observation of Lyon * that a temporary slight loss of pigment
takes place from Arbacia eggs during the period immediately following
fertilization; this outward diffusion of pigment then ceases, indicating
' decrease of permeability. Harvey also states that in Toxopneustes
eges “between ten and fifteen minutes after fertilization the eggs
return to the same condition of permeability, with respect to alkali,
as the unfertilized. There appears to be a second increase at the time

27 R. Lirzie: Journal of experimental zodlogy, 1908, v, p. 375; ¢f. P- 387:
28 Lyon: Loc. cit.

The Physiology of Cell Division. 305 -

of the first cleavage.’’? The evolution of carbon dioxide and the
susceptibility to poisons also follows a rhythm which, as I have already
pointed out, corresponds to the rhythm of changing permeability re-
quired by the present theory. We may therefore infer that in the nor-
mally dividing egg or other cell the permeability undergoes a series
of alterations which might be represented by a curve which would cor-’
respond in its general form to the curve of carbon-dioxide production.
Any artificial means of starting cleavage is presumably successful in
proportion to the faithfulness with which it produces a curve of per-
meability change corresponding to the normal.

I have already expressed the opinion that the parthenogenetic
methods act in this manner.*? The treatment, after the preliminary
membrane formation, with hypertonic sea water, cold, or cyanide,
according to the methods discovered by Loeb, produces, on this hy-
pothesis, a return to the normal semi-permeable condition of the
plasma membrane. Exactly how this result is accomplished it is
impossible to say at present.*! In the egg of Strongylocentrotus the
hypertonic solution fails to produce its effects in the absence of oxy-
gen; so that the physical action of the solution and some chemical
action of the nature of oxidation appear to combine in producing the
effect. How far these conditions are general remains to be determined.

29 Harvey: Loc. cit. 30 See above, p. 295.

31 Hypertonic sea water, by decreasing the volume of the egg, 7. e., concentrat-
ing its constituents, may favor the re-formation of a normal impermeable surface
film. Cold retards the chemical processes which further disintegration, and thus
gives the egg time to recover, and the same may be true of cyanide. Oxidations
are apparently favorable to increase of permeability in the unfertilized egg in
sea water; if they are checked the counter-changes are more likely to restore the
semi-permeable surface film. That a surfdce film of the same nature as in the .
unfertilized egg is actually re-formed after the separation of the fertilization
membrane, is shown by the fact that a second membrane, in all respects like the
first, may be produced by appropriate treatment (cf. Hergst, Biologisches Cen-
tralblatt, 1893, xiii, p. 14; TENNENT: Journal of experimental zodlogy, 1906, iii,
p. 538; Harvey, bid., 1910, viii, p. 363).

It seems to me highly interesting that hypertonic solutions have been used
With success in checking the hemolytic action of bacterial toxins in the intact
organism. This effect is comparable to the above and probably has a similar
basis. Cf. W. D. SutHERLAND: Biochemical journal, 1910, v, p. 1. This result
Suggests that the scientific therapeutics of the future will pay close attention to
the conditions by which the permeability of cells may be kept normal, or restored
to the normal after alteration.

306 FS eenite:

In the following experiments I have studied the action of cyanide
and of hypertonic sea water on eggs previously exposed to the mem-
brane-forming salt solution. These experiments, though incomplete,
have shown that after-treatment with hypertonic sea water increases
the proportion of eggs developing to a larval stage. The effect, how-
ever, in the experiments so far performed has been much less decided
with Asterias than with Arbacia. In the latter form eggs treated for
five minutes with isotonic sodium or potassium iodide or sulphocyanate
solutions and transferred after an interval of ten minutes * to hyper-
tonic sea water (250 c.c. sea water plus 15 c.c. 2.5 m. NaCl) for half
an hour yield a remarkably high proportion of active larve (from 50
to 80 per cent), many of which swim at the surface and appear fully
normal; the hypertonic sea water produces, in fact, the same result
as after membrane formation by fatty acid. Asterias eggs treated
with .55 m. NaCl solution for five minutes, followed, after a brief
interval, by exposure to hypertonic sea water for thirty minutes, also
yield a larger proportion of larve than eggs treated with the isotonic
salt solution alone; but the increase in my last summer’s experiments
was comparatively slight, and the great majority of eggs died in an
early stage of development. One significant difference was noted
between the eggs treated with hypertonic sea water and the others:
the degree of disintegration in the dead eggs after eighteen hours was
distinctly less in the treated than in the untreated eggs, indicating
that the increase in permeability had been checked by the hypertonic
sea water — an observation in conformity with the above hypothesis
that the hypertonic solution acts by decreasing the permeability. After-
treatment with cyanide —a method highly effective with Asterias
eggs in which membranes have been formed through brief warming —
was found to produce little or no increase in the proportion of develop-
ing Asterias eggs and only a slight increase with Arbacia. It would
thus appear that the injurious action of the pure salt solution is less

32 Ten minutes is the most favorable interval between return from the salt
solution to sea water and transfer to the hypertonic solution. Eggs transferred
to hypertonic sea water at twenty, thirty, and forty minutes after membrane
formation show a progressively smaller and smaller proportion of favorable
development.

%8 Apparently it makes little difference whether the membrane-forming solu-

tion has a general action on the colloids of the membrane, or affects specifically
the lipoids.

The Physiology of Cell Division. 307

easily reversed than that of temporary warming; also that after-
treatment with cyanide produces a much less close approach to the
normal conditions than after-treatment with hypertonic sea water.
I hope to continue these experiments next summer.

SUMMARY.

The chief experimental results and conclusions of the foregoing
paper may be briefly summarized as follows:

1. The addition of small quantities of calcium chloride to isotonic
solutions of sodium salts (1) prevents the rapid increase in permeability
produced in the unfertilized eggs of Asterias and Arbacia by the pure
solution, (2) produces at the same time a marked decrease in the
toxicity of the solution, and (3) prevents the membrane formation
and initiation of cell division which are typically induced by the pure
solution. The view is thus confirmed that both the toxic action of
the pure salt sélution and its action in initiating cell division are
due primarily to the production of a condition of increased surface
permeability.

2. This increased permeability is, however, temporary in normal
or in favorable parthenogenetic fertilization. Treatment with hyper-
tonic sea water after the formation of fertilization membranes by salt
solutions results in an increase in the proportion of favorably develop-
ing eggs, especially in the case of Arbacia, of which the majority of
the eggs thus treated may reach active larval stages. Since hypertonic
sea water thus prolongs the life of the egg — an effect comparable to
that of calcium in the above antitoxic action — and since prolonged
life implies a practically normal permeability, the inference is drawn
that the essential effect of such after-treatment is to bring the per-
meability — which has been increased by the initial membrane-
forming treatment — again to the normal. An artificially induced
increase is thus followed after a favorable interval by an artificially
induced decrease of permeability. Without such after-treatment few
eggs succeed in developing beyond an irregular early cleavage stage
and development is abnormal.

EFFECTS OF PRESSURE ON CONDUCTIVITY IN NERVE
AND MUSCLE.

By WALTER J. MEEK ann W. E. LEAPER.

[From the Physiological Laboratory of the University of Wisconsin.]

ECHANICAL stimulation of nerve and muscle as measured by
the resulting muscular contraction has been thoroughly in-
vestigated, but the effect of mechanical factors on conductivity in
these tissues has received scant attention. The literature consists of
few papers and these are full of conflicting statements. That nerve
compression might offer an opportunity for a clearer insight into the
phenomenon of conduction seems never to have been generally appre-
ciated. The purpose of the present work has been chiefly to devise
an instrument suitable for the application of pressure to muscle and
nerve, and to measure the amount of pressure necessary to block
conduction in these tissues. In addition the relations existing between
pressure, time, and strength of stimuli, velocity of conduction under
compression, irritability, and fatigue have also been studied.

The problem as outlined was suggested by Erlanger’s experiments
on artificial heart block. In his work the auriculo-ventricular bundle
was compressed with a special screw clamp. Since the bundle contains
both nerve and muscle tissue, it seemed advisable to investigate the
effects of compression on these tissues separately.

The first work on nerve compression was done by Fontana as early
as 1797. Hermann? quotes Fontana as having pointed out that pres-
sure on a nerve interferes with conduction without causing stimula-
tion. Interest in the subject seems next to have been reawakened by
the clinical methods of nerve stretching which were in vogue during
the middle part of last century. Weir Mitchell? determined roughly

1 HERMANN: Handbuch der Physiologie, 1870, ii, p. 95.
* ? Werr Mircueri: Injuries of nerves and their consequences, Philadelphia,
1872.
308

Pressure on Conductivity in Nerve and Muscle. 309

the amount of pressure necessary to block the nerve impulse. He
found that 18 to 20 inches of mercury acting for ten to thirty seconds
was sufficient to interrupt conduction. Schiff, Laferon, Heidenhain,
Tigerstedt, and Wundt all touched on the problem from time to time,
but the first paper dealing definitely with the question was that of
Liideritz.2 This worker used rabbits and attempted to determine
whether sensory or motor fibres were first affected by compression.
He found that sensory impulses passed through the compressed area
at a time when motor impulses were no longer able to reach the muscle.
Zederbaum * found that pressure on a frog’s sciatic nerve increased
its irritability and that sensory nerves were blocked before motor.
Efron® also found an increase in irritability and conductivity in the
compressed area. Calugareanu ® worked with smaller pressures than
the preceding authors. He found no increase in irritability. The
threshold value of a stimulus applied above a region of compression
was always increased.

Ducceschi’ in tgor published an extensive study of the effects of
compression on conduction in nerve. He determined the pressure
necessary to produce block, and showed that both reflex motor and
sensory impulses are interrupted before motor impulses from stimuli
applied directly above the point of compression. Ducceschi rarely
found an increase in irritability due to pressure. He found that
pressure acts differently on impulses produced by different kinds of
stimuli. With increasing degrees of compression impulses from
chemical stimuli were the first to be blocked, while those from mechani-
cal and electrical stimuli were next affected in the order named.
Muscle curves taken while the nerve was under increasing amounts
of pressure were found to vary only in height.

Bethe * in 1903 verified some of Ducceschi’s data. He frequently
observed an increase in irritability after pressure was applied. In an
attempt to prove that the neurofibrils are the conducting parts of the
nerve, Bethe directed his attention toward the histological effects of

3’ Ltperirz: Zeitschrift fiir klinische Medizin, 1881, ii, p. 97.

4 ZEDERBAUM: Archiv fiir Anatomie und Physiologie, 1883, p. 161.

5 Erron: Archiy fiir die gesammte Physiologie, 1885, xxxvi, p. 467.

® CALUGAREANU: Journal de physiologie et de pathologie, 1901, pp. 393, 413.

7 Duccescui: Archiv fiir die gesammte Physiologie, roo1, lxxxiii, p. 38.

8 BetHe: Allgemeine Anatomie und Physiologie des Nervensystems, 1903,
p. 248.

310 Walter J. Meek and W. E. Leaper.

compression. Semenoff * has also studied compression of nerve, but
his problem was to find whether pressure gave the same changes in
irritability as those described by Wedensky in the phenomenon of
parabiosis.

Many different methods of securing compression were used by the
above-mentioned workers. Liideritz bound the leg of the rabbit with
rubber bands and silk ligatures, including both bone and muscle with
the nerve. Zederbaum applied weights to a platform resting on a
piston which was fitted into a glass cylinder. The foot of the piston
was provided with a piece of hard rubber which rested on the nerve.
Efron used a lever of the second class with weights in a scale pan hung
at the end. Calugareanu employed a device similar to that of Zeder-
baum, the nerve in his experiments resting in a small groove into which
the piston fitted. Ducceschi devised a more refined method in that
the area compressed was greatly reduced. Two small holes were
drilled through a glass plate about .3 mm. apart, and through these
a thread was looped over the nerve. Below the plate the thread was
attached to a scale pan. The latter could be lowered gradually by
means of a screw. The area of compression was thus equal to the
diameter of the thread and the pressure could be applied slowly with-
out injury. :

It will be seen that none of the methods just described is physically
perfect since each involves deformation of tissue. For the most part
they are crushing methods. The actual pressure varies with the size
of the nerve and the friction to be overcome, and it is impossible to
say just how much pressure has been applied per unit of surface. The
point of blocking can never be an exact one as expressed in terms of
weight. From a physical point of view the only perfect way to measure
pressure is by fluid transmission. The first requisite for our work was
to find an apparatus that would transmit fluid pressure to nerve and
muscle.

It proved a task of considerable difficulty to devise an instrument
that would overcome all objectionable features of manipulation and
at the same time give accurate readings. The prosecution of the work
was finally made possible by the ingenious apparatus described be-
low, which was designed by Dr. Erlanger. We wish to express our
indebtedness to Dr. Erlanger for help throughout the entire work.

* SEMENOFF: Archiv fiir die gesammte Physiologie, 1903, c, p. 182.

Pressure on Conductivity in Nerve and Muscle. 311

Fig. 1 illustrates the compression apparatus. A large air tank A
was connected at H with a half-inch pipe leading first to a pressure
gauge and then toacylinder C. This cylinder, which contained about
a pint.of heavy machine oil, was connected by means of a valve V
with a brass tube R, which led to a brass T-tube B. The horizontal
arm of the brass T-tube measured
7 mm. in length and 2.5 mm. in
diameter. Through the arm of this
T-tube a short piece of a carotid
artery from a calf was drawn, re-
flected over the ends, and tied tightly
with a heavy linen thread. The
pressure system was thus closed.
The nerve or muscle to be studied
was then drawn through the lumen
of the blood vessel. On letting air
into the system from the com-
pressed air tank pressure was trans-
mitted by the oil directly to the
blood vessel and thus upon the nerve
or muscle drawn through its lumen.

The pressure gauge was tested and found accurate. Its registering
capacity was 200 pounds, and each division of the scale represented
5 pounds. By watching the gauge and opening valve T of the air tank
any desired pressure could be quickly secured. To remove the pres-
sure it was necessary to disconnect at H. By closing valve V in ad-
vance the pressure could be raised in the oil cylinder independent
of the T-tube. On opening the valve pressure could be admitted
almost instantaneously to the preparation of nerve or muscle in the
T-tube.

At first there was some difficulty in getting a suitable kind of blood
vessel. All the various vessels from the cat, rabbit, and dog were
tried with imperfect success. The carotid artery of the calf finally
proved satisfactory. Sections of some length can be found entirely
free from branches, and if drawn tightly through the tube this artery
will withstand over 100 pounds’ pressure. Usually there was a slow
leakage of oil through the artery if the pressures were high. Unless
the leakage became as much as a drop every three to four seconds the

Figure 1.— Compression apparatus.
Description found in text.

312 Walter J. Meek and W. E. Leaper.

reduction in pressure was considered negligible. The oil itself was
neutral to litmus and had no appreciable effect on nerve or muscle.

Tue AMOUNT OF PRESSURE NECESSARY TO BLOCK THE IMPULSE.

Our work was next directed toward determining the number of
pounds’ pressure necessary to block the impulse in nerve and muscle.
The technique was as follows: For the experiments on nerve a gas-
trocnemius nerve muscle preparation was used. The’ pressure tube
was placed in a moist chamber. The muscle was supported by a
muscle clamp, and the nerve drawn through the lumen of the artery,
which had previously been placed in the tube. The free end of the
nerve rested on platinum electrodes. For experiments on muscle the
sartorius was used. This muscle could be drawn through the artery
and tube in the same way as nerve. One end lay on electrodes and the
other was connected to a lever. In the case of tetanic stimuli no
recording device was employed, the disappearance of the contraction
being noted with the eye. In some experiments single break shocks
from a stimulus selector were used and the contraction recorded with
a muscle lever on a slow drum. In the majority of the experiments,
however, the muscle was connected to a lever which on contraction
of the muscle broke a circuit including an electric chronograph serv-
ing as a signal magnet. The moment of contraction was recorded on
the plate of a spring myograph. The myograph broke the primary
circuit of the induction coil in the usual way.

We were surprised in our first experiments to find the great re-
sistance offered by nerve and muscle to compression. Pressures as
high as 80 and go pounds did not at once block the impulse. Block-
ing, however, resulted after such pressures acted for some time. This
emphasized the time factor, and we took it into account by starting
with a given pressure, usually 40 pounds, and increasing it at the rate
of 5 pounds every half-minute until block appeared. Table I presents
the results of a series of seven experiments on the sciatic nerve.

In the first column are given the initial pressures. From 40 pounds
the pressure was raised at the rate of 5 pounds every half-minute until
block appeared at the number of pounds shown in the second column.
The time of compression in minutes is given in the third column. In

Pressure on Conductivity in Nerve and Muscle. 313

all the experiments, unless otherwise mentioned, a Verdin coil was
used, the secondary having 2000 turns of .7 mm. wire. The secondary
coil in the experiments on nerve was kept 320 mm. from the primary.
Maximal contractions were induced at 400-430 mm.; so the stimuli
were well above the maximal. The stimuli were produced by the

TABLE I.
PRESSURES ON NERVE.
Pressures in pounds.

Time of
compression.

t Initial. Final.

40 65
40 90

40 | 90

40 | 90
40 85
40 90

Average 40

knock-down key of the myograph and were uniform in strength. As
can be seen from the table, an initial pressure of 40 pounds can on the
average be increased to 85.7 pounds before the nerve impulse is blocked
by the compression.

In Table II similar data are given for the sartorius muscle. The
distance of the secondary coil from the primary was 200 mm., which
gave a stimulus somewhat above maximal. All our experiments have
shown that the average pressure necessary to block conduction in
muscle is somewhat lower than it is in nerve. The range of pressures
is somewhat wider, however, in the case of muscle. The greatest
pressure required to produce block was 100 pounds, while 55 pounds
was the lowest recorded. The average for all of our experiments on
muscle is 75 pounds. The seven records in the above table were sub-

214 Walter J. Meek and W. E. Leaper.

mitted because the muscles used were from the same frogs as the
nerves in Table II.

Experiments were next made on curarized muscle for comparison
with the non-curarized (Fig. 3).

TABLE II.
PRESSURES ON MUSCLE.
Pressures in pounds.

Time of
compression.

Initial. Final.

M
40 90 5
40 70 3
40 Ps)
40 5)

4
40 235
40 3

Average 40 : 3.8

Although the lowest pressure recorded, 55 pounds, is not below the
lowest in non-curarized muscle, it will be noted that in no case was
a pressure of more than 75 pounds required to block the impulse, and
the average is 12 pounds below that for normal muscle.

From these experiments it seems that muscle is less resistant to
pressure than nerve. This reduced resistance may, however, be due
in part to the effect of the drug. While curare picks out the nerve
endings specifically, it is by no means inert toward muscle. Possibly
the results are but an expression of this action. We had hoped to
find the difference between nerve and muscle in their reaction toward
compression striking enough to give some idea as to which tissue was
most affected in artificial heart block. The differences are not great
enough to justify any conclusions, at least not until the pressures
used in heart block are accurately known. It is sufficient to say

Pressure on Conductivity in Nerve and Muscle. 315

that in regard to pressure curarized muscle is less resistant than non-
curarized, and non-curarized is less resistant than nerve.
TABLE II.
PRESSURES ON CURARIZED MUSCLE.
Pressures in pounds.

Time of
compression.

Initial. Final.

40
40
40
40
40
40

Average 40

An extensive series of experiments was also made in which tetanic
stimuli were used to determine when block occurred. The same
general facts as stated above were found to apply here also. Nerve
proved the most resistant and curarized muscle least. In these ex-
periments one leg of the frog was ligated before the injection of curare,
and the sartorius of this leg was used as a control against the curarized
muscle of the opposite leg. In every case the non-curarized muscle
proved the more resistant.

From the first it was of course recognized that in determining the
amount of pressure necessary to block the impulse two other factors
were concerned — the length-of time the pressure was applied and the
strength of the stimulus. That a stronger stimulus increases the
amount of pressure or the time that a constant pressure must act in
order to produce blocking, must have been known to the previous
workers, but none seem to have emphasized the fact, and most writers
do not even note the stimulus used to determine the point of block.
To find the relation of the amount of pressure and strength of stimuli

316 Walter J. Meek and W. E. Leaper.

to the time required for block, certain arbitrary pressures and stimuli
were selected.

It is sufficient to give the summaries of these various series of
experiments. In twenty experiments on nerve under a constant
pressure of 40 pounds the average time required to block the nerve
impulse was 7.5 minutes. The longest time observed was 16 and
the shortest 2.5 minutes. Twelve of the observations varied between 6
and 11 minutes. In a second series of five experiments 20 pounds’
pressure was employed. The average time for block was 19 minutes.
From this data it appears that with constant stimuli the time neces-
sary to produce block.is roughly proportional to the pressure applied.

The same general fact was easily shown in muscle, both normal
and curarized. In ten experiments on normal muscle with the coil
at 200 mm. distance, 40 pounds’ pressure blocked the conduction on
the average in 4.4 minutes. The time required for blocking with 20
pounds’ pressure under the same conditions averaged 7 minutes. In
curarized muscle 40 pounds’ pressure blocked the impulse in 2.4
minutes, and 20 pounds’ pressure in about double that time.

In both nerve and muscle the time required to interrupt conduc-
tion is proportional to the strength of stimulus. To demonstrate this
we placed the nerve or muscle under a constant pressure, and when-
ever block appeared increased the strength of the stimuli until maxi-
mal contractions were again made by the recording muscle. One
experiment on nerve subjected to 70 pounds’ pressure gave the fol-
lowing results: maximal stimuli from the coil at distances of 320, 105,
$2, 58, and 32 mm. were blocked in 3, 13.5, 31, 56, and 113 minutes
respectively. The ability of the nerve impulse to pass through an
area of compression therefore varies with the strength of the stimulus.
The term “block” is really a relative one. It would be interesting to
know whether the relation between the strength of stimulus and the
pressure or time of compression is the same as that found by Greene 1°
between strength of stimulus and the action current.

The pressure required for block did not seem to vary with the size
of the muscle in any constant direction. A muscle 3 cm. long and
weighing 87 mg. required 70 pounds’ pressure for block in a given
time, which was exactly the same amount required by a muscle 3.5 cm.
long and weighing 214 mg. No doubt the size of the muscle influences

10 GREENE: This journal, 1898, i, p. 104.

WR mb

Pressure on Conductivity in Nerve and Muscle. 317

the distribution of pressure to some degree, but it would not seem
considerable enough to be a factor in our experiments.

Doers COMPRESSION INCREASE THE IRRITABILITY OF NERVE?

Various results have been obtained by previous workers in attempt-
ing to answer this question. Efron and Zederbaum reported a con-
stant increase in irritability on com-
pression. Ducceschi found it but Mand mA
rarely, and Calugareanu not at all. | |
Increased irritability in this connec- | THA
tion has meant the power of the nerve a oe
to augment the strength of the im- (\ |
pulse as it passes through the area “mm aL arc ria
of compression. That this occurs FICURE 2.— Tracing showing increase

, in height of gastrocnemius after ap-
Efron and Zederbaum showed by Poe : a
plication of 20 pounds’ pressure to

readings from the position of the sclatic nerve. First arrow marks
secondary coil. Zederbaum found that point of compression. Second arrow
increase ivtability begins to appear M=Ms Amal of prea an
when a weight of 75 gm.is placed on ynaximal

the nerve, and steadily increases until

the weight reaches 500 gm. In his experiments with the nerve un-
weighted the stimulus just necessary to give a muscle twitch was
produced by the secondary coil at a distance of 308 mm. from the
primary, while with a weight of 500 gm. the coil gave the same
effect at a distance of 332 mm.

In our work we attempted to show this increase in Teeabalty as
follows: We set the secondary coil at a point where submaximal
stimuli were produced, and then by means of a stimulus selector sent
in submaximal break shocks at the rate of 1 every 2.5 seconds. The
muscular contractions were recorded in the usual way on a slow drum.
After a short series of contractions were recorded pressure was applied
to the nerve. Fig. 2 presents one of these tracings. In this case
the admission of 20 pounds’ pressure to the nerve brought about some
change which resulted in an increased height of contraction. This
does not always occur, but we have found it very frequently with low
pressures. A pressure of 40 pounds usually blocks the impulse from
a submaximal stimulus at once. Pressures of from to to 25 pounds,

318 Walter J. Meek and W. E. Leaper.

on the other hand, frequently show the phenomenon described
above.

Ducceschi was the first to point out that we are not here dealing
with a real increase in irritability. He looks upon the impulse as merely
being augmented by others set up in the compressed area. Increased
irritability in the strict sense would mean some change in the lability
of the protoplasm so that it responds more readily or completely to
its stimulus. There seems to be no reason for believing that the nerve
is rendered more irritable in this sense. More likely it is a matter of
summation in the muscle. Compression itself is a form of stimulation.
This is seen in certain cases by the production of an incomplete tetanus
when the pressure is applied rapidly. It may be assumed that im-
pulses are constantly being received by the muscle from the compressed
area of the nerve. These impulses are summated with those from the
single shock, and an increase in amplitude of the contraction results.
In the same manner compression might result in the summation of
sensory impulses. This is probably the explanation of hypereesthesia
in cases where a tumor or other growth has exerted pressure on a
nerve trunk.

FATIGUE EFFECTS.

Recovery occurred both in nerve and muscle after compression
provided the pressure was not too long maintained. Time records
were not kept of all experiments, but from a large number of obser-
vations it may be said that in general nerve recovered from block
under a constant pressure of 40 pounds in from three to thirty minutes.
In those experiments in which the pressure had been raised rapidly
recovery was also rapid and at times almost instantaneous. The
longer the pressure was applied the slower was the return of function
after the removal of compression. Muscle recovered more rapidly
than nerve, and curarized muscle more rapidly than non-curarized.
These findings were in line with the amount of pressure and length
of time required to produce block in the same tissues.

The recovery, however, was not a complete one. Conduction never
became entirely normal, as could be shown in two ways. In the first
place, a second block could be produced with a much smaller pressure
than at the beginning of the experiment. Forty pounds applied a

Pressure on Conductivity in Nerve and Muscle. 319

second time almost invariably suspended conduction at once. If the
block was not immediate, the time necessary to produce it was much
reduced.

In the second place fatigue effects could be produced such as those
recently described by Tait ™ after subjecting the nerve to cold. This
author found that, after nerve had been frozen and allowed to recover,

eee

ol OUT BA ag ct We Nf eS fe jp We ee

Ficure 3.— Tracing showing fatigue effects of pressure on nerve. Harvard coil. Be-
ginning with the first arrow a series of tetanic stimuli was sent in above the com-
pressed area. At the second arrow, by means of a commutator, the stimuli were
sent in between the muscle and the compressed portion.

on stimulation a series of tetani could be run, each lower than the last
until the nerve was completely fatigued. Fig. 3 illustrates this
phenomenon in our experiments. The nerve was compressed for
thirty minutes with a pressure of 4o pounds. The nerve was then
removed from the apparatus, placed in physiological saline and al-
lowed to recover for two hours. The muscle was then connected to a
muscle lever and the nerve laid over two electrodes, one peripheral
and one central to the compressed portion. A series of short tetani
was then sent in above the compressed area. As can be seen in the
tracing, incomplete tetanic curves were produced, the last stimulation
from the central electrode showing almost complete fatigue. That
this was not due to fatigue of the muscle was shown by the complete
tetanus from the same stimuli applied nearer the muscle than the com-
pressed area. Fatigue of nerve can thus be demonstrated after com-
pression quite as well as after freezing.

1 Tair: Quarterly journal of experimental physiology, 1908, i, p. 79.

320 Walter J. Meek and W. E. Leaper.

Is THE VELOCITY OF THE NERVE IMPULSE INFLUENCED BY
COMPRESSION?

The idea that pressure might slow the nerve impulse was suggested
by the increased A-V interval which is found in artificial heart block.
We may say at once that we do not'have a satisfactory answer to this
question, because we were unable to attack it in the most feasible way,

Stimulation

EET
11.35.00 ————
11.25.00 4 : See

|
a
=
—}
10.37.00 —
=

res.
40 lbs. 2
- Normal.

FicurE 4.— Tracing showing increase in latent period and time of conduction after the
application of 40 pounds’ pressure to nerve. Pressure admitted at 10.3514. Block
in 514 minutes. Recovery some time between 12 and 35 minutes.

namely, by recording the wave of negative variation before and after
its passage through the compressed area. We did find, however, that
the interval between stimulation and contraction increased from the
time pressure was applied until block made its appearance. This
interval is the latent period of the muscle plus the time of conduction
through the nerve. Fig. 4 shows the gradual lengthening of this
interval when the nerve is subjected to pressure. The latent period
was recorded on the spring myograph by an electric chronograph, the
current to which was broken by the contraction of the muscle. The
normal interval between stimulation and contraction as shown by
the eight lower curves is about 1.25 seconds. This increases after
40 pounds’ pressure is admitted to almost 3 hundredths before block
appears. Ducceschi”™ states that there is no change in the muscular
curve other than a reduction in amplitude, but we have found this

12 Duccescut: Loc. cit.

Pressure on Conductivity in Nerve and Muscle. 321

increased period between stimulation and contraction constantly,
both in nerve and muscle.

This increased delay is probably due in most part to a longer latent
period of the muscle as a result of a decrease in the strength of the
impulses which affect it. We failed to lengthen the interval by doub-
ling the area of nerve compressed, which seems to show that the nerve
itself is not greatly concerned. By gradually reducing the strength of
stimuli to a nerve or muscle the same increase in latent period may
be shown. It thus seems evident that compression is a most delicate
way of reducing the strength of the nerve impulse gradually and
slowly. As the impulse weakens, the latent period of muscle lengthens,
and conduction jtself becomes slower in the nerve. Tigerstedt ® has
shown that the latent period of muscle may increase from .007 to .o2
second as the impulse diminishes. Piper™ found a slightly lower
velocity of conduction in nerve for impulses from stimuli near the
threshold value: These results do not eliminate the possibility of
changes in velocity in the compressed area itself, but a study with
the string galvanometer would be necessary for their determination.

SUMMARY.

t. A method has been described which makes possible the fluid
transmission of pressure to nerve and muscle.

2. The average time required to block the nerve impulse in the
frog’s sciatic by means of a constant pressure of 40 pounds was found
to be 7.5 minutes. Under the same conditions 4.4 minutes were re-
quired to establish block in fresh sartorius muscles and 2.4 minutes in
curarized. In both nerve and muscle the time necessary to interrupt
conduction was proportional to the pressure applied and the strength
of the stimulus.

3. Immediately after the application of pressures not exceeding 20
pounds to nerve, stimuli of a given strength frequently caused an
increased height of contraction in the recording muscle. There was
probably a summation in the muscle of impulses set up in the com-
pressed area with those due to the electrical stimulus.

% TicERstEpT: Archiv fiir Anatomie und Physiologie, Sup. 1885, p. 165.
™ Piper: Archiv fiir die gesammte Physiologie, 1909, cxxvii, p. 480.

322 Walter J. Meek and W. E. Leaper.

4. Fatigue of the nerve was demonstrated after recovery from
compression.

5. No change in velocity of conduction in the compressed region
has as yet been detected. The muscular contraction, however, was
reduced in height, and the interval between stimulation and contrac-
tion prolonged as block made its appearance.

oe EEPECTS OF STRETCHING THE NERVE ON THE
RATE OF CONDUCTION OF THE NERVOUS IMPULSE.

By A. J. CARLSON.
[From the Hull Physiological Laboratory of the University of Chicago.]

si years ago, the author, in collaboration with Dr. Jenkins, reported

some observations on the pedal nerves of the slug (Ariolimax),
which seemed to show that when these nerves are stretched within
their physiological limit there is an actual extension of the conducting
substance in the nerve, and a delay of conduction proportional to the
degree of extension, the actual rate of conduction of the impulse thus
remaining the same in the two conditions of the nerve.! These facts
seemed to us “evidence on the side of the view that the conducting
substance in this nerve is in a liquid condition, or at least in a semi-
liquid condition.”’ The following year experiments on a marine worm
(Bispira) were reported.? These confirmed our results on the slug to
the extent that stretching the worm increased the conduction time
in the same length of nerve cord, as compared to the conduction time
in the relaxed or unextended worm. Because of the impossibility of
accurate measurements of the length of the nerve cord in the worm
when not extended, I could not determine whether or not the delay in
conduction of the impulse was proportional to the degree of extension
of the nerve cord, that is, whether the stretching within the physio-
logical limit actually affected the conduction rate. Estimates were
made on this point that indicated the same conditions as were found
in the pedal nerve of the slug, but the uncertain element is the length
of the nerve cord in the non-extended worms, and for that reason little
value was attached to those estimates. The very obvious conclusion
was intimated that “‘the results do not point to any substance in the

1 JENKINS and CaRLsoN: Journal of comparative neurology and psychology,
1904, xlv, p. 85.
2 Carson: This journal, 1905, xiii, p. 351.
323

324 AL) Carmi son:

nerve as being concerned in the impulse, except that in case the neuro-
plasm is of a more fluid consistency than the neurofibrille, they speak
in favor of the former as the substance involved.”

Three years later Bethe * reported some experiments on the leech
(Hirudo), which seemed to question our results on the slug and the
marine worm. Contrary to our results, Bethe finds in the leech that
stretching the nerve cord within the physiological limit does not affect
the conduction time; and he concludes, therefore, that in this stretch-
ing there is no actual extension of the conducting substance, but only
a straightening out of the kinks of the neurofibrillz. These results of
Bethe and his attempt to account for our previous findings on the
basis of experimental errors are the occasion for the present note. It
was planned to carry specimens of Ariolimax to the recent physio-
logical congress in Vienna for actual demonstration, as there seems no
better way to settle a disputed point as to facts. The animals arrived
from California in good condition, but they did not seem to do well
in our laboratory in Chicago, and those brought along succumbed
on the way to Vienna.

The problem resolves itself into two essential questions, namely,
(1) Does the stretching of the nerve within the physiological limit alter
the conduction time? By “physiological limit’? we understand the
degree of stretching to which the nerve or nerve cord may be sub-
jected in normal activity and which does not alter the excitability of
the nerve or the intensity of the nervous impulse. (2) Is there any
relation between the degree of stretching of the nerve and the amount
of change in the conduction time?

As regards the first point, a recent repetition of my previous experi-
ments has confirmed the results then reported. I wish to call atten-
tion to a few typical tracings (Figs. 1-4) secured last August on ani-
mals kindly sent me by Dr. Jenkins of Stanford University. The
technique is essentially the same that was employed in earlier work.
The stimulating electrode remains fixed at the same point of the pedal
nerves near the pedal ganglia. The tracings are left in the same con-
dition as when taken off the kymograph. The difference in the latent
time is apparent on direct inspection, but those interested may draw
the requisite lines and measure the actual delay of the conduction

* BetHe: Archiv fiir die gesammte Physiologie, 1908, cxxii, p. I.

The Effects of Stretching the Nerve. 325

through the stretched nerve. The time signal is the same on all the
tracings, namely, 50 d.v. per second.

The records in Figs. 1 and 2 are practically perfect. The distance
of the muscle lever from the stimulation signal shows that the tonus

ee ee nce
__ Se ies ae

Ficure 1. — Four sevenths the original size. Tracings from the foot musculature of the
slug (Ariolimax) on stimulation of the pedal nerves at a fixed point near the pedal
ganglia. A, nerves relaxed; B, nerves stretched. Delay of conduction in the stretched
nerve.

of the reacting musculature (posterior end of the foot) is the same at
the time of stimulation in each parallel series, and the nearly uniform
amplitude of the contractions shows that the intensity of the nervous
impulses reaching the reacting muscle in the two conditions of the

Ficure 2. — Four sevenths the original size. Tracings from the foot musculature of the
slug (Ariolimax) on stimulation of the pedal nerves at a fixed point near the pedal
ganglia. A, C, nerves relaxed; B, nerves stretched. Delay of conduction in the
stretched nerve.

nerve is practically the same. The records in Fig. 3 are less satisfac-
tory, because the amplitude of the contraction is somewhat greater
with the nerve in the stretched condition, while the tracings in Fig. 4
are typical of poor results, in other words, rejected series. The delay
in the conduction time in the stretched condition is obvious, to be sure,
but the tonus of the muscle and the amplitude of the contraction are too

326 A. J. Carlson.

variable. It is needless to say that in reaching the conclusions in our
earlier work, only records like Figs. 1, 2, and 3 were admitted, as is
shown by the sample tracings then published.

What are the sources of error to be guarded against in these experi-
ments, and are the results possibly due to such errors as have been
overlooked? The variable degree of tonus of the foot musculature is a
troublesome factor. The excitability, and hence the latent period, of

Anny RIAA VINA AAAI LASS AA PRA PALS LID IT STIS TOIT TI IS,

—_—_——
Annu

IADR AIA DRI DALIAN IAAI AALS

Ficure 3.— One half the original size. Tracings from the foot musculature of the slug
(Ariolimax) on stimulation of the pedal nerves at a fixed point near the pedal ganglia.
A, C, relaxed nerve; B, stretched nerve. Delay in conduction in the stretched nerve.

the slug muscle probably varies with the muscle tone, as is the case in
the vertebrates. Bethe assumes that we were ignorant of, or failed to
consider, this elementary fact in the physiology of nerve and muscle,
and so he imagines that we shifted the position of the preparation to
suit any variations in the muscle tonus, and that in the readjustment
we failed to place the writing point of the lever perpendicular to that
of the stimulating signal! Perhaps we erred in assuming that physi-
ologists would take these things for granted. Between preparation of
the animal and the taking of the first record, and between each suc-
ceeding record; time was allowed for the foot muscle to reach a cer-
tain uniform degree of relaxation, as is evidenced by the distance of
the muscle curve from that of the stimulating signal. That does not
mean a complete cessation of the tonus relaxation, but the further re-
laxation is so slow that the muscle curve drawn on the recording sur-
face, travelling at the speed necessary for these measurements, is
practically a straight line (see tracings). Hence there is no unusual
error in determining the beginning of the elevation of the muscle
lever.

The Effects of Stretching the Nerve. 227

The strength of the stimulus (break induction shock) was always
submaximal. Good preparations will yield successive submaximal
contractions of fairly uniform amplitude. If maximal or supermaxi-
mal stimuli had been used, there might have been in the stretched
nerve considerable variations in the excitability and the intensity of
the impulse, with the amplitude of the muscle contraction remaining
practically the same. With the muscle in similar condition of tonus

A =

nA ATP PADDLES DP PP IPOD LL LOD ERA ALL DILL IED LLLP LLP OPAPP EA

Ficure 4. — About one half the original size. Tracings from the foot musculature of the
slug (Ariolimax) on stimulation of the pedal nerves at a fixed point near the pedal
ganglia. A, C, relaxed nerve; B, D, stretched nerve. Typical curves of the rejected
series of experiments.

and degree of fatigue, and the intensity of the stimulus to the nerve
constant and submaximal, if the contractions are of fairly uniform
amplitude it seems safe to conclude that the excitability of the nerve
and the intensity of the nervous impulse also remain fairly constant.
Our criterion for the physiological limit of nerve stretching is, there-
fore, (1) the absence of effects on the intensity of the nervous impulse
and on the excitability; (2) absence of actual excitation (over dis-
tention of the nerve results in muscle contraction; (3) approxima-
tion of the degree of stretching to that attaining in the animal when
normally crawling. This latter is admittedly an approximation. I
do not think that I can come closer to it than about 2 cm.

One source of error was overlooked by us in the earlier work, as
well as by Bethe in his critical review of our data. In the slug prepa-
ration it is necessary to stretch the whole length of the nerve from the
farthest electrode point to which the central end of the nerve is tied.
In consequence, when the nerve is stretched, the portion of it between

328 A. J. Carson.

two points of the electrodes is also stretched, and in consequence the
point or region of actual excitation of the nerve is necessarily a little
farther from the muscle than with the nerve in the relaxed state. There
is no way of eliminating this error even when recognized. But al-
though the rate of propagation of the impulse in these nerves is only
30-40 cm. per second, the delay in the conduction in the stretched
nerve is too great to be accounted for by the slight shifting of the point
of stimulation that actually takes place when the points of the elec-
trodes are only 1 mm. apart.

Lastly, Bethe argues that the complexity of the slug nerve-muscle
preparation renders the results doubtful. Biedermann and Bethe have
shown that there is a peripheral ganglionic plexus in the foot muscu-
lature of many gasteropods. The pedal nerves evidently enter this
ganglionic plexus, and in some gasteropods stimulation of the pedal
nerves may give contraction or relaxation of the foot musculature, or
a combination of contraction and relaxation. I have observed this
both in Helix and in Limax, but never in Ariolimax, although I have
made no special study of it in this slug. I have always failed to obtain
relaxation of the foot musculature, even when in great tonus, by stim-
ulation of the pedal nerves with single induction shocks or the inter-
rupted current. And even in Limax the stimulation of the pedal nerves
invariably gives a definite and uniform contraction of the posterior
. end of the foot and the dorsum, provided these musculatures are re-
laxed at the time of the stimulation. There is, in all probability, in the
foot musculature of Ariolimax a ganglionic plexus similar to that in
other gasteropods, but I fail to see how this vitiates the results under
discussion. Such a mechanism might have made the latent periods of
successive stimulations of the pedal nerve as variable as the latent
time of many reflex reactions, and thus rendered this particular prepa-
ration unavailable for this work. But since this is not the case, we
need not concern ourselves with this ganglionic mechanism in the
present inquiry. As regards the first point, then, the data seem conclu-
sive. Extension of the nerve increases the conduction time without
altering the intensity of the impulse or the excitability of the nerve
fibres. Bethe admits that when the leech is stretched beyond the
physiological limit the conduction time is increased. It is not clear to
me whether Bethe holds that this delay is proportional to the degree of

The Effects of Stretching the Nerve. 329

stretching. Is there any relation between the degree of nerve stretch-
ing and the amount of increase in the conduction time? I have no new
data touching this point. Bethe endeavors to show that our previous
data are wide of the mark. As regards the data on the marine worm,
it was distinctly stated that they were only estimates. But the results
on the slug preparation seem more convincing. The sources of error
in the measurements were recognized. It is difficult to measure with
accuracy the length of the nerve in the relaxed condition. It is diffi-
cult to keep the electrodes nearest the muscle on the same point of the
nerve when alternately stretching and relaxing it. And a further
factor is the rapid slowing of the conduction in the nerve, whether
stretched or not. Under these conditions individual variations are
obviously inevitable. And we made it a special point to- give three
typical experiments in detail, one exceptionally good (Table III) and
two fair (Tables I and II). I fail to see how any of the sources of error
involved could work constantly in one direction. The average is,
therefore, more nearly correct than any single pair of measurements.
The average figures in our sixteen acceptable experiments, being 44
pairs of records on the stretched nerve, and 49 pairs of records on the
relaxed nerve, are:

Length of nerve. Rate of conduction.
Biretchede sels 8.15 cm. 34.6 cm. per second.
eae Sereda. Ss AEA ey kak gee gia ee

The inference seems obvious that the delay in the conduction is pro-
portional to the degree of nerve extension, so that the actual rate of
conduction remains practically constant.’

Attention may now be invited to Bethe’s own results, particularly
to the published tracings and to the experimental methods. When
an attempt was made to extend the observations on the slug prepara-
tions to the worm phylum, I tried all available species of annelids,
including the leech (Aulastomum) and the earthworm, but I finally
settled on the marine worm Bispira as the only workable species. In
my hands the leech proved the least suitable of all. And Bethe’s

4 “Erst bei Dehnung iiber die physiologische Linge tritt eine Verlangerung der
Uebertragungszeit ein. Die Geschwindigkeit der Reizleitung in einen gegebenen
Stiick Bauchmark ist also proportional seiner augenblicklichen Lange’ (p. 32).

> BETHE’S analysis (p. 8) of one of our experiments cited in detail (Table III,
p. 89) is erroneous.

330 A. J. Carlson.

own curves and figures show that the species used by him (Hirudo)
makes no exception. The anterior or reacting end of the leech is in
almost constant motion, so that it is nearly impossible to secure con-
secutive contractions of the same amplitude, or sufficient duration
of rest to obtain a horizontal line from the recording lever prior to the
contractions evoked by the stimulations. That being the case, the
determination of the exact point of rise of the lever, or the latent period,
becomes a matter of more or less guesswork, as is evident on Bethe’s
published tracings (Fig. 10, p. 25). Such tracings can, at best, be
used as the basis only for a rough estimate of the rate of conduction,
as was pointed out seven years ago by Dr. Jenkins and the writer.®

In the second place, the stimuli, both proximal and distal, are sent
through the entire worm by means of the pins fixing the preparation
to the apparatus. This renders it difficult to determine the actual
point or level of stimulation of the nerve cord. And lastly, the Jacquet
chronograph was used as time signal in most of the experiments, ren-
dering the measurements of even so large time elements as 0.01’ only
approximate estimations. It would seem, then, that Bethe’s data do
not even prove his contention for the leech, and much less can they be
used to refute the results on the slug and on Bispira, objects much
better suited to such experiments.

The interpretation of the phenomenon is still a matter of conjec-
ture, except in so far as it points to a fluid condition of the conducting
substance. I have failed in my attempt to extend it to the spinal
nerves of vertebrates, probably because these can be stretched only
within narrow limits before changes in excitability and actual stimula-
tion are produced. When the experimental difficulties are overcome,
the phenomena will probably be demonstrated in the visceral nerve
plexuses in the vertebrates. It probably also obtains as regards the
excitation wave in muscle cells subject to relatively great tonus varia-
tions. As we have seen, even Bethe admits that when the nerve cord
is extended beyond the physiological limit there is a delay of the
conduction, but his figures are not sufficiently correct to determine
whether the delay is directly proportional to the degree of stretching.

6 Jenkins and Caritson: Journal of comparative neurology and psychology,
1903, Xili, p. 265.

THE PRODUCTION OF GLYCOSURIA BY ADRENALIN
IN THYROIDECTOMIZED DOGS.

By FRANK P. UNDERHILL,

[From the Sheffield. Laboratory of Physiological Chemistry, Yale University,
New Haven, Connecticut.]

a failure Of adrenalin to induce glycosuria after removal of
the thyroids (the parathyroids being intact) reported by Ep-
pinger, Falta, and Rudinger ! was not corroborated by Underhill and
Hilditch.2 In the experiments of the latter adrenalin called forth the
appearance of sugar in the urine of all the thyroidectomized animals
tested.

In a reply by Falta and Rudinger® the following criticisms con-
cerning the work of Underhill and Hilditch are offered: (1) with a
single exception the latter authors injected larger quantities of adren-
alin than were employed by Eppinger, Falta, and Rudinger. The
positive outcome of the single exception is given the doubtful ex-
planation that some thyroid secretion might still be in circulation
such a short time (two days) after thyroidectomy. On the other
hand, when in the same animal adrenalin in the proper dosage called
forth glycosuria eighty-one days after removal of the thyroids, Falta
and Rudinger point out that ‘‘Es darf aber nicht vergessen werden,
dass bei Hunden sich gar nicht so selten akzessorische Schilddriisen
langs des Oesophagus nach abwarts finden, welche bis zum Pericard
herabsteigen kénnen.”* (2) Falta and Rudinger claim that comparison
can be made with their experiments only when the irritability of the
thyroidectomized animal to an electrical current has been determined

1 EppINGER, Fatta, and RupINGER: Wiener klinische Wochenschrift, 1908,
p. 241; Zeitschrift fiir klinische Medizin, 1908, Ixvi, p. 1, and 1909, Ixvii, p. 1.
* UNDERHILL and Hitpitcu: This journal, 1909, xxv, p. 66.
3 Fatta and RupIncErR: Zentralblatt fiir die gesamte Physiologie und
Pathologie des Stoffwechsels, to10, xi, p. 81.
4 Fatta and RupDINGER: Loc. cit., p. 83.
331

232 Frank P. Underhill.

in an endeavor to discover latent tetany. Underhill and Hilditch
made no such observations. (3) Finally, according to Falta and
Rudinger, Underhill and Hilditch made no autopsies to demonstrate
whether perchance thyroids or accessory thyroids were still present in
these animals.

In reply to the criticisms outlined above it should be stated, first
of all, that apparently Falta and Rudinger read only a review® of
the work by Underhill and Hilditch. In the original article ® there
are given at least two instances, out of a possible four, in which it
may be seen that the same or a smaller quantity of adrenalin than
that employed by Eppinger, Falta, and Rudinger induced glycosuria
in dogs after thyroidectomy. Falta and Rudinger explain sugar ex-
cretion in the single exception noted by them (Dog A ‘ or Dog I) by
assuming that the adrenalin was administered too soon after removal
of the thyroids (two days). This explanation will not suffice, how-
ever, for the appearance of dextrose in the urine when adrenalin was
injected according to the Eppinger, Falta, and Rudinger standard of
dosage in the case of Dog D.® This injection occurred six days after
thyroidectomy, and the statement is made by the above-mentioned
authors that at least three days after thyroidectomy adrenalin failed
to produce glycosuria.

Falta and Rudinger admit in the case of Dog A (Dog I), in which
glycosuria was caused eighty-one days after removal of the thyroids,
that there can be no question here of a latent tetany, but they inti-
mate that accessory thyroids may be present. At the time of the
publication of our former investigation it was considered undesirable
to kill Dog A (Dog I) in order to determine the presence of accessory
thyroids. Since then, however, an autopsy has been performed upon
this animal with this end in view. Every piece of tissue in any way
bearing a resemblance to thyroid tissues was carefully preserved
and sent for identification to Professor H. Gideon Wells of the Uni-
versity of Chicago. In his report concerning these tissues Professor

5 UNDERHILL: Zentralblatt fiir die gesamte Physiologie und Pathologie des
Stoffwechsels, 1909, x, p. 641.

6 UNDERHILL and Hitpircu: Loc. cit.

7 In the article by UNDERHILL and Hixpircu the animals were called A, B, C,
etc. In the review by UNDERHILL Dog A was called Dog I, Dog B was designated

Dog II, etc.
8 UNDERHILL and Hitpitcu: Loc. cit. ‘

The Production of Glycosuria by Adrenalin. 333

Wells makes the statement that he finds no evidence of thyroid tissue
or accessory thyroids; nothing except parathyroid tissue was in evi-
dence. During the present summer an autopsy was also performed
upon Dog C (Dog III). The result demonstrated the presence of
two somewhat hypertrophied parathyroids but no trace of thyroid

TABLE I.

Time after Adrenalin chloride Dextrose in
thyroidectomy. | per kilo injected. urine.

mgm. gm.
March 3 1.0 0.0
16 months
March 15 1.0 4.79
21 days 1.0 0.76
223 4 days 1.0 3.67

ARS) © 3 days 1.0 : 14.38

1 This animal was Dog A employed in experiments by UNDERHILL
and HirpircH. The dog never showed any signs of abnormality.
The body weight was 12.1 kilos.

2 A dog of 8 kilos in splendid nutritive condition. Both thyroids
were removed, leaving two parathyroids intact. Fed a mixed diet.
Autopsy revealed absence of thyroid tissues.

3 Well-fed dog of 13.2 kilos. No evidences of thyroid tissues on
autopsy.

4 Dog of 12 kilos, in good condition. No thyroid tissues on autopsy.
None of these animals showed any abnormality.

tissue, nor could any thyroid tissue be found at the autopsy of Dog D.
The results of the autopsy of Dog B (Dog II) were noted in our
former communication.

The criticism of Falta and Rudinger concerning the points just
discussed are not extremely potent in view of the autopsy findings
now reported, together with the observation that all dogs gave glyco-
suria after removal of the thyroids on treatment with adrenalin, and
that two of the four reacted positively with the same dosage employed
by Eppinger, Falta, and Rudinger. Nevertheless, in order to decide
the question even more definitely further experiments have been un-
dertaken the results of which may be seen in Table I. The methods
employed were identical with those outlined in our former commu-

334 Frank P. Underhill.

nication, except that in the observations here reported all adrenalin
injections were made subcutaneously. It may be seen from these
data that adrenalin chloride administered subcutaneously to dogs in
the dosage of 1 mgm. per kilo body weight is capable of provoking
the appearance in the urine of significant quantities of dextrose,
three, four, and twenty-one days,, and thirteen months after thy-
roidectomy. In no case were there any abnormal manifestations,
nor could any thyroid tissues be found on autopsy.

In the present investigation, as in the previous one, we have deemed
it unnecessary to determine the response of the thyroidectomized
animal to electrical stimulation in order to discover latent tetany.
The observation that none of the animals selected behaved in an
abnormal manner, together with the fact that Dog A (Dog I) and
Dog C (Dog III) of our previous experiments lived more than six-
teen months without tetany, that like conditions obtained for Dog D,
killed ten days after thyroidectomy because of an abscess at site of
injection, and that Dog 21 of the present investigation was allowed
to live more than twenty-one days after removal of the thyroids —
these facts all speak against the idea that latent tetany was present.

The criticism that adrenalin injection was made too soon after
operation in the case of Dog A (Dog I) of our former investigation
will not hold for our present experiments, since in no case was adrenalin
introduced under three days after removal of the thyroids. A survey
of the work of Eppinger, Falta, and Rudinger reveals that in several
instances their own injections were made three and four days after
thyroidectomy.

In the paper by Falta and Rudinger two new experiments ° are
detailed designed to corroborate former statements. Concerning the
first experiment the question may well be asked, ‘“‘ Why was the dose
of adrenalin, 10.5 mgm. (less than 1 mgm. per kilo for a 13-kilo dog)
divided into two portions (6 mgm. and 4.5 mgm.) and these injected
on two separate days?’’ Such a procedure does not add weight
to the statement of Eppinger, Falta, and Rudinger that 1 mgm.
adrenalin per kilo body weight administered subcutaneously or in-
traperitoneally into thyroidectomized dogs is incapable of causing
glycosuria. The failure of these quantities of adrenalin to provoke
the appearance of sugar in the urine can be duplicated at times in

® Fatta and RupiINnceER: Loc cit., p. 82. y

The Production of Glycosuria by Adrenalin. 335
normal dogs. In fact, even 1 mgm. per kilo often fails to induce
glycosuria (see Dog 22, Table II), and in Dog A (Dog I) without
thyroids (see Table I) it may be seen that this dose failed on March
3, whereas on March 15 it caused the excretion of 4.79 gm. dextrose.
From our experience with normal dogs we have been led to the con-
clusion that, although in general 1 mgm. adrenalin per kilo body
weight administered subcutaneously or intraperitoneally is capable
of causing glycosuria, animals are frequently encountered in which
this dose provokes no glycosuria. On the other hand, these same
animals at other times behave in the usual way and react to doses of
I mgm. adrenalin per kilo.

The second experiment of Falta and Rudinger shows that 5 mgm.
adrenalin injected intraperitoneally before removal of the thyroids
caused a slight glycosuria. After thyroidectomy the same quantity
injected intraperitoneally failed. Later 10 mgm. introduced sub-
cutaneously into two places also failed to provoke glycosuria. In
several experiments we have noted the absence of dextrose in the
urine following the intraperitoneal introduction of 5 mgm. adrenalin
into normal dogs smaller than the one employed by Falta and
Rudinger. The fact that 5 mgm. in this instance failed to produce
glycosuria does not necessarily mean that this has been due to thy-
roid removal. When the subcutaneous injection was given, the
animal weighed 16.2 kilos, and yet only 10 mgm. adrenalin were ad-
ministered. If Falta and Rudinger desired to offer evidence in sup-
port of their former statement that 1 mgm. adrenalin per kilo will
not cause glycosuria in thyroidectomized dogs, why did they not
inject enough adrenalin to comply with their own conditions? Again
in a footnote the following is given concerning the dog of the second
experiment: ‘‘Der Hund bekam vor dem 2. Versuch auch Pituitri-
num infundibulare. Dieses hat nach unseren Untersuchungen keinen
Einfluss auf den Kohlehydratstoffwechsel.”’!° Nevertheless, in an
experiment the results of which may be so significant it would have
been much better to have eliminated this last unnecessary conflict-
ing factor.

The experiments of Falta and Rudinger do not support the original
statement of Eppinger, Falta, and Rudinger that 1 mgm. adrenalin
per kilo administered subcutaneously or intraperitoneally fails to

10 FaLTA and RuDINGER: Loc. cit.

336 Frank P. Underhill.

provoke glycosuria in thyroidectomized dogs, since in neither of the
protocols reported is there any indication that these authors intro-
duced 1 mgm. adrenalin per kilo body weight.

In an article by Grey and de Sautelle™ the conclusion is drawn
that after thyroidectomy in dogs glycosuria evoked by adrenalin is
much smaller than in the normal animal. The method of procedure
adopted by these investigators was as follows: Dogs were kept upon
a fixed meat diet for several days. They were then injected with
adrenalin, after which thyroidectomy was performed. During re-
covery from the operation a mixed diet was fed. Then meat was
again given and adrenalin administered a second time. In the two
experiments recorded less sugar in the urine was obtained after thy-
roidectomy, as a result of adrenalin injection, than was excreted by
the normal dog.

The results obtained by these authors, namely, the appearance of
sugar in the urine of thyroidless dogs after administration of less
than 1 mgm. adrenalin per kilo, stand in direct opposition to those
reported by Eppinger, Falta, and Rudinger, but they are in perfect
harmony with the observations of Underhill and Hilditch. Grey and
de Sautelle were also unable to find any thyroid tissue at autopsy.

On the other hand, the experiments cited hardly warrant the con-
clusion drawn by the authors, namely, that after thyroidectomy the
glycosuria, produced by adrenalin in the normal animal, is greatly
reduced. The investigation was evidently carefully planned and
executed, but was based upon an assumption the correctness of which
is questionable. Consequently the conclusion drawn is not firmly
established. From the data presented it is apparent that these
authors assumed that if the same normal dog is kept under constant
conditions and given equal doses of adrenalin at two different times
the quantity of sugar excreted after these injections should be approxi-
mately the same. If that assumption was not made, then the ex-
periments are purposeless. All experimental evidence, however,
points against such an assumption, for the same normal dog under
constant conditions does not necessarily excrete the same quantity
of sugar with the same dosage of adrenalin given at two different
times. If a normal dog will not invariably respond to adrenalin

™ Grey and DE SAUTELLE: Journal of experimental medicine, 1909, xi, p. 659.

The Production of Glycosuria by Adrenalin. 337

twice alike, it is fallacious to attribute a lessened elimination of sugar
to the loss of the thyroids without at least the support of a great
number of experiments all showing the same marked tendency. In
an experiment having as its object the study of the influence of the
thyroids upon carbohydrate metabolism, it is, therefore, apparent that
quantitative changes in sugar excretion should be given little weight.

As corroboratory to the views just expressed the data in Table II
are submitted. In these experiments the plan followed was very
similar to that outlined by Grey and de Sautelle, and although the
details differ somewhat the end aimed at, to keep conditions of diet
constant, was attained. The animals had been fed upon meat for
several days before the experiments began. They were then fed
upon a constant mixed diet for a period of five days. Then adrenalin
was given subcutaneously, — 1 mgm. per kilo body weight. On the
day of the injection the usual quantity of water was given but no
food. As a rule sugar elimination ceased within twenty-four hours
after adrenalin administration. The dogs were then fed upon meat
until five days before the second adrenalin injection. During these
five days the mixed diet was again given. As before, no food was
offered on the injection day. After the second adrenalin administra-
tion the thyroids were completely removed, but at least two para-
thyroids, one on each side, were left intact. This was confirmed at
autopsy. During the period of recovery from the operation the dogs
received the meat diet. Five days before the third adrenalin injec-
tion the mixed diet was fed, and as previously no food was given on
the day of injection. By such a régime the animals remained prac-
tically constant in weight. In every instance 1 mgm. adrenalin per
kilo was administered subcutaneously in the region of the lower ribs.
The subcutaneous injection possesses the following advantages:
animals do not die so frequently as with the intraperitoneal injection,
and are not so likely to vomit or have diarrhoea or bloody urine. In
this particular point we differ from Grey and de Sautelle, since their
injections were made intraperitoneally. The principle, however,
is identical in the two cases, since the mechanism involved in each
case is the same. Furthermore, there is no basis for assuming that
adrenalin given intraperitoneally will show a different behavior
concerning the point under discussion than adrenalin administered
subcutaneously.

338 Frank P. Underhill.

The data presented in Table II demonstrate conclusively that
adrenalin administered twice to the same normal animal under like
conditions does not necessarily provoke the same degree of glycosuria
in the two instances. Moreover, in every case reported adrenalin
induced glycosuria after thyroidectomy, and the quantity of sugar

TABLE Il.

SUGAR IN URINE IN GRAMS.

Before thyroidectomy.
After
thyroidectomy.

First injection. | Second injection.

March 3 0.0

March 15 4.79
Died during

1.27 operation

9.70 0.76
0.0 3.67
22 14.38

1 This animal was a dog of 7.0 kilos. The details concerning the
other dogs are given in the footnotes of Table I, p. 333. In each ex-
periment 1 mgm. adrenalin per kilo was injected subcutaneously.

eliminated after the operation was not uniformly decreased. In
fact, in two of three experiments detailed more sugar was excreted by

the thyroidectomized dog than appeared in the urine of the normal
dog.

CONCLUSIONS.

Renewed investigation concerning the efficiency of adrenalin in
provoking glycosuria in thyroidectomized dogs leads to a reiteration
of our former conclusion that adrenalin chloride administered sub-
cutaneously in doses of 1 mgm. per kilo body weight causes a signifi-
cant glycosuria in dogs deprived of both thyroids but retaining at
least two parathyroids. The criticisms of Falta and Rudinger with

respect to our former experiments have in no way invalidated this
conclusion.

The Production of Glycosuria by Adrenalin. 339

In the investigation by Falta and Rudinger, put forth in support
of the conclusions deduced by Eppinger, Falta, and Rudinger, they
have failed to comply with the conditions laid down by the latter.
Consequently the results offered by Falta and Rudinger cannot be
accepted as proof that adrenalin administered to thyroidectomized
dogs in doses of 1 mgm. per kilo is incapable of causing glycosuria.

The observations of Grey and de Sautelle are in harmony with our
own results and stand in direct opposition to the position taken by
Eppinger, Falta, and Rudinger. The validity of the conclusions
drawn by Grey and de Sautelle, however, may be questioned, since
the investigation was based upon an assumption the correctness of
which has not yet been established.

Experiments are reported demonstrating that adrenalin admin-
istered subcutaneously to normal dogs in doses of 1 mgm. per kilo
causes a widely varying degree of glycosuria. The same quantity of
adrenalin introduced into thyroidectomized dogs under like condi-
tions is capable of inducing as great or even a greater glycosuria than
occurs with the normal animal.

STUDIES IN EXPERIMENTAL GLYCOSURIA.— VI. THE
DISTRIBUTION OF GLYCOGEN OVER THE LIVER
UNDER VARIOUS CONDITIONS. POST MORTEM GLY-
COGENOLYSIS.

By J. J. R. MACLEOD anp R. G. PEARCE.

[From the Physigqlogical Laboratory, Western Reserve University, Cleveland, Ohio.]

CONTENTS.
I. The distribution of glycogen over the liver immediately after and some time
ARTETA Cal ite ae. ao pem ote, at al. etumeieaer. aimee Reer av etreahaoe @ <4 fe ea taceteet 342
II. The distribution of glycogen over the liver at varying periods after feeding with
eanponyCrareeiGh LOOCmei es aise) Stenger foie) yeeleiotde. Fe saleh, of for Goinarets Sah 346

illee Rosi moriem ely COsenalysis, | sry.) suse, -eee tt 20), ee ss A ae
(a) Comparison of the rate of glycogenolysis in the liver during ether anzs-
thesiavandcattercdedtny. <5 ei" e wren Ste cate Ae owe ee 353
(b) The time of onset of post mortem glycogenolysis and the reaction velocity 356
(c) Analysis of some of the factors that determine the velocity of post mor-

fevarely Gar EnOlySisnmey me, ut cae eee ve, taiken eel cy ten 359

(d) The influence of stimulation of the great splanchnic nerve on post mortem
eG PENG IVES oo oil vc Acie! EA oe ds pak doniseg “glade, dbp olsbjs epee peers 362
MVAEBOUCIUSIONS] sy Me a homie a is= WA Meath bas antes ee Gaines seme. tone ee ors ven aie 364

i studying the influence of various conditions on the glycogenic
function of the liver, it is customary to make a comparison of the
amount of glycogen contained in a portion removed just prior to the
experimental condition under investigation with the amount found
in another portion removed during or immediately after it. The
difference between these values indicates the amount of glycogenoly-
sis in a given time, and this difference is conveniently expressed as a
percentage of the original amount of glycogen present. This percen-
tile glycogenolysis, as we may call it, is then compared with that oc-
curring in a liver kept for a similar time as nearly as possible under
normal conditions.

Although this method has been frequently employed by various
workers, there is some considerable uncertainty regarding several
fundamental data on which the calculations involved in it depend.
These are especially with regard to:

341

342 J. J. R. Macleod and R. G. Pearce.

1. The distribution of glycogen over the liver, particularly during
ether anesthesia and following the taking of food.

2. The average course of post mortem glycogenolysis (7. e., its time
of onset and the velocity with which it proceeds).

Having employed the above-described principle for estimating
glycogenolytic activity in connection with some of the work which
has been in progress in this laboratory during the past two years, and
finding that the above-mentioned fundamental values might vary
considerably, quite independent of the experimental conditions un-
der investigation, we have thought it important to repeat many of
the observations, more especially with the object of ascertaining to
what cause these accidental variations are due.

I. THe DISTRIBUTION OF GLYCOGEN OVER THE LIVER IMMEDIATELY
AFTER AND SOME TIME AFTER DEATH

Regarding the previous work in this connection the following
articles are of importance: Kiilz,’ using the Briicke-Kiilz method,
found in the dog’s liver divided into four parts the following per-
centages of glycogen: 5.18, 5.05, 5.0, and 4.96; the greatest differ-
ence was 0.26, or 4.24 per cent. In another dog’s liver divided into
three parts the percentages were 2.45, 2.25, and 2.19; greatest differ-
€nce; 0:26; OF 11-9 per cent.

Cramer,” working on rabbits’ liver divided into three parts, found
the following percentage amounts of glycogen: 0.9038, 0.9528, and
0.9438; the greatest difference was 0.490, or 5.1 per cent.

Schéendorff * found in the dog’s liver a percentage of 18.82 in one
portion and 18.33 in another; difference, 0.49, or 2 per cent.

Grube,* at the instigation of Pfliiger, undertook a more thorough
investigation of this question. He used dogs that had been starved
from one to five days and then fed with a mixed diet containing an
excess of carbohydrate. The dogs were killed by hemorrhage at a
variable time after feeding, and the following results were obtained:

Kitz: Zeitschrift fiir Biologie, 1886, xxii, p. 183.
CRAMER: Zeitschrift fiir Biologie, 1888, xxiv, p. 85.
SCHOENDORFF: Archiv fiir die gesammte Physiologie, 1903, xcix, p. IgI.

1
2
3
* GruBE: Archiv fiir die gesammte Physiolugie, 1905, cvii, p. 483.

Studies in Experimental Glycosuria. 343

1. Dog fed for three days after a period of five days’ starvation. Liver
divided into four parts. Average amount of glycogen, 16.6 gm.; greatest
difference, 0.86 gm., or 5.4 per cent.

2. Dog killed twelve hours after feeding, following two days’ fast. The
entire liver was minced, and four portions of the mixed mince analyzed.
Average percentage, 7.4; greatest difference, 0.055, or 0.76 per cent.

3. Dog killed fourteen hours after feeding, following thirty-six hours’
fast. Entire liver minced, and four parts of mixed mince taken for analy-
sis. Average percentage, 6.4; greatest difference, 0.039, or 0.6 per cent.

4. Dog killed fourteen hours after feeding, following a twenty-four-hour
fast. The peripheral portions of four different lobes were analyzed sepa-
rately. Average percentage of glycogen, 9.73; greatest difference, 0.098, or
fen per cent.. 4,

5. Dog killed six hours after feeding, following a twenty-four-hour fast.
Portions from three different lobes were analyzed. Average percentage of
glycogen, 0.19 (one result was however very low, viz., 0.139); greatest dif-
ference, 0.048, or 24.3 per cent. It is suggested by Grube that the unusually
low value which one of the portions in the liver of Dog V showed is prob-
ably due to this portion having contained a large amount of connective
tissue, blood vessels, etc. It is to be noted that it is only in Experiments 1,
4, and 5 that the glycogen in different lobes is compared.

Sérégé * found in dogs that during the first two hours following the
taking of food (that is to say, during digestion in the stomach) there
was more glycogen in the left half of the liver than in the right; during
the next six to eight hours (that is to say, during digestion in the in-
testine) the distribution was reversed, there being more glycogen in
the right lobes than in the left. From twelve hours on, the liver
again contained more glycogen in the left lobes. Sérégé employed
the Frankel method (the precipitation of proteins by trichloracetic
acid), which Pfliiger has shown to be quite unreliable.

Lépine,® without however giving any experimental evidence, con-
siders it is definitely settled that the distribution of glycogen over
the liver is irregular.

These results, taken as a whole, would indicate that the distribu-
tion of glycogen over the liver does not usually vary by more than
5 per cent; the greater variation of 12 per cent, observed by Kiilz,

® SEREGE: Comptes rendus de la Société de Biologie, 1905, lviii, p. 52.
6 LépINE, R.: Le diabéte sucré, 1909, p. 99 (Paris, Félix Alcan).

344 J. J. R. Macleod and R. G. Pearce.

being exceptionally great, and the high variation in Grube’s last
result being partly due to the experimental error incidental in
estimating such small quantities of glycogen as this liver contained.

In the present research the portions of liver for analysis were re-
moved as simultaneously as possible, cut into thin slices, pressed free
of blood between filter paper, and treated by Pfliiger’s process for the
estimation of glycogen.

The following are, briefly, the results:

Rabbit I.— Fed on preceding day with carrots; killed by stunning.
Liver immediately excised and placed in incubator.
Percentage amount of glycogen (dextrose) in:

Quadrate lobes (immediately after death) 11.05 (20 min. later) 10.31

left lobess(omin® after). -o. 5. Grae & 10.80 (20 min. after) 10.21
Diferenceser2 *. 4 ee ees 0.10
Difference in per cent of (akwettas amount 2.20 0.97

Rabbit II.—Same procedure followed as in Rabbit I. Percentage
amount of glycogen (dextrose) in:

Immediately Twenty Fifty min.
after death. min. later. later.
Qradrate lame: . 7s. c/c ik eee Oe ee oeoae 8.111 6.733
JES 12) ee a rane omen + ieee hee 8.793. 7.960 5.an
Diflerence: .6 .. 4: id) he Ocenia fo ea 0.801
Difference in per cent of (ageer amount 0.53 1.86 11.8

Dog I (Exp. 97). — Specially fed the day before on cane sugar; killed
by bleedipg. Liver left im situ, but animal kept on warm tank. Tempera-
ture of front lobes, 30° C.; temperature of back lobes, 36° C.

Percentage amount of pe (dextrose) in:

Before Ten min. Twenty min.
death. later. later.
Prouitowess.. ve Gai ea. ee tetas 7.664 7-316
BACK TODES s,s. WE sy es ee ee 6.750 7.025 6.234
Pxtierenes tere a See Oza s 0.639 1.136
Difference in per cane af Taree amount 6.2 8.3 15-4

Dog IT. — Dog fed with bread and meat; died while under anesthetic - |
during experiment of injecting N/8 lactic acid into vena pancreatico-
duodenalis.

Studies in Experimental Glycosuria. 345

Percentage amount of glycogen (dextrose) in:

Before Onehr. Two hrs. Three hrs. Four hrs. Five hrs. Six hrs.
death. later. later. later. later. later. later.

Bett lobe ~. . 4.675 4.016 3.726 3-610. 3-750 3-450 3-233
Right central. 4.620 4.165 3.866 3.570 3.230 3.186 3.033
Wiitereice . 0.055 (+... ins 0.040 0.526 0.264 0.200

Dog III. —Dog not specially fed; killed in course of another experi-
ment. Liver excised.
Percentage amount of glycogen (dextrose) in:
Immediately Twenty min.

after death. after.

Left lobe. NO eae em Rea ae 2.506 2.118
cmuteerreraee a2! 3° Some 13 eels 2 RD Sy tote sh ELOSD
Retaer Rea ek. Pregl Ho yrs oe! eg Cage al ESS
(Scie) AS) s Aon error Sane yee Pera oe ib Nguucree eed s ERT
Greatest difference .. . 0.955

Greatest difference in per ape af bees nat 38.2

Dog IV. — Dog not specially fed; had been used (under deep anesthesia)
in three-hour experiment on blood pressure.
Percentage amount of glycogen (dextrose) in:

Left lobe (back). . . . 3.290 Caudate (front) . . . 2.990
ere ilobe (front), +... 3-140. +Right central ~~ - 73.180
Peltrcented 2 4-820, bapht lope]. 2. . '2 3-070
Caudate (back) . 23770 = opieciiate on”. i0a.) 2. © 2.800
Greatest ditched Ae Ue Ee 73) SGC

Greatest difference in per one af nest pocnt 16.1

Dog V.— Dog fed with bread and meat; died of hemorrhage in course
of operation for making Eck fistula. At time of death had been under
anzsthesia for about half an hour. The liver was placed in the incubator
at 40° C. between the periods of removal of portions for analysis-

Percentage amount of glycogen (dextrose) in:

Immediately Fifteen min. Thirty min. Forty-five

after death. later. later. min. later.
Rete tobe ..5 2-3 Eh ri Se TGS 2.98 2.62 2.38
Right central... 0. 2... 3.82 2.89 2.61 2.40
DOG ce oe rir - 0.04 “0.09 0.01 0.02

Difference in per cent of
larger amount .. . 1.05 3.02 0.38 0.82

346 J. J. R. Macleod and R. G. Pearce.

Several conclusions of practical importance regarding the distribu-
tion of glycogen over the liver can be drawn from these observations:

1. When the liver is removed from an animal (rabbit or dog) that
has just been placed under anesthesia, the percentage amount of
glycogen in the different lobes does not differ by more than about
6 per cent, the difference being usually much less than this. These
differences are probably due to varying amounts of connective tissue
and of blood in the different portions of the organ.

2. At varying periods after death, the post mortem disappearance
of glycogen proceeds at first to an equal degree in the different por-
tions’ of the liver, provided the liver be excised and kept at equitable
temperature (Rabbit I, Dogs II and V).

3. If the liver be left in the body after death, post mortem glyco-
genolysis proceeds at a variable rate in the different portions of it, so
that considerable differences are found in the percentage amount of
glycogen in the different lobes. These results are partly due to differ-
ences of temperature between the deep and the superficial lobes
(Dog I).

4. When, prior to death, the animal has been for some considerable
time on its back under anesthesia, a considerable difference in the
glycogen content is found in the different portions of the liver. What
it is that occasions these differences is not quite clear, but it is prob-
ably associated with unequal blood supply, for it will be shown later
that the rate of glycogenolysis in liver tissue is very much influenced
by the amount of blood in the viscus (Dogs ITI and IV).

II. THe DISTRIBUTION OF GLYCOGEN OVER THE LIVER AT VARYING
PERIODS AFTER FEEDING WITH CARBOHYDRATE-RICH FOoop.

It has already been pointed out that Sérégé claims to have found
an unequal distribution of glycogen in the liver during the different
periods of digestion. The results on which this conclusion is based
have been seriously called in question by Pfliiger,’ who further quotes
Grube’s observations (Grube, loc. cit.) as absolutely refuting Sérégé’s
conclusions. This is not, however, the case, for Grube’s observations
do not bear directly on this point. The shortest period after feeding

7 Priticer, E. F. W.: Das Glycogen, 2d ed.

Studies in Experimental Glycosuria. 347

during which Grube determined the amount of glycogen in the difler-
ent portions of the liver was six hours, but in that case only traces of
glycogen were found, so that nothing final can be concluded from the
result. The next period was twelve hours, and in this, as in the later
periods, equality of distribution was observed, thus disproving at
least one of Sérégé’s conclusions, namely, that after absorption is over
the left lobes are richer in glycogen than the right.

It is undoubtedly of great importance to definitely decide this
question, not only because of its practical bearing, but on account of
its physiological interest. Should Sérégé’s observations be corrobo-
rated, then would it be evident that the mechanism which controls
the deposition of glycogen in the liver is not a perfectly acting one:
it would indicate that one portion of the viscus becomes partially
filled with glycogen before any more of this substance is stored in
other portions. Sérégé attributes this inequality of distribution to
the fact that the blood of the mesenteric vein proceeds more or less
directly to the right lobes without becoming mixed with that of the
splenic vein which proceeds to the left lobes. During absorption
therefore the right lobes receive the greater proportion of absorbed
food stuff. Although, on a priori grounds, such a piece-meal deposi-
tion of glycogen in the liver is improbable, yet the possibility of its
occurrence must be admitted, and there is nothing in Grube’s results
which absolutely refutes the hypothesis. The following experiments
were performed in connection with this question :

Five dogs, each weighing about to kilos, and as nearly as possible
alike in age and breed, were starved for five days. On the sixth day
each dog was fed by the stomach tube with soup having cane sugar
dissolved in it, corresponding quantities being given to each dog.
The following was the further plan of the experiment:

8.30 A.M. Each dog given about 250 gm. cane sugar in soup.

9.30 “ Again given some sugar.
10.30 “ Again given some sugar.
11.30 “ Dog I killed by chloroform.
1.20 P.M. Dog II killed by chloroform.
ZEBO Remaining dogs given some sugar.
3.20 “ Dog III killed by chloroform.
5.20 “ Dog IV killed by chloroform.

8.30 A.M. Dog V killed by chloroform.

348 J. J. R. Macleod and R. G. Pearce.

The following are the results:

Dog I. — Weight, 9.950 gm. Time since food first given, three hours.
Time since food last given, one hour. Liver weight accidentally lost.
Percentage amount of glycogen (dextrose) in:

erelobe ays. 4 ere heels 3.500 3
Sra (1)
3.530 3-515
estvecotral«m > 2. fase 3.549 .
4640 3.594 (3)
Caudatenat: ae te Rest hat, a. 3.505
485 (1
oe 3-485 (1)
Right central 3.310
mab 3-335 (2)
| Bsc al (0) 0 ge 2:
ae eae 3.045 (1)
3.040
Greatest idifierence .-<,npieding cos) em bene 0.549

Greatest dif. in percentage of largest amount . 15.5

Dog II. — Weight, 11.400 gm.; liver weight, 418 gm. Time since fed,
three hours. Weight of liver, 3.66 per cent of body weight.
Percentage amount of glycogen (dextrose) in:

bett lobe: - °)..9 << eee ee 4.600

4.670 4.635 (2)
Lett central: 1.5 % (Sage es baa 209
-L7E
ie 4.171 (5)
ete: ath s.r) hems al oe: 4.104 4.104
Right central nic. io dpe. cen 4.149
ED °
ee 4.149 (0)
Rightlobe:.. Gaui) eee Be gk :
igh abe ues 4.126 (1)
4.119
Greatestrdifierenceuen <x \isoea cos a Boole 0.531

Greatest difference in percentage of largest amount 11.2

Dog IIT. — Time since first fed, five hours. Weight, 10.200 gm. Weight
of liver, 341 gm. Weight of liver, 3.34 per cent of body weight.
Percentage amount of glycogen (dextrose) in:

8 The figures in parentheses indicate the approximate percentage error between
the duplicates.

Studies in Experimental Glycosuria.

Left lobe

Left central
Caudate
Right central

Right lobe .

Greatest difference

. . .

a: ee .446

ue 3.481 (2)
Povis cent 3.468 3.468
4:42 ZgE2
wee yahein ¥

mp 3.094 (3)
oe Ree eee 3-054 3.054

Be SE aN see" Se ae te 0.427

Greatest difference in’percentage of largest amount 12.2

349

Dog IV. — Time since last fed, three hours, but seven hours since main
feeding. Weight, 9.300 gm.; weight of liver, 409 gm. Weight of liver, 4.39

per cent of body weight.

Percentage amount of glycogen (dextrose) in:

Left lobe

Left central
Caudate
Right central

Right lobe
Greatest difference

Greatest difference in percentage of largest amount

Dog V.— Time since last fed, eighteen hours.
weight of liver 395 gm. Weight of liver, 3.65 per cent of body weight.

gate! Pee. prema

har ears .10

ee 9-393 (5).
hg lke 852

vee eee al
ere ee 10.310

ee. 10.198 (2)
sie alge hay ia rae 9721 (x)
gs ee 8.970 8.970
EE a ad pia ree a 1.228

12.04

Percentage amount of glycogen (dextrose) in:

Left lobe

Left central
Caudate
Right central .

Right lobe
Greatest difference

Greatest difference in percentage of largest amount

CA ae kee eee et

RE ee bor ee,

ee 7-525 (2)
attr ee ee 6.612

6.804 6.708 (3)
Ga tae seat aC 6.770

aes bres)
Cage) ea ae .O1

noe 6.801 (5)
Ne ee 6.700 6.700
fy eb OE Cel ae = yn a ge 0.825

10.9

Weight, 10.800 gm.;

350 J. J. R. Macleod and R. G. Pearce.

It is perfectly evident that in all the above observations more gly-
cogen was found in the left than in the right lobes. This difference
cannot be attributed to the supposed inequality of blood supply to
the different lobes during absorption, for then, according to Sérégé,
the right lobes slould have contained more glycogen than the left
in Nos. 1, 2, 3, and 4. The observed differences are undoubtedly duc
to the occurrence of post mortem glycogenolysis in the liver during
the process of removal and weighing of portions for analysis. In the
above observations the portions of liver were removed in the order
in which the results are reported, and it took from twenty to thirty
minutes to make all the necessary weighings. After the portions of
liver had been weighed they were placed in flasks and potash added
simultaneously to all, but since, as we shall show later, the post mor-
tem process is more rapid in the intact viscus than in sections of it,
there was opportunity for considerable variation in the amount of
glycogen between the first and the last removed portions. It would
have been more accurate to cool the liver in ice immediately after its
removal. This was done in the following two experiments:

Dog VI.—Starved five days. Fed with 10 gm. per kg. cane sugar and
killed by chloroform six hours later. The liver was immediately excised
and placed in freezing mixture. Wedge-shaped portions were then removed
and analyzed, beginning with left lobe.

Percentage amount of glycogen (dextrose) in:

LEST 4 (0) ot ay eee eee Rem Ty

ae 10.845 (0.3)
Benrcential | ie etels-. weet oh. emia 10.98
ee 10.890 (1.6)
Catidiateicas Mi naga GR Aneel 10.86 (2.7)
es Lo, 7EOwess
PUI COME hy Mi act oe cans ee Li.34
eG 11.350) (GuEy}
Right lobe: -45%,1.% sce eee ee jo)
x i eee 11.610 (1.5)
Ti52
Greatest difference.) 2 . fo. 9 ss . 2a 0.900
Greatest difference in per cent of largest amount . 7.6

Dog VII. — Treated in same way as Dog VI, but killed in twelve hours
after feeding and portions of liver first of all removed from right lobes.

Studies in Experimental Glycosuria. 351

Percentage amount of glycogen (dextrose) in:

BONG ea pais a) se) ¥ 14.25 :
13.88 14.065 (2.5)
etteeeNtTabes sn. ek ss a TALBS
14.700 (2—
ee 4.700 (2—)
Peeled Miers oly. Meth of aos 14.70 be
Oe 14.840 (2—)
Mewicewural ti log. (et ey ae 14.28 ee)
13-99
ee ree. Map y Toktee sa: tay dl Pare rs 14.43,
Gis 14.445 (0.2)
14.46
Greatest difference .....- +++ - +s i, O98
Greatest difference in per cent of largest amount . 5.25

In the light of these last two experiments (6 and 7) it is evident that
the differences in glycogen content of the various lobes observed in
Dogs I, Il, III, and V were largely due to post mortem glycogenolysis.
The experimental error in the case of Dog IV was higher than usual,
but even in this case a great part of the observed difference is un-
doubtedly due to the same cause. In the last two experiments (6 and
7) the differences observed are about the same as those already given
as the normal variation due to inequality of connective tissue and
blood (see pp. 344, 345).

Taking the results as a whole, we may conclude that even when
every precaution is taken against unequal post mortem change the
distribution of glycogen over the liver is not perfectly uniform. It
may vary by from 5 to 7 per cent, being, however, no greater during
absorption than at other times. It is impossible from these investi-
gations to refute Sérégé’s statement ® that during absorption from
the intestine glycogen is deposited more quickly in the right lobes
than in the left (indeed some of our results would seem to confirm
it, cf. No. 6), but-it is clear that the differences, if they exist, are of
small magnitude, being entirely masked in most of our experiments
by post mortem glycogenolysis.

9 In Dog VI it would at first sight appear that Sérégé’s observations are con-
firmed (i. e., that during absorption from the intestine glycogen is deposited more
quickly in the right than in the left lobes). The difference is, however, too small
to warrant any such conclusion.

352 J. J. R. Macleod and R. G. Pearce.

III. Post MortrEmM GLYCOGENOLYSIS

In certain preliminary observations on the rate of disappearance
of glycogen from the liver after death, we were struck with the fact
that it is an extremely variable process. The causes of the variations
are obscure, and it was deemed essential for further progress to sub-
ject them to a more thorough investigation. In a previous contribu-
tion by one of us (in collaboration with H. O. Ruh) #? in which the
rate of post mortem glycogenolysis in the liver during stimulation of
the splanchnic nerve is compared with that occurring without such
stimulation, the following proviso is made in connection with the dis-
cussion of results: ‘Before drawing any final conclusions from the
experiments here recorded, it is evident that we must be furnished
with more reliable and extensive data regarding the course of post
mortem glycogenolysis.”’ It is pointed out in that connection that
very little indeed is known about this process, either as regards the
time of onset, or as to when it attains its maximum velocity.

In practically every form of hyperglycemia increased production
of dextrose by the liver is the initial cause of the condition. It is true
that in the severer forms of so-called experimental diabetes (pan-
creatic, phloridzin, etc.), dextrose goes on accumulating in the blood
after all the glycogen in the liver has been used up, but even in the
late stages of at least some of these conditions, a serious derangement
of the glycogenic function of the liver is known to exist, as evidenced
by the fact that glycogen is no longer deposited in this viscus even
when large amounts of carbohydrate food are ingested. What is it,
therefore, which makes this glycogenic function so susceptible to de-
rangement? Although we have at present no justification for believ-
ing that the cause of post mortem glycogenolysis is the same as that
which brings about this process in experimental diabetes, yet it is evi-
dent that more accurate information regarding the former process —
the cause of its onset, the nature of the conditions which accelerate
or retard it, etc. — cannot but be of value to us in arriving at the ex-
citing cause of the ante mortem process.

In a previous investigation by one of us on the cause of asphyxial
glycosuria it was shown that in an incubated mixture of minced
liver and blood glycogenolysis is much accelerated by the presence

10 Macteop, J. J. R., and Ruu, H. O.: This journal, 1908, xxii, p. 397.

Studies in Experimental Glycosuria. 353

of an excess of carbon dioxide. It was further pointed out that it is
probably by increasing the acidity of the mixture that carbonic acid
produces this effect. After death acid substances (lactic acid) de-
velop in the tissues (including the liver — Magnus Levy), and the
thought immediately presents itself that the onset of post mortem
glycogenolysis is dependent upon the development of acid. Before
concluding that such is the case, however, much more must be known
about the nature of the post mortem process than at present exists;
its exact time of onset, its course, whether it is associated with an in-
crease in the amount of glycogenase in the liver, etc.

With regard to the onset of glycogenolysis, Pavy and Bywaters ™
have recently expressed a similar view to the above. There is nothing
in their paper, however, which corroborates it beyond the observation
that the acidity of alcoholic extracts of liver increases after death and
that injection of sodium carbonate solution in the portal vein pre-
vents post mortem glycogenolysis.

As already pointed out in the paper just referred to, the variations
in hepatic glycogenolysis that can be brought about experimentally
during life may likewise be associated with varying degrees of acidity
in the liver cells; thus the hyperglycogenolysis following muscular
work may, like the increased respiratory activity, be due to the pres-
ence in the blood of acid products of muscular contraction.

a. Comparison of the rate of glycogenolysis in the liver during ether anzs-
thesia and after death.— Dogs fed on the previous day with bread,
meat, and cane sugar were used for these experiments. Each ani-
mal was anesthetized with ether and a piece of liver removed as
quickly as possible from one of the right lobes, after which the dog
was placed on the warm operating table and the anesthesia main-
tained, other portions of liver being removed at regular intervals for
a period of about one and a half hours. The animal was then bled
to death and kept on the warmed operating table, portions of liver
being removed as previously. Each portion of liver after removal
was cut in thin slices, pressed between filter paper to remove blood,
weighed and dropped into an equal volume of 60 per cent KOH.
The glycogen determinations in the different portions were conducted

1 Pavy, F.W., and Bywaters, H. W.: The journal of physiology, 1910, xli,
p. 168.

354

J. J. R. Macleod and R. G. Pearce.

simultaneously, every detail of the process being exactly the same for

each portion.

The following table depicts the results of three such experiments:

Condition of
animal.

Ether
anesthesia

Dead
(hemorrhage)

Ether
anesthesia

Dead
(chloroform)

Ether

anesthesia

Dead
(hemorrhage)

TABLE I.

Per cent gly-

cogen (as dex-
trose) in por-
tions of liver
removed every

fifteen minutes.

gm.
6.36

6.46
5.76
5.94
4.98
4.16
3.93
3.51
SWith
9.17
8.92
8.22
8.60
192
(Be
7.41
6.94
6.95
6.56
3.90
S31
3.13
3.28
2.67
1.86
1.45
1eS2
1.065

|
|
:
|
|

1 Velocity constants calculated from the equation K =

, Percentage
amount of gly-
cogen (dex-
trose) disap-
peared during
each period.

Percentage
amount of gly-
cogen disap-
peared for
equal periods
before and af-
ter death.

Velocity
constants.!

0.40
0.59
0.18
0.23
0.71
0.41

0.13
0.255

per cent.

001537
.001407
001769
00250
00236

.00187
.00077

00105
00112
.00102
00115
00100
00106

(.00474)
.00317
00243
.00274
00427
.00477
.00448

1.605 .00470

amount of glycogen left after the time T.

1 G ‘
T log. CG? where C, is the

Studies in Experimental Glycosuria. 355

Several conclusions of importance can be drawn from these results:

1. Inan anesthetized animal (dog) there is a rapid disappearance
of glycogen from the liver.

2. The rate of disappearance varies markedly in different lobes.

3. There is sometimes an acceleration in the rate of glycogenolysis
when the etherized animal is killed.

The figures in the first two columns of the table reveal a remarkable
irregularity in the rate of disappearance of glycogen from the various
lobes of the liver during ether anesthesia. Such a result shows clearly
that in the anesthetized animal the usual method for determining the
rate of hepatic glycogenolysis is utterly unreliable (Croftan ”). The
irregularity in glycogenolysis is maintained after death, which, as
we shall see later, is not the case when death is produced by stunning,
or in some other sudden way. It cannot of course be inferred that
normal hepatic glycogenolysis likewise proceeds in an irregular man-
ner. We have already seen (p. 351) that while glycogen is being
rapidly stored up in the liver there is an almost uniform rate of ac-
cumulation in the different lobes. The above observations indicate
that when glycogen is being rapidly broken down quite another con-
dition obtains, the process being markedly variable in intensity in
different lobes.

With such irregularity in the rate of hyperglycogenolysis in the
various lobes it is necessary, in comparing the rate of glycogenolysis
before and after death, to employ values in which the variations will
be as much minimized as possible. This has been done in two ways:

1. By finding how much glycogen (per 100 gm. liver) has disappeared
for periods of one hour before and after death.

2. By calculating the velocity constant for the different time peri-
ods observed. The latter value will of course be influenced by the
irregularity above noted, but since it expresses the glycogenolysis as
a ratio of the original amount of glycogen present, the error will be
less marked, and the calculated value will furnish a reliable criterion
of the velocity with which glycogenolysis is proceeding at any given
moment. It allows for the decreasing amount of glycogen capable of
glycogenolysis.

By both methods of calculation it is evident that in Experiment 5
death caused an acceleration in the glycogenolysis. Under ether

2 Crortan, A. C.: Archiv fiir die gesammte Physiologie, 1909, cxxvi, p. 407.

350 J. J. R. Macleod and R. G. Pearce.

alone the process was gradually getting slower to become greatly ac-
celerated immediately after death. The velocity constant in Experi-
ment 3 also shows post mortem acceleration, but in Experiment 3 a,
throughout which glycogenolysis was distinctly slow, no change was
produced by death. This experiment differed from the others in the
fact that death was produced by an overdose of chloroform instead
of by hemorrhage. The chloroform, in the blood left in the liver,
probably retarded the glycogenolysis.

b. The time of onset of post mortem glycogenolysis and the reaction
velocity. — To investigate these questions, the animal should be sud-
denly killed without previous administration of anesthetic. This
has been done in rabbits by stunning in the present research. In the
case of the dogs used, the ether anesthesia was as short as possible.

The particular questions which we have sought to answer are these:

1. At what period after death does glycogenolysis become marked?

2. Does this process, after it appears, attain its maximum intensity
gradually or quickly?

3. Having attained its maximum intensity, does it remain constant
at this or gradually fall away?

The velocity constant at different stages of the reaction supplies
us with the most useful information from which to answer the above
questions. This constant will be greater when the conditions are
most favorable for the glycogenolytic process, and in calculating it
allowance is made for the progressive decrease in glycogen.

In making use of the velocity constant for this purpose it is impor-
tant to consider how this behaves during the hydrolysis of glycogen
solutions im vitro by means of diastase. Researches of such a nature
have been conducted by Philoche.”

When very strong preparations of diastase were used (1 to 50), it
was found that glycogen was completely converted into maltose in
about fifty hours in a 2 per cent solution. When feebler diastase
preparations were used, however, the process ceased long before this
stage. Investigation showed that between glycogen and maltose in
the hydrolytic process are large amounts of dextrines into which
glycogen is readily transformed, but from which, after some time,
maltose ceased to be produced. This inhibition in the process was not
due to adsorption or destruction of the diastase. Under the circum-

18 PHILOCHE, CuH.: Journal de chimie physique, 1908, vi, p. 359.

Studies in Experimental Glycosuria. 357

stances it was impossible for Philoche to calculate the velocity con-
stant. If we take for comparison sake the velocity constants as
determined for the action of diastase in starch, we find that these
gradually decline. When large amounts of substrat are present,
however, the first part of the process forms a linear curve and the
latter part a logarithmic.“

Taylor ™ has computed the velocity constant in the case of the
liver of clams kept at a constant low temperature. He found it to
progressively decrease, even although he employed for its computa-
tion the disappearance of glycogen and not the appearance of re-
ducing sugars.

In the present research we have determined the constant in the
liver of two specially fed dogs, killed as quickly as possible by ether
and hemorrhage. After death in the case of one of these the liver was
quickly excised and cut in very small pieces and placed in an incu-
bator at body temperature. Portions were removed for estimation
of glycogen at periods of one hour each. We recognize that some
source of error is incurred by using weighed amounts of liver for the
values. This source of error might have been lessened by making
nitrogen estimations and using these, as Taylor has done, for the
standard of amount of substance taken. For our purpose, however,
as the results show, such precaution is unnecessary.

The following table depicts the results:

t (hours). C, (gm. glycogen 1 log. nat. = t (hours). C, (er. glycogen log. nat. &
t

in 100 gm. liver). in too gm. liver). ¢ G:
° (C 4.345) 5 2.924 .0343
I 3.826 0554 6 2.615 .0367
2 3.620 .0396 a 2.265 .O404
3 3.480 .03 21 8 25S .O417
4 3.100 .0366 9 1.800 0425

The process was most rapid during the first hour after death, then
slowed off until the seventh hour after death, when it began to in-
crease again. This increase is probably due to commencing putre-
faction. The results do not, however, show any evident decrease in
the activity of the ferment after the first hour.

4 Brown and GLenpinninc, T. A.: The journal of the Chemical Society,
1902, Ixxxi, p. 388.

19 Taytor, A. E.: The journal of biological chemistry, 1908, v, p. 315.

358 J. J. R. Macleod and R. G. Pearce.

In another experiment conducted in the same manner with the
difference that the liver instead of being cut into small pieces was left
intact the following results were obtained:

t (hours). ee ee ee = log. nat. e t (hours). Fe ae ae = log. nat. e
I 4.165 .04490 "6 4 3-230 .03883
2 3.866 .03084 5 3.186 03227
3 3-576 03697 6 3-033 103045

In this case there is a progressive decline in the constant.

We may conclude from the results that the post mortem process
when once established proceeds at a practically uniform rate for
several hours. =

At what period after death does the process set in, and how long does it
take to attain its maximal velocity? —It has proved a most difficult
task to obtain data from which these questions can be answered.
The periods of time intervening between the removal of the pieces of
liver must be so short and the observed changes in glycogen con-
tents are so small that after allowing for the experimental errors in-
volved in such determinations (see p. 346) there is considerable un-
certainty in the observations. It is only in animals killed without
previous anesthesia that these observations are of any value. Of the
several experiments which we have conducted there are only two,
on rabbits, which we will publish at the present moment:

Rassirt I.
t (min.). Ci. + log. nat 2
C,
° 10.800 4 sare
20 10.210 .0O120
40 9.466 .OO142
60 g.100 .00123
Rasesir II.
° 8.793 the
20 7.960 .00215
50 6.733 .00232

Controls taken from right

Studies in Experimental Glycosuria. 359

These preliminary results indicate that the process starts within
twenty minutes after death and that by this time it has attained its
maximal velocity.

The only studies of the same nature as the above which we can
find recorded in the literature are by Pavy,!’ Dalton,!® and Seegen.!®
The most important results were obtained by Pavy. In well-fed
rabbits this investigator found that if a portion of liver were re-
moved immediately after death by pithing, and instantly cooled in a
freezing mixture, it did not yield on analysis any more dextrose than
that usually found in the other tissues (up to 0.2 per cent). If a few
minutes were allowed to elapse before the portion of the liver was
removed, a great increase in dextrose (up to 1.29 per cent) was noted.
By leaving the liver until next day, a further increase (up to 3.68 per
cent) had occurred. The amount of sugar present in the liver on the
following day is of course quite unreliable as indicating the degree of
glycogenolysis which had meanwhile occurred, because considerable
glycolysis must have taken place. Pavy concluded that the glyco-
genolytic process must become greatly slowed after a few minutes
because 10 to 12 parts per 1000 sugar had accumulated during this
time out of a glycogen supply which could yield, say, 50 parts per
tooo. This would mean “that the whole [of the glycogen] would
disappear in about three quarters of an hour if the production of
sugar took place at the rate above mentioned.’ The question cannot,
however, be so cursorily dismissed, for it is clearly shown by our re-
sults that for several hours after death the process is still at its maxi-
mum velocity.

There is no doubt, as our observations testify, that the rate of post
mortem glycogenolysis varies considerably in different animals, even
when they are of the same species. These variations have prompted
us to investigate some of the factors which influence the post mortem
process.

¢. Analysis of some of the factors which determine the speed of post
mortem glycogenolysis. — There are two factors which are known to
have an influence on the rapidity of the glycogenolytic process in the
liver. These are (1) the amount of blood in the viscus, and (2) the

17 Pavy, F. W.: The physiology of the carbohydrates, London, 1894, p. 138.
18 Cf. BERNARD CLAUDE: Lecons sur le diabéte, 1877, p. 351.
19 SEEGEN, J.: Die Zuckerbilding in Thierkérper, 2 ed. (Berlin), 1900, p. 62

360 JY, R. Macleod “and ik. G.. Peatec:

connection of the liver with the nervous system. To study the rela-
tive importance of these factors we have proceeded as follows:

An Eck fistula was established in a specially fed anesthetized dog.
Portions of liver were removed from two lobes, and each of these por-
tions was further subdivided into three parts. In one of these the
glycogen was determined, another portion was placed in the incu-
bator at body temperature, and the third portion was placed in blood
in the abdomen of the animal. At the end of an hour the glycogen
content was determined in:

1. The portion placed in the incubator.
2. The portion placed in blood in the abdomen.
3. The liver left im situ in the animal’s body.

The liver left im situ was in some cases isolated from the nervous sys-
tem by cutting all hepatic nerves; in other cases these nerves were
left intact.

The results of these observations are given in Table IT.

The figures recorded in Table II are chosen from experiments
in which there was no uncertainty as to the accuracy of the re-
sults. The temperature of the incubator varied somewhat in dif-
ferent cases and the variations are noted. It is seen that the glycogen
disappeared in every case more rapidly in the portions of liver that
were bathed in blood in the abdomen than in the portions that were
placed in the incubator, and from which as much blood as possible
had been pressed out. This difference is of course readily understood.
It is due to the presence of glycogenolytic ferment in the blood. Ex-
actly similar results have been recorded by Bial.2° Pavy™ also ob-
served that the addition of blood to washed liver materially increases
the amount of sugar produced. As illustrating the importance of the
presence of blood in accelerating the process, the following experi-
ment is of interest. One half of the liver was washed free from blood
by perfusing o.g NaCl solution through the branches of the portal
vein running to it. Portions of liver were then removed from the
washed lobes and placed in 0.9 NaCl solution in the incubator, and
portions from the unwashed lobes were placed in blood at the same
temperature. These were incubated for one hour, at the end of which

20 Brat: Archiv fiir die gesammte physiologie, 1894, lv, p. 434.
at PAVY L0G. c2t., D: Tag.

Studies in Experimental Glycosuria. 301

time it was found that 17.5 per cent of the glycogen had disappeared
from the blood-free lobes, whereas 22.1 per cent had disappeared from
the lobes incubated in blood (Experiment 134).

TABLE II

PERCENTILE GLYCOGENOLYSIS IN ONE HOUR IN THE LIVER OF THE DOG, AS AFFECTED
BY THE AMOUNT OF BLOOD LEFT IN IT AND CONNECTION WITH THE NERVOUS Sys-
TEM (RESULTS GIVEN AS DEXTROSE).

Per cent glycogen disappeared:

Per cent
glycogen
at start.

, From portion
of liver
placed in in-
cubator.

B.
From por-
tion of
liver placed
in blood in

Cc.
From
liver left
in situ.

Difference
between
A&C

Difference
between
B&C

abdomen.

0.7857 21.4
(Temp. 39.5)
7.8

(Temp. 42°C.)

28.0

10.9425 10.0

*qno soAIOU
oye dazy

9.2300 10.1

121 5.0800 33.5
(Temp. 39°C.)

122 8.605

124 4.249 12.5
(Temp. 42-44)

125
130

2.240

sodrou st) edoFT

4.286

Average of last three .

Much more marked, however, are the differences between the gly-
cogenolysis in the amputated portions placed in blood and in the liver
left in situ. It is difficult to see how these differences can be depend-
ent upon differences in the amount of blood, for the portions placed in
the abdomen were thoroughly cut up so that the blood might come
in intimate contact with the liver tissue. The blood in the abdomen
was collected from the femoral artery and the pieces of liver were
placed in it before it became coagulated. It was thought that this |
difference might be due to nervous influence which would act on the
liver for at least some minutes after the portal circulation had been

362 JOS. RR. Macteod and fo °G. Pearce:

cut off. To test this hypothesis, certain of the observations were con-
ducted with the hepatic nerves intact, and others with these nerves
entirely severed. There was no difference in the results of these
two types of experiments; that is to say, the difference between the
glycogenolysis in the intact liver, and in an isolated portion kept in
blood in the abdomen, was just as great when the liver was discon-
nected from the nervous system as when these connections were
intact.

We must conclude, therefore, that the exaggerated glycogenolysis
which occurs in intact liver as compared with that occurring in dis-
connected portions of the same liver is not due to any influence of the
nervous system. It may be dependent upon the more thorough in-
timacy of contact between blood and liver cell existing in the intact
viscus as compared with that which is possible in the case of isolated
portions of liver placed in blood.

The practical importance of these results in connection with all
researches in which comparison is made of the rate of glycogenolysis
under different conditions cannot be too strongly emphasized. In-
deed it was while making observations on the influence of stimulation
of the great splanchnic nerve on post mortem glycogenolysis that -the
necessity of such observation became apparent to us.”

d. The influence of stimulation of the great splanchnic nerve on the rate
of post mortem glycogenolysis. — In order to demonstrate the influence
of nervous control over the glycogenic function, one of us in con-
junction with H. O. Ruh™”™ has already published experiments in
which the rate of glycogenolysis was compared in livers during stimu-
lation of the great splanchnic nerve, with that which occurs when
this nerve is not stimulated. The comparisons seemed to show a
more rapid glycogenolysis when the nerve was stimulated. It is
evident, however, from the observations above reported that a com-
parison between the glycogenolysis in different livers is not a very
reliable criterion on account of the considerable variations in speed
with which this process proceeds in different livers. It must also be
remembered in all these experiments that what we are really studying
is the rapid post mortem process, and it may well be that this is in
itself as rapid as can be, so that any nervous influence over it is en-
tirely masked. To control the previous results, another series of ex-

2 MACLEOD, J. J. R., and Run, H. O.: This journal, 1908, xxii, p. 397.

Studies in Experimental Glycosuria. 363

periments were therefore undertaken in which the above-observed
differences in the rate of glycogenolysis in intact and isolated livers
were compared with the same differences when the great splanchnic
nerve was stimulated.

The dogs were prepared as described above with the difference that
the great splanchnic nerve was stimulated at frequent intervals dur-
ing the hour intervening between removal of the portions of liver for
analysis.

The following table depicts the results:

TABLE III.

PERCENTILE GLYCOGENOLYSIS IN LIVER IN SITU DURING STIMULATION OF THE GREAT
SPLANCHNIC NERVE, COMPARED WITH THAT IN PIECES OF THE SAME LIVER PLACED
IN BLooD IN THE ABDOMEN.

Per cent glycogen disappeared:

No. of Per cent
glycogen at A. :
: B. Difference
sae From pieces of From intact between

liver placed in :
ddonant liver. A&B.

experiment.

1311 ‘ 17.9 30.2 1273
1334 : 28.6 50.1 Aes
135 : 14.2

Average

1 Blood pressure low, with slight rise on stimulation.
2 Blood pressure normal, marked rise on stimulation.

The average difference in these experiments between the glyco-
genolysis in the intact and isolated liver is practically the same as in
the previous observations where the nerve was not stimulated.

These results show that artificial stimulation of the splanchnic nerve
does not accelerate post mortem glycogenolysis in intact liver.— We are
aware that this conclusion is out of harmony with that contained in
the previous article referred to above, but when we allow for the
undoubted variations in speed with which glycogenolysis proceeds in

364 J. J. R. Macleod and R. G. Pearce.

different livers, even when these are treated alike, it is readily seen
that the method of procedure adopted in the earlier work is unreli-
able and uncertain. We recognized this while doing the previous
work and recorded the results without drawing any final conclusions
from them. Post mortem glycogenolysis proceeds with such rapidity
that the influence of nerve control is undemonstrable on it. So far,
therefore, the only means of demonstrating this control is in the in-
tact animal by making observations on the reducing power of the
blood before and during stimulation of the nerve.

CONCLUSIONS.

The percentage amount of glycogen in the different lobes of the
liver varies by about 5 per cent. This variation is partly due to er-
rors in the method of estimation (Pfliiger) and partly because of an
unequal amount of connective tissue and of blood in different por-
tions of the organ. These differences become much greater (a)
under ether anesthesia; (b) when the liver is left zm situ in the dead
animal.

The differences are not materially affected by absorption of carbo-
hydrate food from the itestine.

After death in an etherized animal there is usually, but not al-
ways, an acceleration in the rate of glycogenolysis; this varies in dif-
ferent lobes.

It has, so far, been impossible to determine the exact time of on-
set of post mortem glycogenolysis, but it is certainly well established
within twenty minutes after death.

Once established, post mortem glycogenolysis proceeds at a uni-
form speed for several hours after death, being dependent solely on
the amount of glycogen remaining in the viscus (temperature, etc.,
being constant). The value of K in the equation for reaction ve-
locity remains therefore practically constant for several hours after
death.

Post mortem glycogenolysis is much more active in intact than in
cut up liver.

In cut up liver glycogenolysis is much more rapid when the liver is
in contact with blood than when it is blood free.

Studies in Experimental Glycosuria. 365
The greater glycogenolysis in intact liver is not due to any influence
which the nervous system might have during the few minutes after
death in which it could still exert an influence.
Stimulation of the great splanchnic has no constant influence on
the course of post mortem glycogenolysis.

THE METABOLISM OF DOGS WITH FUNCTIONALLY
RESECTED SMALL INTESTINE.

By FRANK P. UNDERHILL.
(WitH THE CoLLaBorATION oF CHESTER J. STEDMAN anv JESSAMINE CHAPMAN.)

[From the Sheffield Laboratory of Physiological Chemistry, Yale University,
New Haven, Connecticut.]

4 fee necessity of removing varying lengths of intestine from
man has made imperative extensive investigations concerning
the influence of such surgical procedures upon nutritional processes.
This is particularly true with respect to the mechanisms involved in
digestion and absorption. For obvious reasons the experimental
demonstration of the effects incident to resection of portions of the
enteric tract has been made for the most part upon the dog and cat.
Of particular importance in this connection have been the observa-
tions recorded by Harlay,! Senn,? Trzebicky,? Monari,* De Filippi,”
Erlanger and Hewlett,® Flint,’ and others.§ From these investiga-
tions it may be stated that extirpation of portions of the small intes-
tine generally entails a decreased absorption of the nitrogenous and
fatty constituents of the food. The extent of lessened absorption
depends upon the relative length of intestine removed, and also upon
the period which has elapsed after the operation, that is, whether
compensation has been established. The composition of the food

1 Hartay: Proceedings of the Royal Society, London, (B), 1899, lxiv, p. 255.

2 Senn: Experimentelle Beitrige zur Darmchirurgie, Basle, 1892, quoted.

3 TrzeBIcKy: Archiv fiir klinische Chirurgie, 1894, xlviii, p. 54.

4 Monart: Beitrige zur klinischen Chirurgie, 1896, xvi, p. 479.

5 Der Fixrepr: Archives italiennes de biologie, 1894, xxi, p. 445.

6 ERLANGER and HEWLETT: This journal, 1901, vi, p. 1.

7 Fiint: Transactions Connecticut State Medical Society, 1910, p. 283. A
complete discussion of the earlier literature upon this subject both in connection
with the human subject and with the lower animals may be found in this article.

8 Lonpon and Dirriew: Zeitschrift fiir physiologische Chemie, 1910, lxv,
p. 213; CARREL, MEvyER, and LEVENE: This journal, 1910, xxv, p. 439.

366

The Metabolism of Dogs. 367

may play a significant rdéle, since it has been established that large
quantities of fat bring about a much poorer utilization of food nitro-
gen than occurs when smaller quantities of fat are ingested. Dimin-
ished utilization of fat is particularly noticeable in these experimental
animals. With respect to carbohydrate absorption the observations
are somewhat at variance, since in some instances it has been reported
that the feces held reducing substances, and in other experiments
none were found.

In the investigation to be reported study of three problems was
held in view, (1) the absorption of foodstuffs after functional removal
of varying lengths of small intestine, (2) a study of food absorption at
different intervals after the operation, and (3) the determination of
carbohydrate utilization under varied conditions.

EXPERIMENTAL.

Description of the animals employed. — The dogs used in these ex-
periments were placed at our disposal through the kindness of Pro-
fessor Joseph Marshall Flint, and for convenience will be designated
Dog A, Dog B, and Dog No. 12. In these three instances a portion
of the small intestine was short-circuited.

Dog A was a water spaniel of 8.3 kilos in splendid nutritive condi-
tion. Two weeks after the operation, when she came into our posses-
sion, January 26, 1910, the wound was well healed. The stools of this
animal were fairly well formed, and throughout the entire period of
observation diarrhoea was not once noticed. On autopsy about nine
months later it was found that the entire intestine was 412.5 cm. long
and that 162.5 cm. or 39 per cent had been short-circuited.

Dog B was a mongrel bitch weighing 13.5 kilos when she came into
our possession, January 26, 1910. The animal was in fair nutritive
condition, but suffered from persistent diarrhoea, discharging copious
liquid stools of exceedingly foul odor. On her entrance to the labora-
tory two weeks after the operation the wound was well healed. On
autopsy about nine months later measurement showed the entire
length of intestine to be 525 cm., of which 350 cm. had been short-
circuited, — 66 per cent.

Dog No. 12 was the animal called Dog No. 12 in Professor Flint’s
report. She was received into the laboratory January 8, 1909, al-

368 Frank P. Underhill.

most six months after the operation. At that time she weighed 7.4
kilos; a distinct loss of weight which was never regained. Diarrhoea
was persistent and the appetite was ravenous. On autopsy it was
found that of the 324 cm. of small intestine 235 cm. or 73 per cent
had been short-circuited.

Methods. —In general the methods followed were.those usually
employed in metabolism experiments in this laboratory. Urine was
collected in twenty-four-hour periods by catheterization. The water
content of the faeces was determined by drying them upon the water
bath under acid alcohol and preserving them air-dry. Carbohydrates
in the feces were estimated according to the method of Tsuboi.?
Food was given in two portions daily. One meal was at nine o’clock
in the morning, the other at four o’clock in the afternoon.

DESCRIPTION OF EXPERIMENT I.

In this experiment Dog A and Dog B were employed. For several
days previous to the investigation the animals had received an ade-
quate mixed diet. The experiment was planned to determine the
ability of these dogs to maintain nitrogenous equilibrium upon suffi-
cient mixed diets the composition of which was to be radically altered
at intervals with respect to the content of fat and carbohydrate. As
originally planned, each period of the experiment was to extend over
five days. Owing to fecal contamination of the urine of Dog B, it
was practically impossible to obtain urine and feces unmixed for five
consecutive days, with the exception of the first period. Accordingly
in the other periods balances have been made covering the consecu-
tive days on which the urine was uncontaminated. Although it was
out of the question at times to include the excreta in the balance,
food was given as usual and all other conditions remained constant,
so that when a balance_of less than five days is reported the results
are probably fairly representative of the animal’s condition through-
out the entire period. The periods of this experiment have been
designated Periods 1, 2, 3.

Diets. Dog A.— The diet for Period 1 consisted of 100 gm.
meat, 50 gm. cracker meal, 20 gm. lard, 10 gm. bone ash, and 150 c.c.

* Tsusor: Zeitschrift fiir Biologie, 1897, xxxv, p. 68.

The Metabolism of Dogs. 369

water, the total nitrogen content of which amounted to 4.39 gm.,
or .56 gm. nitrogen per kilo. The estimated fuel value was 505
calories, or 64 calories per kilo.

In Period 2 the diet was made up of 100 gm. meat, 75 gm. cracker
meal, 36 gm. lard, 10 gm. bone ash, and 230 c.c. water, containing
4.81 gm. nitrogen and 750 calories, or 0.62 gm. nitrogen and 96
calories per kilo.

The composition of the food in Period 3 was as follows: 10o gm.
meat, 100 gm. cracker meal, 10 gm. bone ash, and 300 c.c. water. The
diet contained 5.23 gm. nitrogen, and a calculated fuel value of 520
calories, or 0.69 gm. nitrogen and 67 calories per kilo body weight.

Further details concerning Experiment I, Dog A, may be found in
Tables I and III, pages 370 and 372 respectively.

Diets. Dog B.—JIn the first period this dog was placed upon a
diet consisting of 200 gm. meat, 80 gm. cracker meal, 30 gm. lard,
20 gm. bone ash, and 300 c.c. water. This diet contained 8.44 gm.
nitrogen and had an estimated fuel value of 837 calories, or 0.63 gm.
nitrogen and 63 calories per kilo body weight.

In Period 2 the diet was made up of 200 gm. meat, 120 gm. cracker
meal, 50 gm. lard, 30 gm. bone ash, and 4oo c.c. water, and contained
9.11 gm. nitrogen and 1180 calories (calculated), or 0.71 gm. nitrogen
and 93 calories per kilo.

The diet in Period 3 had the following composition: 200 gm. meat,
160 gm. cracker meal, 30 gm. bone ash, and 450 c.c. water. The food
contained 9.78 gm. nitrogen with an estimated fuel value of 875
calories, or 0.79 gm. nitrogen and 7o calories per kilo body weight.
For further details of Experiment 1, Dog B, see Tables II and III,
pages 371 and 372 respectively.

DISCUSSION OF RESULTS OF EXPERIMENT I.

Throughout the entire first experiment, Dog A maintained body
weight and furnished large positive nitrogen balances (see Tables I
and III). Nitrogen utilization was practically the same as that of a
normal animal upon a mixed diet. The effect of increasing carbohy-
drate and fat intake had little influence upon nitrogen equilibrium,
although fat utilization on this diet was somewhat diminished. A
further increase is to be noted when lard was entirely removed from

370 Frank P. Underhill.

TABLE I

EXPERIMENT I. — Doc A, 39 PER CENT INTESTINE RESECTED.!

Prrtop 1.— Foon: 100 cm. Meat; 50 Gm. CRACKER MEAL; 10 Gm. Larp; 10 em.
Bone AsuH; 150 c.c. WATER.

Urine. Feces.
Nitro-
Body : :
Date. ae gen in Weight
weight. “food. | Vol- Posey ann aK G ve Nitro- | Ether
ume. gen. | Moist ee tent: gen. |extract.
1910. kilos. gm. tice gm. gm. gm. per cent. gm. gm.
Jan. 29 7.8 4.39 160 1.95 Bi) 2ef 18

oe 7.7 4.39 160 3.06
ol ee | 4.39 180 San He 29.0 Se 2.66 | 11.96
Feb. 1 ia 4.39 100 2.70 36.5 22-5 38

SS ge thee | 4.39 100 3.21 59.6 | 38.0 36

Periop 2.— Foon: 100 cm. MEAT; 75 Gm. CRACKER MEAL; 36 Gm. LARD; 10 cm.
Bone AsuH; 230 c.c. WATER.

Feb. 3 Tal 4.81 200 3.21
7.7 4.81 200 2.93 Chee 46.7 39
rr | 4.81 200 4.08 5 ae 26.8 ons 3.68 | 235

4
5

G6 ffeil 4.81 200 3.03 40.8 21.0 48
7 ed | 4.81 200 3.43 53.8 28.7 46

Periop 3.— Foon: 100 cm. Meat; 100 cu. Cracker MEAt; 10 cm. Bone AsHu;
300 c.c. WATER.

Feb. 8 7.7 D229 230 4.26 36.8 17.9 ot
ae 7.6 5.23 240 4.02 55.6 26.0 53
= 10 7.6 5:23 240 3.90 ae she des 3.67 9.80
pas Voi 5.23 280 3.84 32.6 20.4 37
nelle Fe Bes: 310 4.26 97.5 43.7 55

1 Tn all experiments the urine showed an acid reaction to litmus.

The Metabolism of Dogs. 371

TABLE II.

EXPERIMENT I.— Doc B, 66 PER CENT INTESTINE RESECTED.

Pertop 1.— Foon: 200 cm. Meat; 80 cm. CRACKER MEAL; 30 Gm. LARD; 20 Gm.
Bone Asu; 300 c.c. WATER.

Urine. ; Feces.
Nitro-
Body : ;
Date : gen in Weight.
ht g
weig food Vol- bili ; meee Nitro- | Ether
hee gen. | Moist. oy tent. eee a
1910. kilos gm. c.c. gm, gm. gm. per cent. gm. gm.
Jan. 29 13.3 8.44 550 9.24 23.9 9.7 59
ou 13.0 8.44 350 6.90 93.8 SPAT 44

iD ot 13.0 8.44 420 5.40 | 104.7 50.2 52 3.35 | 29.26

Feb. 1 12.9 8.44 320 8.16 | 119.8 45.8 62
oo ad 12.7 8.44 310 8.16 73.2 36.8 50

Preriop 2. — Foon: 200 cm. MEAT; 120 cm. CRACKER MEAL; 50 cu. LArpD; 30 Gm.
Bone AsuH; 400 c.c. WATER.

Feb: 5 125 9.11 350 7.86 ae se Sus | |
30 12.6 9.11 410 8.24 | 153.5 81.5 47 |43.41 |15.66

ei | 12:6) 9.11 450. | 7.89 | 114.2 64.5 43

Periop 3.— Foon: 200 cm. Meat; 160 cm. CRACKER MEAL; 30 Gm. BonE AsH;
450 c.c. WATER.

Feb. 10 12.4 9.78 430 FBO | ESS.7 59.3 62

{3.7 5.48
aie 12.3 9.78 360 7.86 | 129.7 50.2 61

the diet and carbohydrate intake again augmented. Carbohydrate
utilization was perfect upon all diets of this experiment.

With Dog B there was a gradual but steady loss of body weight
upon diets which would have been entirely adequate for a normal
animal of the same weight. In the first period (five days) a negative
balance of 0.99 gm. nitrogen or minus 0.19 gm. nitrogen per day was

372 Frank P. Underhill.

obtained. During this period nitrogen of the food was utilized to the
extent of 87 per cent. The fat utilization was 85 per cent, while that
of the carbohydrate was perfect — that is, no trace of carbohydrate
could be demonstrated in the feces. In the second period (three

TABLE III.

SuMMARY. — EXPERIMENT I.

Doc A, 39 PER CENT

Nitrogen.
Excreta. : Balance.
Food.
Urine. Feces. Total. a d. Per day.
21.95 14.07 2.66 16.73 +5.22 +1.04
24.05 16.68 3.68 20.36 +3.69 +0.74
26.15 20.28 3.67 23.95 +2.20 +0.44
is Doc B, 66 PER CENT

42.20 37.86 5.33 43.19 —0.99 —0.19
PUPS: 23.99 3.41 27.40 —0.07 —().02
19.56 15.66 3.37 19.03 +0.53 +0.26

days) the dog was in almost perfect nitrogenous equilibrium, a total
negative balance of only 0.07 gm. or 0.02 gm. nitrogen per day being
obtained. During the period both carbohydrate and fat intake had
been markedly increased, nitrogen increase being slight. . In spite of
these increases the utilization of the three components of the diet re-
mained practically unchanged. The third period (two days) reveals
a positive nitrogen balance of 0.53 gm. nitrogen, or plus 0.26 gm.
nitrogen per day. No lard was fed during this period, but carbohy-
drate was much increased. The utilization of nitrogen was not quite
so good as in previous periods. . Fat utilization was greatly dimin-
ished. A portion of this apparent diminution may probably be ex-

The Metabolism of Dogs. 373

'
plained by the presence in the feces of ether soluble intestinal ex-
cretory products which in the absence of truly unutilized food fat
causes a distortion of the percentage utilization. Although the car-
bohydrate intake was twice that of the first period, no trace of sugar-

“of TABLE. Mi.

SUMMARY EXPERIMENT I.

INTESTINE RESECTED.

Fat (ether extract). Carbohydrate.
'

eT Food

2 Food. Feces. Utilization. (calcu- Feces. | Utilization.
, lated).

per cent. gm. gm. per cent. gm. gm.* per cent.

87 125.50 11.96 90 172 0) 100

84 207.55 27.35 86 273 0 100

86 29.55 9.80 66 364 0 100

INTESTINE RESECTED.

87 199.45 29.26 85
87 181.59 15.66 91
82 22.50 5.48 7

yielding substances could be detected in the faeces. A point of interest
in connection with the feces is the variable water content.

From these experiments upon animals with different lengths of
intestine put out of function shortly after the operation only small
differences can be detected in the ability of the two dogs to utilize
their food, although in one case, Dog A, only 39 per cent of the en-
tire small intestine was not functionating, whereas with Dog B 66
per cent was non-functional. Furthermore, notable increases in fat
intake appeared to cause little or no change in fat, nitrogen, or car-
bohydrate utilization. Large increases in carbohydrate did not
result in impaired utilization. The carbohydrate utilization was

374 Frank P. Underhill.

very much better than that of normal dogs upon practically the same
diets. For instance, unpublished experiments of Dr. Mary D. Swartz
make it evident that only 90 to 95 per cent of ingested carbohydrate
is utilized in the normal dog. Only in the case of fat can the utiliza-
tion be called poor.

TABLE IV.

EXPERIMENT II. — Doc A, 39 PER CENT INTESTINE RESECTED.

Pertop 1.— Foon: 100 cm. Meat; 50 cm. CRACKER MEAL; 10 cm. Larp; 10 cm.
Bone Asug; 150 c.c. WATER.

Urine. Feces.

Body
weight. Vol- | Total Water Ether
ume. | nitro- - con- is extract.

gen. tent.

1910. kilos. pe: per cent.

May 25| 10.5 4.43 75 3.77 . 43
“ 26| 10.4 4.43 80 3.70 47
So 2a val Ore. 4.43 65 3.28 34

PEeriop 2. — Foon: 100 cm. Meat; 75 Gm. CRACKER MEAL; 36 cm. LARD; 10 cw.
Bone Asuw; 230 c.c. WATER.

May 28 85 | 3.44 16
cee! 3.51 49
sete!) 3.42 41
eyes \; 3.28 43

June 1 3230 68

DESCRIPTION OF EXPERIMENT II.

At the completion of Experiment I the dogs were allowed to run in
large airy cages and were fed upon adequate mixed diets. As time
progressed, it became noticeable that Dog B developed an almost in-
satiable thirst. Diarrhoea was persistent, the stools discharged being
notably clay-colored. A constant but gradual loss of weight also

The Metabolism of Dogs. 375

occurred. On the other hand, Dog A appeared to be perfectly normal

and steadily gained in weight. The rest period for these animals

extended from February 12 to May 23. From this time until June 2,
TABLE V.

EXPERIMENT II. — Doc B, 66 PER CENT INTESTINE RESECTED.

Pertop 1.— Foon: 200 cm. Meat; 80 cm. CRACKER MEAL; 30 cm. Larp; 20 Gm.
Bone AsuH; 300 c.c. WATER.

Urine. Feces.

Body Peas oe orn Jeig | |
weight. | Total pee Water Nitro- | Ether

| nitro- con- |
"| gen. | Moist. tent. Gens) arere:

kilos. gm. c.c. gm. gm. = per cent. |
10.2 8.50 205. |) 7.83 137 64

10.4 8.50 150 | 6.39 223 70
10.4 8.50 175. | 7.78 140 58

10.2 8.50 190 | 7.37 166 64

10.2 8.50 150=- |. 7.05 165 64

Pertop 2. — Foon: 200 cm. Meat; 120 cm. CRACKER MEAL; 50 cm. LARD; 30 Gm.
Bone Asu; 400 c.c. WATER.

170 6.99
175 6.96
6.96 8.63
7.26
7.56

two periods of five days each were carried out upon diets practically
identical with those of Experiment 1. The food fed in these two
periods corresponded with that given in the first two periods of Ex-
periment 1. Owing to slight differences of nitrogen content of the
meat, the total nitrogen intake varied slightly from that in the previ-
ous experiment. See Tables IV, V, and VI.

376 Frank P. Underhill.

DISCUSSION OF RESULTS OF EXPERIMENT II.

During the second period of observation (Table VI) Dog A furnished
only positive nitrogen balances. This animal had 39 per cent of its
small intestine short-circuited. Fat utilization may be fairly com-

TABLE VI.

SUMMARY. — EXPERIMENT II.

Doc A, 39 PER CENT

Nitrogen.
Periods. Excreta. Balance.
Food
Urine. Feces. Total. Per period. | Per day.
1 13.29 10.75 1.60 LESS +0.94 +0.31
3.44 20.42 +3.98 +0.79

Doc B, 66 PER CENT

43.35 —0.85 —0.17
+1.72 +0.34

pared with that of a normal dog on a mixed diet and was better than
in Experiment I several months earlier. Increase of fat intake had
little if any influence upon utilization of any of the foodstuffs. Car-
bohydrate utilization remained perfect. The body weight of Dog A
was 7.8 kilos at the beginning of Experiment I, and at the end of Ex-
periment 2 had increased to 10.5 kilos — a gain of 2.7 kilos.

It is at once apparent from an inspection of Table VI, Dog B, that at
a period three months after functional resection of two thirds of the
intestine fat utilization was much lower than shortly after the opera-
tion. Increasing fat intake within somewhat narrow limits did not
markedly impair utilization of any of the foodstuffs. Carbohydrate
utilization was still perfect. Nitrogen utilization, however, was some-
what lowered and appeared to undergo a still further slight diminution

ct ae “ve

The Metabolism of Dogs. 377

by increase in fat intake. At this later period of observation the dog
furnished a slight negative nitrogen balance for the first five days and
a positive nitrogen balance for the second period. The body weight
at the beginning of Experiment I was 13.3 kilos and at the end of
Experiment II was 10.0 kilos —a loss of 3.3 kilos.

TABLE VI.

SUMMARY. — EXPERIMENT II.

INTESTINE RESECTED.

Fat (ether extract). Carbohydrate.
ive Food
A rae Food. Feces. Utilization. (calcu- Feces. | Utilization.
ion.
lated).
per cent. gm. gm. per cent. gm. gm. per cent.
87 75.48 5.88 92 109 0 100
86 206.20 13.22 | 94 273 0 100

INTESTINE RESECTED.

83 201.28 66.96 66 292 | 0 100
81 301.90 102.37 66 437 | 0 100

DESCRIPTION OF EXPERIMENT III.

The food received by Dog No. 12 was the usual mixture of raw
meat, cracker meal and lard which was fed several days previous to
the actual period of observation and in sufficient quantities to main-
tain a normal dog in nitrogenous equilibrium. Water was given ad
libitum. Preliminary trials demonstrated the separate collection of
urine and feces to be almost impracticable owing to the persistent
diarrhoea. To overcome this obstacle the animal was fed small quan-
tities of finely ground agar-agar with the food —a procedure which
resulted in the passage of stools which while not formed were also
not fluid. The number of defecations was just as great as without

378 Frank P. Underhill.

the agar, but the character of the stools was so entirely altered that
contamination of the urine was prevented. In all probability the
insoluble agar produced this change in the texture of the feces by
imbibing the water from the intestinal contents.

Utilization of nitrogen, fat, and carbohydrate. — In the first observa-
tion with this dog it was planned, to bring about nitrogenous equi-

TABLE VII.

EXPERIMENT III, Pertop 1.— Doc No. 12, 73 pER Cent INTESTINE RESECTED.

eS
2a
=]
o)

Date. 1909.
Body weight.
Nitrogen in
food
= 100.

nitrogen
Ammonia
nitrogen.
Indican
fehlings
sol.

£| Nitrogen.

oy
5

per cent.

83
84

co
~
Oo
—s
oo
~
oO

librium as nearly as possible and then to determine the utilization of
the different foodstuffs. To accomplish this purpose a few days
previous to the real observation the dog received a diet consisting of
200 gm. meat, 120 gm. cracker meal and 10 gm. lard containing 8.92
gm. nitrogen and furnishing approximately 775 calories, or 1.27 gm.
nitrogen and 110 calories per kilo body weight, amounts which would
be far in excess of the requirements for a normal dog of this size.
This diet was continued through the first day (January 14) of obser-
vation and was then reduced, since the animal appeared to have
difficulty for the first time in devouring these large amounts of food.
The new diet contained 200 gm. meat, 80 gm. cracker meal, 10 gm.
lard, and ro gm. agar. The nitrogen content amounted to 8.24 gm.
This diet was eaten readily.

During the four days of observation, the results of which may be
seen in Tables VII and IX, it is apparent that the dog was not in a
condition of nitrogenous equilibrium in spite of the previous ingestion
of large quantities of food, the nitrogenous balance for the four days

The Metabolism of Dogs. 379

being minus 2.80 gm., or minus 0.7 gm. nitrogen per day. Turning
to the utilization of nitrogen and fat, it may be observed that both
were poor, nitrogen being utilized to the extent of only 74 per cent fat

TABLE VIII.

EXPERIMENT III, Pertops 2 anp 3.— Doc No. 12, 73 PER Cent INTESTINE
RESECTED.

Pertop 2. — Foop per Day: 100 cm. Meat; 25 cm. GELATIN; 80 Gu. CRACKER MEAL;
10 cm. Larp; 10 cm. AGAR-AGAR. Tota NITROGEN = 8.24 GM.

Urine. Feces.

' Indican Weight.

Total fehl- Water

Nitro- Ether
gen. extract.

nitro- ess *|=5 === is ee cons
bens ey Moist. | Air-dry. ao

gm gm. gm per cent.

6.38 112 20 82
6.57 wy 48
7.49 51
10.10 18
6.72 24

— Foop per Day: 50 cu. GELATIN; 80 Gm. CRACKER MEAL; 10 GM. LARD;
10 cm. AGAR-AGAR. ToTAL NITROGEN = 8.96 cm.

8.91

to the extent of 72 per cent. These figures agree well with those of
Erlanger and Hewlett. In spite of the extremely foul odor of the
feeces the quantity of indican eliminated through the urine was not
excessive. The figures for ammonia nitrogen are high when com-
pared to the output of the normal dog. Another point of interest
was the rather large percentage of water contained in the feces of
this animal. Normally the water content of the air-dry feces rarely

380 Frank P. Underlull.

exceeds 75 per cent, whereas those passed by this animal had a water
content of 80 to 85 per cent (see also Table VIII). It is not unlikely
that the water content of the feces may have been increased by the
presence of agar, which would imbibe a certain quantity of water and
prevent its absorption. It is hardly probable, however, that this
is the sole explanation since the fluidity of the feces passed when

TABLE IX.

SUMMARY. — EXPERIMENT III, Doc No. 12, 73 PER CENT INTESTINE RESECTED.

Nitrogen. Fat (ether extract).

Excreta. Balance.

ae Utili-
Utili- Food. za-
Per Per | zation. tip
‘| period.| day. :

Urine. | Feces.

gm gm. gm per cent. per cent.

gm. | 7 . mng.
27.74 | 8.70 —2.80 | —0.70 74 67.60 72

37.26,:| 7219 —3.25.| —0.64|} 82 | 96.50 84

31.41 | 8.91 —4.48 | —1.12|} 75 | 59.20 58

no agar was given was such that a large water content was obvious.
By varying the amount of carbohydrate even up to four times the
requirement for a normal dog of the same weight, no change in the
utilization of this foodstuff could be observed. It has been demon-
strated repeatedly in this laboratory that a diet sufficient for a normal
dog of this weight may consist of 200 gm. meat, 30 gm. cracker meal,
and 25 gm. lard. Dog No. 12 received approximately this diet, then
the cracker meal was increased to 80 gm. and the lard reduced to
10 gm. Finally, the cracker meal was still further increased to 120
gm. In all cases carbohydrate utilization was complete.!? In these
experiments designed to test carbohydrate utilization agar was not
given with the food for obvious reasons.

THE INFLUENCE OF GELATIN FEEDING UPON THE ELIMINATION OF
URINARY INDICAN.

In a previous communication " it was demonstrated that the re-
placement of meat by gelatin in a mixed diet results in a diminution

10 Cf. FLINT: Loc. cit. 1 UNDERHILL: This journal, 1904-1905, xii, p. 176.

The Metabolism of Dogs. 381

in the excretion of indican in the urine of the dog. This observation
is in accord with the now well-established origin of indican. In Table
VIII (Periods 2 and 3) are given the results of observations made
with a view of determining whether a dog with a short-circuited in-
testine would behave in a manner similar to a normal animal when a
portion or all the meat of the diet was replaced by gelatin. It is ob-
vious from these figures that the substitution of gelatin for meat was
without significant influence upon urinary indican elimination. With
this animal only negative nitrogen balances were obtained. In the
second gelatin period nitrogen utilization amounted to 75 per cent,
which is comparable to the utilization obtaining when meat was fed
(see Table IX, Period 1). The fat utilization in the first gelatin
period (see Table IX, Period 2) when some meat was fed was much
better than when the meat was entirely replaced by gelatin. In the
first instance utilization amounted to 84 per cent; in the second
gelatin period (see Table IX, Period 3) only 58 per cent of the fat was
utilized. It would appear that meat in the diet of this animal had a
tendency to aid fat utilization.

SUMMARY.

From the foregoing observations it is apparent that as much as
39 per cent of the small intestine of a dog may be short-circuited
without causing significant detrimental changes in the utilization of
the various foodstuffs, and the animal may gain in weight. This.
statement is equally true when observations are made either at a
period shortly after operation or at a period several months later.

When as much as 66 per cent of the small intestine has been func-
tionally resected, the nutritive condition of the animal presents an
entirely different aspect. Under these conditions fat utilization is
particularly decreased and the dog displays a decided tendency to
furnish negative nitrogen balances. A small though steady loss of
weight is especially noticeable. Food utilization is in general ap-
parently much better immediately after the operation than at a later
period. In neither animal did a material increase in fat intake cause
significant change in the utilization of this or other foodstuff.

When about three quarters of the small intestine of the dog has
been short-circuited, food utilization for the most part is seriously

382 Frank P. Underhill.

impaired, at least at a period several months after the operation. This
is particularly true for fat utilization. Indican elimination through
the urine is not materially altered under these conditions by replace-
ment of meat in the diet with gelatin, an observation directly opposed
to that obtained with the normal dog.

The animal with a short-circuited intestine displays a greater
ability to utilize carbohydrate than does the normal dog. Even
though the carbohydrate intake may be much, in one case several
times, greater than the normal animal requires, carbohydrate utiliza-
tion is complete whether the test is made shortly after the operation
or months later. This observation may prove of practical importance
in the dietary treatment of the human subject who has undergone
extensive intestinal resection.

THE INFLUENCE OF THE PRECEDING DIET ON THE
RESPIRATORY QUOTIENT AFTER ACTIVE DIGES-
TION HAS CEASED.

By FRANCIS G. BENEDICT, L. E. EMMES, anp J. A. RICHE.

[From the Nutrition Laboratory of the Carnegie Institution of Washington, Boston, Mass.]
N laboratories where calorimetric measurements are impracticable
much use has been made of a study of the respiratory exchange as
a general index of the total metabolism and of the energy transforma-
tions in the body. Up to the present time practically all of the con-
tributions to our knowledge of the respiratory exchange in experiments
of short duration have been derived from experiments made with the
Zuntz apparatus by Zuntz and his co-workers. Since there is a great
difference of opinion as to the practicability of using a respiration ap-
paratus which employs a special appliance for breathing, such as a
mouthpiece, nosepiece, or mask, especially when the subjects are un-
trained, it is particularly fortunate that we have so large a number
of observations made by the most skilled manipulators of the Zuntz
apparatus, including Zuntz, Loewy, Mueller, Caspari, and particu-
larly Durig and Magnus-Levy. By means of these observations our
knowledge of the respiratory exchange with men at rest and under
different conditions of muscular activity has been much amplified.
Using the numerous methods for determining carbon dioxide in
air, experiments have been made by different investigators on the
carbon-dioxide excretion of man, since the measurement of this factor
presents no serious difficulty. As has been frequently pointed out,
however, the carbon-dioxide excretion in a given time cannot be con-
sidered an accurate index of the katabolism, since the amount of en-
ergy produced in the body for every gram of carbon dioxide exhaled
varies widely, depending upon whether the carbon dioxide is derived
from the carbon of protein, fat, or carbohydrate. For every gram of »
carbon dioxide resulting from the oxidation of protein in the body
383

384 Francis G. Benedict, L. E. Emmes, and J. A. Riche.

there are liberated 2.9 calories; for every gram of carbon dioxide re-
sulting from the oxidation of fat, 3.4 calories; and for every gram of
carbon dioxide resulting from the oxidation of carbohydrate, 2.57
calories.

The ordinary diet of mankind contains varying amounts and vary-
ing proportions of fat and carbohydrate, and it has been customary
to assume that under ordinary conditions the carbohydrate is for the
most part burned first. For periods following inanition this has been
completely disproved.1_ During periods when no food is given it is
clear that the combustion must of necessity be largely of body fat,
although here again the carbohydrate material in the body stored in
the form of glycogen may be such as to furnish by its oxidation a con-
siderable proportion of the total carbon-dioxide excretion in the course
of aday. Thus it was found in fasting man that as much as 181 gm.
of glycogen may be burned in the body in the first day of complete
inanition.?

Certain methods of studying the respiratory exchange take into
consideration the important factor of oxygen consumption, this being
particularly true of the Zuntz apparatus. When the energy per gram
of oxygen absorbed is considered, it is found that the differences ap-
pearing in the carbon dioxide, and depending upon the character of the
material katabolized, do not exist in the case of oxygen. With oxygen
the calorific value of 1 gm. remains nearly the same irrespective of
whether the oxygen is used to burn protein, fat, or carbohydrate.
Consequently, in lieu of calorimetric experiments, an accurate meas-
urement of the oxygen intake is taken as a reasonably exact index of
the total energy transformations during the time of an experiment.

This index of the total katabolism is, however, by no means so
satisfactory as an apportionment of the metabolism between the pro-
tein, fat, and carbohydrate, — an apportionment that may be made if
one knows the nitrogen excretion during the period of experimenting,
the carbon-dioxide exhalation, and the oxygen consumption. Given
these three factors, it is possible, by means of the method devised by
Zuntz, to compute the energy transformations by so-called ‘indirect
calorimetry.’ It is not possible here to enter into a discussion of the

1 BENEDICT: Carnegie Institution of Washington, Publication No. 77, 1907,
Pp. 532.
2 BENEDICT: Loc. cit., p. 464.

Influence of Preceding Diet on Respiratory Quotient. 385

comparison between direct and indirect calorimetry in short experi-
ments, as an extended series of experiments on this point is now in
progress in this laboratory.

Entirely aside from the energy transformations, however, one of
the chief advantages of obtaining the respiratory exchange by a
method which permits an estimation of the oxygen consumption in
addition to the carbon-dioxide production is that it enables the ap-
portionment of the katabolism between carbohydrate and fat. In
studying questions regarding the relative efficiency of fat and carbo-
hydrate as a source of muscular work, such an apportionment is of
prime importance, as has already been demonstrated in discussing the
transformations of material during diseases resulting from disturb-
ances of metabolism, such as diabetes.®

The calculation of the total katabolism expressed in terms of pro-
tein, fat, and carbohydrate may be carried out with reasonable ac-
curacy when the output of nitrogen and carbon dioxide and intake
of oxygen are known. A general picture of the character of the me-
tabolism, showing the nature of the material burned inside the body
without a quantitative estimation of the exact'amounts burned, may
be obtained by a simple inspection of the ratio of carbon dioxide to
oxygen, namely, the respiratory quotient. When carbon in carbohy-
drate is burned, the volume of carbon dioxide resulting from the
combustion is exactly equal in volume to the oxygen absorbed, and
hence the ratio of carbon dioxide to oxygen is 1. When fat is burned,
a very much larger volume of oxygen is consumed than of carbon di-
oxide eliminated, since a portion of the oxygen is used to oxidize the
hydrogen of the fat molecule and form water. With a diet rich in
fat, therefore, the ratio is usually found to be not far from 0.71. These
calculations may be approximately verified by feeding animals pure
fat and pure carbohydrates. It is a little more difficult to feed pure
protein, and the calculation of the theoretical respiratory quotient for
protein varies somewhat with different writers. It is usually accepted,
however, as being not far from 0.81.4. Consequently, if the respira-
tory quotient is in the neighborhood of 1, it is obvious that the katabo-
lism must be largely of a carbohydrate nature. If it is in the neigh-

3 BENEDICT and JosLIn: Carnegie Institution of Washington, Publication No.
136, IQIO, Pp. 203.
4 Benepict: This journal, 1909, xxiv, p. 350.

386 Francis G. Benedict, L. E. Emmes, and J. A. Riche.

borhood of 0.7, it is likewise clear that it must be largely of fat. In
practically all of the diets which man is accustomed to use, protein
rarely forms more than 15 per cent of the total katabolism; hence, for
many comparisons, the protein katabolism can be neglected, and the
general conclusions drawn that respiratory quotients in the neighbor-
hood of 0.7 to 0.75 indicate a predominantly fat katabolism, and
respiratory quotients between o.9 and 1 indicate a predominantly
carbohydrate katabolism.

With these marked variations in the respiratory quotient, depend-
ing upon the nature of the material katabolized, it was soon seen that,
to obtain results of any significance in a study of the respiratory ex-
change, the experiments should be made a sufficient length of time
after the last meal to eliminate the question of the increased katabo-
lism invariably following the ingestion of food. This increase in
the katabolism is most marked after protein is eaten, but a positive
increase is also observed when carbohydrate and fat are ingested. As
a result of many experiments made by Magnus-Levy and others, it
has been the consensus of opinion that twelve hours after the last
food is taken the active work of digestion has ceased, and the katabo-
lism is therefore assumed to be constant for each individual under like
conditions of body activity. A distinction has been made between
the value obtained twelve hours after the last meal and the value ob-
tained after a fast of twenty-four or more hours, the first value being
the so-called ‘‘Niichternwert”’ of the Germans. As a matter of fact,
however, a large number of experiments on fasting men *® have shown
that the katabolism is far from constant on the first two days of com-
plete inanition, and that only on the third day is constancy approxi-
mated. Hence there must be a regular transition from the time the
last food was taken until complete inanition takes place, the first step
being the active process of digestion, and the second when body
material (fat, and particularly glycogen) is drawn upon. After the
glycogen has in large part been reduced there is, then, a period of rela-
tive constancy in the katabolism, for it is chiefly of fat, with a small
draft upon the remaining store of body glycogen and a relatively
constant protein katabolism throughout the whole period.

An examination of the literature of experiments made twelve hours

5 BENEDICT: Carnegie Institution of Washington, Publication No. 77, 1907,
pp. 456 ef seq.

Influence of Preceding Diet on Respiratory Quotient. 387

after the preceding meal shows that the respiratory quotient may be
very considerably higher on one day than on another. Again we
find series of experiments in which the respiratory quotient was re-
markably constant, not only from day to day, but also from season to
season. Inasmuch as in the majority of cases the respiratory quo-
tient is relatively constant from day to day and from month to month
with the same individuals, it has been the custom of many writers to
determine this value once for all for a base line and consider it as an
average value in subsequent experiments. For example, when study-
ing the influence of the ingestion of food or muscular work upon ka-
tabolism, the values found twelve hours after the last meal are de-
ducted from the values found under the conditions of the experiment,
and the difference ascribed to the effect of the ingestion of food or to
the effect of work as the case may be. Since these average values
play an important réle in many subsequent calculations, it can be
seen that it is of the utmost importance to know to what extent they
represent the true average, what variations may commonly be ex-
pected, and, if possible, to what the variations are due. A study of
these average values has accordingly been made in this laboratory,
and the results are here presented and discussed.

GENERAL PLAN OF THE EXPERIMENTS.

The primary object of this study was to note the effect of the pre-
ceding diet upon the respiratory exchange twelve hours after the last
meal was taken. Consequently, the earlier experiments were so de-
signed as to have the study of the respiratory exchange made approxi-
mately twelve hours after a meal ® which had consisted either in large
part of carbohydrates or, on the other hand, of relatively small amounts
of carbohydrate. The diets, then, were distinctly carbohydrate-rich
at one time and carbohydrate-poor at the other. In order to eliminate
the possible influence of an excessive ingestion of protein, the protein
in the diet of the preceding meal was in no instance materially in-
creased, although care was taken to have the carbohydrate-poor diet
contain enough fat to keep up the energy requirement for the day.
As a rule, the meal on the evening preceding the experiment was usu-

6 In no case was the experiment made less than twelve hours after the last
meal.

388 Francis G. Benedict, L. E. Emmes, and J. A. Riche.

ally eaten about six o’clock. The subject then came to the laboratory
the next morning without eating, and immediately lay down upon the
couch, preparatory to a series of experiments.

This method of studying the influence of the preceding diet has a
number of disadvantages. In the earlier experiments the subject was
simply instructed to make the last meal of the day before either car-
bohydrate-rich or carbohydrate-poor. No attempt was made to con-
trol the diet exactly and, indeed, subsequent calculation based upon
the reported amounts eaten shows that there were enormous varia-
tions in the amount of carbohydrate ingested, — variations that were
not suspected by the subject. Experiments subsequent to October 1,
Ig10, were made on a somewhat different plan. In most of these the
diet on the evening before was controlled with reasonable rigidity. In
a number of cases actual weighings of all food materials were made,
and the diet so apportioned as to secure approximately exact amounts
of carbohydrate, large or small.

It is obvious that the difficulty was to be found more particularly
in the meals in which low carbohydrates were desired. Usually it
could be assumed with every guarantee of certainty that the subjects
would eat large amounts of carbohydrate if so instructed, but the
proper selection of a carbohydrate-low diet was not easy for inexperi-
enced individuals. The admirable plan instituted and followed by
Atwater in his researches at Wesleyan University in Middletown,
Conn., of having all experiments preceded by a three-day period in
which the same diet was ingested as on the experimental day, would
have been followed here had the time and the expense allowed it.
Under these conditions and these only can we be certain of the condi-
tion of the body and the plane of nutrition prior to the experiment.
This is particularly the case if it happened, as it is more than likely
that it did happen, that the subjects had been on certain days engaged
in active muscular work and perhaps had eaten an insufficient lunch
at noon owing to pressure of other work, thus necessitating a heavy
draft upon body glycogen. Furthermore, although there might have
been a large amount of carbohydrate ingested in the evening meal,
it is not at all improbable that a considerable proportion of this may
have been stored as glycogen, as has been suggested by Johansson.’

7 JOHANSSON and HELLGREN: Festschrift fiir Olof Hammarsten, Upsala, 1906,
vii, p. 8.

Influence of Preceding Diet on Respiratory Quotient. 389

On the whole, however, these experiments were designed to be so ex-
tensive and to include so many individuals as to eliminate these acci-
dental possibilities as far as possible. The subjects for the most part
were living upon diets that were reasonably uniform and regular.

METHODS.

The respiration apparatus used in this investigation was devised by
one of us 8 and had been fully tested prior to these experiments. With
this apparatus the subject, lying upon a comfortable couch, breathes
through two nosepieces of special design into a closed air circuit. By
means of a small ‘rotary blower the current of air in the ventilating
pipe is kept moving at the rate of 35 litresa minute. After the expired
air enters the air current, it is passed through a set of absorbers, the
first containing sulphuric acid, which removes the water vapor given
off from the lungs; the second containing soda lime, which absorbs
the carbon dioxide in the exhaled air; and the last, sulphuric-acid,
which removes the water absorbed by the air current as it passes
through the moist soda lime. Since this dry air would be irritating to
the mucous membrane in the nose and throat, moisture is added by
passing it through an air-moistening device. The deficiency in oxy-
gen is also made up by adding oxygen from a cylinder of the gas. The
air is then in approximately normal condition, though free from car-
bon dioxide, and is returned to the subject for breathing.

The air in the ventilating current is kept at atmospheric pressure
by means of a tension equalizer, in which a rubber diaphragm falls or
rises as air is withdrawn or added to the system. Any loss in volume
due to the absorption of oxygen by the subject is made up by the oxy-
gen admitted from the cylinder, so that when the experiment is con-
cluded, the volume of air in the system is the same as the initial
volume. The amount of carbon dioxide given off by a subject may be
determined by noting the difference in weight of the carbon-dioxide
absorber before and after an experiment, also of the sulphuric-acid
container in which the water from the soda lime is absorbed, and find-
ing the algebraic sum of the two. The loss in weight of the oxygen
cylinder, corrected for the small amount of nitrogen in the oxygen,

8 BENEDICT: This journal, 1909, xxiv, p. 345.

390 Francis G. Benedict, L. E. Emmes, and J. A. Riche.

gives the weight of the oxygen consumed by the subject. By this
means the carbon-dioxide excretion and oxygen absorption are rapidly
and accurately determined without the necessity of using a complex
gas analysis apparatus.

A sample of the form used for recording the data of an experiment
is reproduced here to give an idea of the simplicity of the calculations
required to obtain the results. As will be seen, at the head of the
form are recorded the number of the experiment, the date, name, or
initials of the subject, and time of beginning and ending of the experi-
ment, as well as its duration. In the left-hand division the records
are made of the weights of the carbon-dioxide absorber, the water ab-
sorber, and the oxygen cylinder, together with the simple calculations
for obtaining the weights of the carbon dioxide excreted and oxygen
consumed. The calculations necessary to reduce these weights to
volumes by means of the factors 0.5091 and 0.6998 are recorded at
the right below the respiration record, also the calculation of the
respiratory quotient. In the two divisions at the bottom the calcula-
tions for obtaining the amounts of carbon dioxide excreted and oxy-
gen consumed in cubic centimetres per minute are given, while space
is reserved below for any miscellaneous observations which may be
worthy of record. The records of the pulse and respiration rates are
given in the upper divisions. The time required for ten respirations
is recorded as well as the rate of respiration per minute, and the
average respiration and pulse rates.

STATISTICS OF EXPERIMENTS.

Although the statistics with regard to the body weight, age, etc.,
have no particular value in a study of the influence of the preceding
diet on the respiratory quotient or on the character of the katabolism,
they are included in Table I, as the experiments may be of interest
in other connections.

Experiments with seven subjects are included in this report. With
six of these, the experiments were made with the respiration appara-
tus; four of these six subjects were also subjects of respiration calo-
rimeter experiments which are included in this report, together with
respiration calorimeter experiments made with one other subject.

Influence of Preceding Diet on Respiratory Quotient. 391
RESPIRATION EXPERIMENT NO. I.
May 6, 1909.
Start = 5... “40.007Ar M.
Stioject, sks: {ind 4 Oa5.02' A. M.
Duration . 15 min. 2 sec.
mee; Pere MINY 2 Ls ee GxtiaKesp) 1050 Seer Sess 12
Pee per min, es a 6 |) Resp ieasoisee. oes! 12
Peeper ming)... wae Go: | Resp: tor 46: sece=. 2 3 13
BRMEMACEE |S. Sand 0515 Wee Wacoal 63 ELAR Cia ae i ci, aie eee = 12
- ee ee
H.SO,4 End De osrO! | ALOR. SOU 5 5 5 ab ay te 70680
No. 19 SEAEEOS i), atsamy EG EO =
Log. total CO. = 74974
1.81
Se Be End 5250.76 Log. vol. CO. =
| VOL. eae EA ae 45654
G a ieee OR e40:05 ; ;
3.81 TLagtyval.' Oo = 0.2. 52072
1.81
Total CO. 5.62 Remere Os Sto Gi aa 92682
Ozcyl. ee a hata SHE 22
No. 27427 (End 7340.40
4.82 EOS xO Sees hyena eat sack 844096
N. corr... . sOTOE I Kase total Oars se ah 68476
Total O2 4.839
Respiratory Quotient: 84. Tor vole O2=" 3544s 52072
Eogwol. CO, = . 2... s A5o5a° Loe-vol/ O22 52072
Log. time in min. = 17705 | Log. time in min. = 17705
Log. c.c. COe per min. = 27949 | Log.c.c.O2permin.= . 35267
eCe,cOsyperminy= . 2%. EOO))|. @.¢, Op per min 9 5: 225
EEE co ticteeeel ae ill Ae edo an Ra Re i eI ape Ri BB bP

All of the subjects were engaged in scientific work in the laboratory,
and were thoroughly familiar with the methods of experimenting and
the object of the research. Their intelligent co-operation was thus

assured.

392 ~Francis G. Benedict, L. E. Emmes, and J. A. Riche.

The experiments reported in this article are of two kinds. As pre-
viously stated, most of them were made with a respiration apparatus
and only the carbon-dioxide output and oxygen intake were deter-
mined. To supplement these experimental researches with the res-
piration apparatus, a number of experiments with a respiration

TABLE I.

Statistics OF AGE, HEIGHT, AND AVERAGE WEIGHT OF SUBJECTS.

Weight with-
out clothing.

Subject. ; Height.

cms.

183
182
172,
LZ

calorimeter, which were made for another purpose, have been here
included. In certain calorimeter experiments it is perfectly feasible
to combine a test on the influence of the preceding diet on the respi-
ratory exchange with some other special subject for investigation.
Accordingly, in a number of such experiments, the subjects were pro-
vided with carefully controlled diets containing either large or small |
amounts of carbohydrate. These respiration calorimeter experiments,
however, are presented only in abstract, as the values for the respira-
tory quotient alone are of significance in this connection.

The details of the respiration experiments with the different indi-
viduals are given in Table II, and those with the respiration calorime-
ter in Table III. The respiration experiments were made primarily
for this study, and each daily series represents comparable experi-
ments under like conditions of muscular activity and preceding diet.
The experiments were usually of fifteen minutes’ duration and fol-
lowed each other at intervals of approximately thirty to forty-five
minutes. While with certain of the subjects the uniformity in the

Influence of Preceding Diet on Respiratory Quotient. 393

oxygen consumption from experiment to experiment was not so con-
stant as could be desired, thereby indicating slight differences in
muscular activity, it was believed that the large number of experi-
ments here made, many of them on the same day, would more than
offset the differences in muscular activity, and the average value for
the whole day must represent very nearly the actual conditions of
katabolism existing at the time the experiment was made. This is
particularly the case in this discussion, as the respiratory quotient is
the only factor of significance, and it has been maintained by Zuntz
and his associates? that even moderately severe muscular work does
not alter the respiratory quotient.

During the experiments the subjects were lying quietly upon a
couch, and the greatest muscular relaxation was insisted upon. A
close examination of the figures in Table II shows, however, that
frequently the first experiment of a series gave results distinctly
abnormal as compared with the subsequent experiments. This is
explained by the fact that there may have been a somewhat larger
carbon-dioxide exhalation due to irregular respiration, and that in
subsequent experiments these irregularities disappeared.

NITROGEN EXCRETION.

In experiments in which the diet is undergoing marked alterations
it is necessary to demonstrate rather than to assume that the protein
intake on the day before is not abnormally increased or decreased,
as such change would influence the respiratory exchange and mask
any effect of a variation in the amount of carbohydrate in the preced-
ing diet. Accordingly a study was made of the urine passed at the
time of the experiments, that is, for the period representing the twelfth
to the eighteenth hours after the last meal. The nitrogen excretion
per hour in these experiments has been calculated from the results of
the Kjeldahl determinations and presented in Table IV. It is here
seen that, considering the subjects were for the most part subsisting
on uncontrolled diets, there is a remarkable uniformity in the nitrogen
excretion per hour in practically all of the experiments with the same
individual, thus showing conclusively that there could not have been
any marked alteration in the protein proportion of the diet the day

9 HEINEMANN: Archiv fiir die gesammte Physiologie, 1901, Ixxxiii, p. 441.

*yuatjonb
Aioyerdsayy

394 Francis G. Benedict, L. E. Emmes, and J. A. Riche.

€hZ 7ST 850 | sz SZ £60 | €8T TLT
bee €8T. | (oT) S60 | ¢92 iA 760 | S8T PLT
Lt “os | 7p Ae S 760 | £67 SEZ a 181 "8
Lt 102 760 | 62 87Z 960 | LST 6LT
97Z LLY (ystH) BuO | ZS7 Tz =| (ystH) 760 | S8T OLT
St O6T g AIL 57 960 | OFZ oez | 77 ARI 6 680 | Z8T 991
€h7Z ee S60 | FZ ££Z EO10="|- C85 Sol
OFZ 61Z | (q3tH) £60] 7SZ ££Z 160 7ST SOT
FEZ 022 J Ae ¢ 130 | 6F2 917 060 | £6 ELT (ystH)
SEZ LST 83'0 | 9Sz $272 #80 | 602 OLT | HT ‘AON ir
SSZ LOZ (ystH) 680 LSZ 877 (y3tH) OL
OFZ L1@ | 6¢ ‘adv 6 160 | Tz Pee | O02 API 8 PL0 | ZSZ SST
6£7 O1Z 16°0 7£7 O17 810 L¥Z raat! (MoT)
9772 Sst (qs1H) 68'0 SET 607 910 LtT L8T | ST ‘AON Zz
FEZ Loz | Lz “Ady T S0'| 7Sz 9072 680 | 982 672
6061 cr fei 9 83'0 | 6£Z 602 160 | SZ T£Z
672 8ST 8'0 | 962 L0Z (MoT) 980 | £97 97Z
61Z | #98T 830 | ‘8&2 00Z | eT Ae l 460 | OFZ FEZ
97 061 (ystH) €30| . £57 FIZ 160 | £82 677
SEZ POL | OF “AON C $80 Sh $07 68'0 €SZ S7Z (qstH])
0SZ 661 (Mo’T) if €3'0 T£Z 161 (43TH) 9 130 8SZ S7z_ | TT ‘AON I
S74 16E | 81 AON | “DTA Te) | con 16r | OLAV | “AAT || FBO | 652 81Z Ol6t |"a'D ‘A
‘oo ‘20 2°9 *o" a0 oo
ct ~) oO o iS)
wn Q © p wn (@) si rac} n QD Oop
=O Com) Bao on (e) a) eo ep) a(S) go oct n
peo iehS| 8h | eee ee ake leks] oF | 288] oF |e bS],83| of | p82
Bee |BS5| Pub | Fas | SU SSS 1588 | eo | Fee] St [Bee |S) ooh | fae
Ege isoa| So | SRG | as |eor iso] 8G2)/ 559 | SS [ECB leod| age eS
Pogifos| a2 |88e | Fs | Foe Fes) “se | Bee | BS |Foa iss] “ae | Ree
os ae eae 2 St sey 5 4. Bp 7 Neh pov As: stall r4. ae te
ee ek Siege ee) ee peel alee See ‘:

(‘SQLVaVddy NOILValdsay)
, ‘SALVAdAHOPUV) NI MOT XO HOI SLAICG] ONIMOTIO“ GOOY LAOHLIM SINAWIMAdXY NI FONVHOXY AYOLVAldSAyY

Il WIavL

Influence of Preceding Diet on Respiratory Quotient.

6L°0
t3'0
08°0

LLE
8h
897

(ystH)
J ounf

LLY
Tol

OST
sot

(MoT)
Q] “AON

396 Francis G. Benedict, L. E. Emmes, and J. A. Riche.

before. Hence the effect of the protein ingested in the last meal can
be considered as negligible in this discussion. An examination of the
figures shows that while there may be a general tendency for the

TABLE III.

RESPIRATORY QUOTIENT IN EXPERIMENTS WITHOUT Foop FOLLOWING Diets HIGH oR
Low IN CARBOHYDRATES. ' (RESPIRATION CALORIMETER.)

High 89 a. High
High 88 High

Low .80 | Low
High 92 Low
High : High
Low j High

Low P 5 Low

Low 4 Low

nitrogen excretion to be somewhat lower following a high carbohy-
drate diet, nevertheless there are so many irregularities that no general
deduction on this point can be drawn.

DISCUSSION OF RESULTS.

The most important comparison to be made in connection with
these experiments is the relationship between the amount of car-
bohydrate in the last meal of the preceding day and the respiratory
quotient on the morning of the experiment. To make this compari-
son all the more clear, the results of all the experiments are brought
together in Table V in such manner as to show which days were pre-
ceded by a high carbohydrate diet and which by a low carbohydrate
diet. In this table the average respiratory quotient for each experi-
mental day, both with the respiration apparatus and with the respira-
tion calorimeter, is given. It is clear that the respiratory quotient

Influence of Preceding Diet on Respiratory Quotient. 397

TABLE IV.

NirrocGen ExcreTION PER Hour IN EXPERIMENTS WITHOUT Foop FOLLOWING DIETS
HicH or Low IN CARBOHYDRATES.

Carbo- : Carbo- :
Subject. | Date, | Hydtates | Ntreton |Subject. | Date. b¥StBteS | excretion
meal. DEF hour. ein eee hour.
1910 gm. 1910 ae a
eG: b: 4) Nov. 11 High 400 L. E. E. | Nov. 26 High AA4L
TAS \ Lew: || 408 «29 | Low | .567
fee. C. 1 Oct. 27 Low 450 Weems Low 450
aes High 472 ae: Low 456
Nov. 3 Low Sit V.G. | Oct. 24 Low 462
SO it) High 318 a 26 High .316
NY) High 383 Nov. 4 Low 324
Dec. 6 Low 481 ere ih High 274
pee Cx ae High 531 TS Low 487
Mar. 23 Low 494 OO High 306
July 12 | High 387 ae hoe: Gratien: 458
Marie High 358 ae 71S, High 464
16 Low 454 May 1 High 488
Wim fa *o8 | High 625 « 6 | High 383
130 High 462 Se We Low 554
May 3 High 448 Ze al High 473
con ee High 471 fom: High 494
ct) High 480 cary al High 449
eels Low 387 599 Low 550
1910
20 High 515 Miss S. | Apr. 4 Low 257
DD, High 538 foe te Low 431
June 1 High 531 ores High 347
ier D High .658 June 21 High .300
AG High 478

398 Francis G. Benedict, L. E. Emmes, and J. A. Riche.

which was determined twelve hours after a meal consisting of a diet
rich in carbohydrates was in general higher than when the last meal
contained but a small amount of carbohydrates.

TABLE V.

AVERAGE RESPIRATORY QUOTIENTS IN EXPERIMENTS WITHOUT FooD FOLLOWING DIETS
HicH or Low IN CARBOHYDRATES.

High carbohy- | Low carbohy- High carbohy- | Low carbohy-
drate. drate. drate. drate.

Subject. lies
Date. R. Q. Date: »||RoO3) Daten eG:

1910 1905 1909
Nov. 11 76 .E.E.| June 9] 0.84

Oct. 31 85 0.80
Nov. 8 85 0.86

Nov. 22 79 “iG: 0.90 | Oct. 24
1909
Feb. 4 .80 0.94 |Nov. 4
1910
July 12 83 0.83 | Nov. 18
1909
Nov. 14 Ae Re 0.86 | May 12
1909

.| Apr. 26 0.83 | “ 29

1910

eS) 0.93 | June 2
0.84

0.84
0.85
0.86
0.85
0.91

A general average for each subject is given in Table VI. In this
table the method of averaging may be open to criticism, as with two
of the subjects, L. E. E. and J. R., an unequal number of experiments
was used in obtaining the average values, the experiments following
a high carbohydrate diet exceeding those which followed a diet low in

Influence of Preceding Diet on Respiratory Quotient. 399

carbohydrates. With the other subjects approximately equal numbers
of experiments were made on each nutritive plane, and the general
averages of the table are therefore probably not far from representing
the actual conditions at the time of the experiment.

In making experiments with so complicated an organism as the
human body it is extremely difficult to foresee the exact conditions

TABLE VI.

AVERAGE RESPIRATORY QUOTIENTS WITH DIFFERENT INDIVIDUALS IN EXPERIMENTS
WITHOUT Foop FOLLOWING Diets HicH or Low IN CARBOHYDRATES.

Respiratory quotient. Respiratory quotient.
Subject.

Subject.

ee eee
High carbo- | Low carbo- High carbo- | Low carbo-

hydrate. hydrate. hydrate. hydrate.

89 ; oy Gre oto 89 83
86 79
88 82

for experiments in all instances and thus arrange ideal preliminary
periods and preliminary body conditions. In all probability a serious
error was made in these experiments in not insisting upon a longer
preliminary plane of nutrition. We have here, therefore, only the
general impressions derived from a large number of experiments, and
can explain only inadequately the several anomalous experiments
which stand out so prominently in the series.

The amounts of carbohydrate ingested with the last meal on the
days with a rich-carbohydrate diet were in many instances very large,
occasionally over 300 gm. Since, however, the subjects were accus-
tomed to eat the hearty meal of the day at night, this is not an exces-
sive amount of carbohydrate, and not more than might have been
taken at many meals, since 300 gm. corresponds to about 1200 calories,
or from one third to one half of the daily requirement. In certain
instances the amounts of carbohydrate were from 400 to 500 gm. It
is possible, therefore, that under these conditions we may have again
to deal with a delayed carbohydrate absorption out of the intestinal
tract, and it may be that the active digestion had not ceased, although

400 Francis G. Benedict, L. E. Emmes, and J. A. Riche.

these experiments were for the most part made not twelve, but four-
teen hours after the last meal.

The possibility of the high respiratory quotient after a high carbo-
hydrate diet being due to the delayed absorption and combustion of
carbohydrates in the alimentary tract, is somewhat difficult of proof.
In fact, the evidence is rather against this theory, since in certain ex-
periments, namely, with T. M. C., where a high carbohydrate diet
was given, the high respiratory quotient continued throughout rather
a lengthy series of experiments which lasted a good many hours.
There was apparently no tendency for the respiratory quotient to
fall rapidly, as would be expected at the termination of carbohydrate
digestion in the alimentary tract. These experiments point out, as
do almost no others, the necessity for long experiments, preceded by
a plane of nutrition carefully studied beforehand.

Such evidence as has been accumulated in previous experiments in
which carbohydrates have been ingested shows that the respiratory
quotient is rarely above the initial value ten hours after the ingestion
of relatively large amounts of carbohydrate, thus indicating a rapid
absorption and digestion. Indeed, Magnus-Levy ’° states his inability
to increase the respiratory quotient in a man that had been given a
very rich carbohydrate diet (rice, etc.) for two days previous to the
experiment. In the light of the experiments here presented Magnus-
Levy’s statement is very difficult to explain. It seems hardly probable
that by reason of preliminary drafts upon body glycogen the store of
glycogen in his subject was so lowered that two days of rich carbo-
hydrate feeding could not replenish it.

Anomalous experiments of this kind occasionally appear, and it
seems difficult to predict exactly what the respiratory quotient will
be after a given diet with carbohydrate-rich and carbohydrate-poor
food. Our experiments show that, in general, when a carbohydrate-
rich diet is given, there is an increase in the respiratory quotient,
although certain experiments seem to show no effect whatever. In
certain of these experiments, however, it has been shown, after a care-
ful calculation, that the diets on the evening before were not so rich
in carbohydrate as the subjects thought. Thus, on one occasion when
the subject, V. G., was given a so-called carbohydrate-rich diet, it

10 Macnus-Levy: Archiv fiir die gesammte Physiologie, 1894, lv, p. 25.

Influence of Preceding Diet on Respiratory Quotient. 40%

was found that the total amount of carbohydrates ingested on the
evening before was but 200 gm. Under these conditions the respira-
tory quotient did not vary materially from that in experiments with
a carbohydrate ingestion of but 50 to 60 gm.

Obviously the previous body condition plays a very important réle.
The extent to which the body storage of glycogen has been drawn
upon, the muscular activity of the day previous to the experiment,
possibly the temperature of the surrounding air, the general diet of
the individual for several days before, in fact, anything which con-
tributes to a disturbance of the storage of glycogen in the body, may
alter the influence of the ingestion of a carbohydrate-rich meal. If
the glycogen storage in the body is at a low point, the ingestion of a
carbohydrate-rich meal does not result in an increased respiratory
quotient in accordance with the amount ingested, as a not incon-
siderable proportion of the carbohydrate may be stored immediately
as glycogen. Until this glycogen storage has been replenished the
combustion of carbohydrate in the food may be delayed. On the
other hand, with individuals subsisting without food and remaining
quiet in a respiration chamber, the store of glycogen may last for
some time. From these data we may infer, then, that muscular activ-
ity may play an important réle in affecting the storage of glycogen.

POST-DIGESTIVE KATABOLISM.

These experiments on the effect of the preceding diet on the res-
piratory exchange suggest many ideas with regard to a study of the
character of the katabolism after active digestion has ceased. Recog-
nizing the influence of the ingestion of food upon katabolism, it is of
great importance and value to know what is the character of the
katabolism after this digestive activity has ceased. Assuming that
the carbohydrate, fat, and protein have been converted into their
soluble forms and withdrawn from the alimentary tract, and active
peristalsis, secretion, absorption, etc., have ceased, what is the char-
acter of the katabolism and how can it change? The katabolism
under these conditions is obviously made up of the disintegration of
the three main compounds of the body, — the protein, the fat, and
the carbohydrate. As has been shown previously in the discussion
and in the table giving the nitrogen excretion, the protein katabolism

402 Francis G. Benedict, L. E. Emmes, and J. A. Riche.

remains essentially constant throughout all experiments. This is
assuming, obviously, that the protein katabolism follows closely the
curve of nitrogen excretion in the urine, an assumption that is com-
monly made if not absolutely correct. It is, then, clear that the
changes must take place in the relative proportions of the fat and
carbohydrate oxidized. Twelve hours after a rich meal of either fat
or carbohydrate, the body contains a relatively large amount of
fat which has not undergone proportionately any great change.
Even with no food the amount of fat drawn upon during one day or
in twelve hours would not materially deplete the storage. On the
other hand, while the storage of carbohydrate is quantitatively very
much less than fat, it is subject to fluctuations in its content which
may amount to an enormous percentage of the original storage. It
has been assumed that the amount of carbohydrates in the body of
an average man is, roughly, not far from 400 gm. This is probably
a low rather than a high estimate. If, as has been demonstrated, in
twenty-four hours without food 180 gm. of carbohydrate can be drawn
upon, it will be seen that nearly 50 per cent of this storage may dis-
appear after twenty-four hours of complete inanition. It is obvious,
therefore, that in a study of the character of the katabolism after
active digestion has ceased, we must pay particular attention to the
drafts upon body glycogen. Inasmuch as these drafts affect the
total storage by a very large proportion, it is also easily seen that a
determination of the amount of these drafts may be of great signifi-
cance, and that the respiratory quotient as an index of the character
of the katabolism may throw much light upon the storage of carbo-
hydrates or glycogen existing in the body.

Any alterations in the relative proportion of carbohydrate and fat
burned are due almost exclusively to the fluctuating amount of carbo-
hydrate storage rather than of fat storage. The fat storage is very
large and is sufficient to supply the body with energy for complete
maintenance for many days. This being the case, then, we may
again examine the transitional period from the end of active digestion
until complete inanition takes place. At the end of active digestion
it may be assumed that the body is fully charged with carbohydrate
if a carbohydrate-rich diet has been given for some time before. The
immediate drafts upon this carbohydrate are probably very rapid,
and inside of twenty-four or forty-eight hours nearly one half may

Influence of Preceding Dict on Respiratory Quotient. 403

have been used; accordingly the proportion of the total katabolism
during this time supplied by carbohydrates may be very large. In-
deed, it has been demonstrated in one experiment “ to have amounted
to 37.8 per cent of the total energy for the twenty-four hours following
the twelve hours after the last meal. At the end of the first twenty-
four hours there may be still a continued draft upon body glycogen
which with man reaches, on the average, 40 gm. in twenty-four hours.
After this second twenty-four hours is over, we have a relatively con-
stant plane of metabolism which involves the katabolism of about
20 to 25 gm. of glycogen up to and including the seventh day. Ex-
periments with fasting men determining the glycogen metabolism
longer than seven ‘days have not as yet been made. It would appear,
therefore, that the characteristic transition between whole digestion
and inanition was on the first and second days of inanition, and it is
to be questioned whether the ‘‘niichtern” value has any significance
whatever, except as representing, possibly, the first moments after
active digestion has ceased. It is obviously very difficult to distin-
guish between the exact moment when the end of digestion takes
place and the drafts upon body glycogen begin to be made. It is
highly probable that this is more or less of an interchangeable situa-
tion, and that toward the end of digestion there may already be
drafts upon body glycogen. This would be the case if, as a result of
muscular activity or even of minor muscular activity in bed at rest,
the carbohydrates in the food were not sufficient to supply the drafts
upon the previously stored glycogen.

On these grounds, therefore, we can assume that if all of the factors
tending to increase the storage of glycogen are constant and all of the
factors tending to decrease this storage are constant, there will be a
sufficient equilibrium established in which the body would have a
normal amount of glycogen present. If this glycogen is not drawn
upon or added to, we would have a constant plane of nutrition, and
one would expect that the respiratory quotient at this plane of nutri-
tion would remain constant. This being the case, it is not difficult to
realize that in many series of experiments, with a relatively constant
diet and with approximately the same muscular activity from day to

1 BENEDICT: Carnegie Institution of Washington, Publication No. 77, 1970,
P. 497.

404. Francis G. Benedict, L. E. Emmes, and J. A. Riche.

day, the respiratory quotient would not materially alter. Such a
constancy in the respiratory quotient would be expected, particularly
in experiments made on successive days and in the same period of-
the year, which would insure no change in the character of the diet.
These are the conditions which undoubtedly obtained in the series
of experiments reported by Reach.” His experiments, eleven in
number, extended from June 4 to July 13, and show respiratory
quotients which ranged only from 0.772 to 0.865, — certainly a strik-
ing uniformity.

In view of these considerations, therefore, it is questionable whether
it is logical to speak of a particular state or plane of nutrition as cor-
responding to the ‘‘niichtern,”’ or whether this state followed by the
fasting state actually exists. The most sharply defined condition is
that of the first two days of inanition after ordinary diet, before the
metabolism settles to a normal plane, which apparently continues for
five days at least without marked alteration.

From the noticeable alterations in the respiratory quotient follow-
ing diets of different character, it would seem feasible to conduct ex-
periments with the special view of altering the glycogen content of
the body. This could be done either by the ingestion of carbohydrate
to increase the glycogen storage or to decrease it by a low carbohy-
drate diet accompanied by muscular work, or possibly by the in-
fluence of cold baths. Considering the importance that the storage
of glycogen has taken in studying the metabolism of diabetics, it will
be seen that the susceptibility of the body to glycogen storage may be
of great value in the diagnosis and prognosis of disease. In fact,
studies of the respiratory quotient by modern technique would lead
one to believe that the significance of the relatively small amounts of
glycogen in the body may prove to be of the greatest importance in
studying many problems in normal as well as pathological metabolism.
The influence of muscular activity, training, etc., with regard to the
varying alterations of the glycogen storage in the body, should all
prove problems of the greatest value. It is not impossible, for ex-
ample, that ‘‘training”’ might prove to be a process of adjusting the
body to a condition in which there will be excessive glycogen storage,
and that this storage will subsequently be used for the muscular

ReacuH: Landwirtschaftliche Jahrbiicher, 1908, p. 1ogr.

Influence of Preceding Diet on Respiratory Quotient. 405

work accompanying the strenuous muscular exercise. The body is
thus prepared for the sudden drafts upon body material. Certainly,
during muscular exercise there is a draft upon body material and
perhaps the most labile substance is the glycogen. The relationship,
therefore, between the glycogen storage and muscular activity is one
that should prove of great interest.

THE RESPIRATORY EXCHANGE AS AFFECTED
BY BODY POSITION.

By L. E. EMMES anp J. A. RICHE.
[From the Nutrition Laboratory of the Carnegie Institution of Washington, Boston, Mass.]

HE well-known influence of muscular activity upon the respira-
tory exchange has led investigators for purposes of comparison
to determine the value for metabolism at the lowest attainable degree
of muscular activity, usually when the subject is lying on a couch or
sofa, and always some time after the preceding meal. This funda-
mental value has proved of much importance in determining the
effect of various agencies upon metabolism, particularly muscular
activity and the ingestion of food. In the majority of the experi-
ments, the subjects have been lying awake in a state of enforced
muscular rest. The values thus obtained are somewhat abnormal for
the practical purpose of estimating the energy requirements of in-
dividuals, in that people usually do not lie absolutely quiet while
awake; consequently their resting metabolism would be somewhat
higher than that of experimental subjects voluntarily restricting
muscular movements while lying on a couch. Investigators are at
present not in accord as to the influence of sleep on metabolism,
some maintaining that the metabolism is not lowered during sleep,’
and others,” that it is distinctly lowered.

There are relatively few observations with regard to the metabo-
lism of an individual sitting upright in a chair. ,Two experiments on
Caspari* are reported in which there was no apparent alteration in
the metabolism when the subject was sitting as compared with that

1 JoHAnsson: Skandinavisches Archiv fiir Physiologie, 1898, viii, p. 85.
2 BENEDICT and CARPENTER: Carnegie Institution of Washington, Publication
No. 126, 1910, p. 242.
8’ ZuNTz, Loewy, MULLER, and Caspari: Hoéhenklima und Bergwanderungen,
Berlin, 1906, Table XII.
406

Respiratory Exchange as Affected by Body Position. 407

when he was lying down, but the experiments were not made on a
strictly comparable basis and cannot be used as a criterion.

Johansson‘ found in a series of experiments in which the carbon
dioxide output alone was determined that the average values per
hour when the subject was sitting were about 6 per cent greater than
when he was lying down.

Benedict and Carpenter > have made a number of comparisons of
the metabolism of individuals when lying in a bed calorimeter and
when sitting in a chair inside of another type of calorimeter. These
comparisons have been made in a number of ways. When the average
of a large number of experiments with individuals asleep inside of a
respiration chamber is compared with another large number of in-
dividuals sitting up in a chair, they find an increase in the sitting
metabolism of some 35 to 40 per cent over that when the subject is
lying down. This increase must not be considered as measuring the
difference in metabolism between the body positions of sitting and
lying, for the subjects when lying down were asleep during the night
hours and therefore the metabolism was at a minimum. On the other
hand, when sitting in the chair they were not always absolutely quiet,
and, indeed, at times they might have moved considerably.

In a more recent series of experiments, as yet unpublished, Bene-
dict and Carpenter ® have compared the metabolism of individuals
when lying quietly awake in the daytime in a bed calorimeter with
that of individuals sitting very quietly in a chair calorimeter. The
increase noted in these experiments is much less than in the first
series referred to, the metabolism being on the average from 20 to
30 per cent greater with the chair calorimeter than with the bed
calorimeter.

In experiments on muscular work it has been necessary at times to
determine the metabolism while standing. This has recently been
done by Reach’ in Durig’s laboratory. In his experiments the sub-
ject was not only standing free but also leaning forward or backward
against a support.

4 JoHaAnsson: Loc. cit., p. 118.
5 BENEDICT and CARPENTER: Loc. cit., p. 252.
§ An abstract of these comparisons is given by BENEDICT and JosrIn: Car-
negie Institution of Washington, Publication No. 136, 1910, p. 175.
Reacu: Landwirtschaftliche Jahrbiicher, 1908, p. r1oo.

408 L. E. Emmes and J, A. Riche.

Zuntz has always used a position in which the subject lies on a
couch with the greatest muscular relaxation. Johansson has almost
invariably used a similar position with enforced muscular relaxation,
but aside from the few experiments of Caspari and Johansson, no
extended series of experiments has as yet been reported with men
sitting upright in a chair. Individuals sitting up in a chair quietly
reading might be considered by'some persons to have even a lower
metabolism than when lying in bed. It is believed, for example, that
there is less work performed in respiration when the person is sitting
up than when lying down, as in the lying position, the chest must be
raised and lowered at each respiration. On the other hand, when
sitting in a chair, there is invariably some muscular tension to hold
the position, as is evidenced by the general relaxation and instability
of the upper portions of the body when a person falls asleep sitting
in a chair.

The experiments reported here were made for the purpose of study-
ing the metabolism as affected by the body position. They were
planned with a special view to finding out if the metabolism increased
when the subject was sitting as compared with lying, particularly if
in the sitting position the head was comfortably supported, the mus-
cular activity was kept at the lowest possible point, and there were
no muscular movements other than those of involuntary respiration.
A comfortable armchair with a special headrest was used for the
sitting experiments, the headrest being so adjusted that the nosepieces
of the respiration apparatus described by Benedict ® could be inserted
without any discomfort to the subject. The study was made in con-
nection with the series of experiments reported in the preceding
paper ° on the effect of the previous diet on the respiratory quotient
twelve hours after the last meal.

The subjects came to the laboratory without breakfast and lay
down upon the couch in preparation for a series of experiments.
Usually three or four experiments were made in a series; the details
are given in the preceding paper.’ After the experiments were com-
pleted, the subject then changed his position and sat upright in the
chair, and a second series of three experiments followed. The details
of these latter experiments are given in Table I. But two subjects

8 BENEDICT: This journal, 1909, xxiv, p. 364.
* BENEDICT, Emmes and RIcHE: This journal, rorr, xxvili, p. 383.

Respiratory Exchange as Affected by Body Position. 409

were used in these experiments, L. E. E. and J. R., and the series are
so numbered as to show that they correspond with a similar group of
experiments in the preceding article; thus, series 1a with L. E. E.
followed immediately series 1 with the same subject in the article of
Benedict, Emmes, and Riche.

TABLE I.

RESPIRATORY EXCHANGE OF SUBJECTS WHILE SITTING IN CHAIR.

Subject,
and num-
ber of
series.

Carbon
dioxide
excretion
per min.

Oxygen
consump-
tion per
minute.

| Respir’y
quotient.

Subject, | Carbon
and num-| dioxide
ber of | excretion |

tion per

minute. |

Respir’y
quotient.

series. | per min. |

c.c.

213

c.c.

ZT
254
274

c.c.

226
218

c.c.

247

221 242

213
212 267

250

la
2a

261

pT
3a 249
242
243
258

255

aeact

No difficulty was experienced in carrying out these experiments;
both the subjects for the most part were comfortable, and the metabo-
lism was reasonably uniform. Only occasionally do we find noticeable
differences in the carbon dioxide production or oxygen consumption
in a series of ‘two or three experiments. The greatest variation is
that noted in the second experiment with L. E. E. (1 a), where the
oxygen consumption was noticeably less than in either of the other
two experiments. Usually the agreement is fully as satisfactory as
could be expected under the conditions of the experiment.

A comparison of the metabolism when lying on a couch with that

AIO L. E. Emmes and J. A. Riche.

when sitting in a chair is given in Table II, in which we have the values
not only for carbon dioxide excretion and oxygen consumption per min-
ute, but the percentage increase while sitting has also been computed,

TABLE II.

COMPARISON OF THE AVERAGE PULSE RATE AND RESPIRATORY EXCHANGE OF SUBJECT
SITTING IN CHAIR AND LyING ON COUCH.

Carbon dioxide Oxygen
Subject, and ee Paice excretion. consumption.
number of sey rate per :
series. peer: minute.
Per Increase Per Increase
minute. | in chair. | minute. | in chair.

cic per cent. c.c. per cent.

Lying 59 ye a 242

Sitting 71 216 a9 268 10.7
Lying 56 212 a 235
259
240
249
232
252

Sitting

Lying

Sitting

Lying

Sitting

233
eee Sitting 77 219 9.0 248 6.4

Lying

“we 9 Lying 67 04s al) aoe 246
“ « 94! Sitting 74 208 2.0 249 “1.9
eo 3 Lying 68 P19 orn 236
“ «© 3q| Sitting 73 225 2.7 256 8.5
wow 4g Lying 63 i049 onda 226

254

Sitting 70 207 0 ns

using as the basis the metabolism of the subject while lying on the
couch. For the carbon dioxide excretion, the values show a range
from —2.8 per cent in series 2a with L. E. E. to +9.9 per cent in
series 3 a with L. E. E. For the oxygen consumption there is uni-
formly an increase, ranging from 1.2 per cent in series 2 a with J. R. to

Respiratory Exchange as Affected by Body Position. 411

12.4 per cent in series 4 a with J. R. The table also shows the average
pulse rate determined in each series of experiments. An inspection of
the data shows an increase in the pulse rate in the experiments while
sitting.

Comparing all of the data with both subjects, we obtain the follow-
ing results:

Carbon dioxide Oxygen

Pulse. c.c. per min. c.c. per min.
LS eR eh Pete A ee od oe 209 236
SULT ae en te OAC ey a 218 254
LIANE 1S Sigua Ann Set ayn 8 9 18
Percentage,of increase in metabolism . . . . 4.3 7:6

Inasmuch as the oxygen consumption is commonly considered
as the best index of metabolism, it is seen that our experiments in-
dicate an increase in metabolism amounting to 8 per cent when the
metabolism in the lying position is compared with that when the sub-
ject is sitting upright. These values may probably be considered as
subject to slight corrections. For example, the experiments with
the couch were always made earlier in the morning than experiments
with the chair; consequently, since the metabolism shortly after ar-
riving at the laboratory is frequently somewhat larger than after the
subject has been lying down for several hours, the metabolism meas-
ured on the couch would be slightly larger than the normal value
determined at the period of the day when the sitting experiments
were made. This would tend to lower somewhat the apparent in-
crease in metabolism. We believe, however, that this error cannot
be very large, and would not be more than 1 or 2 per cent of the total
metabolism.

In the series of experiments reported by Benedict and Carpenter,!?
there was an average increase of some 35 to 40 per cent in the metabo-
lism while the subject was sitting in a chair as compared with the
metabolism when the subject was lying in bed asleep. As has been
pointed out before, the validity of this increase is vitiated by the fact
that when in bed the subjects were all asleep with consequent mini-
mum metabolism, while in the chair they were somewhat active. On
the other hand, in the series of experiments made by Benedict and

“ 10 BENEDICT and CARPENTER: Loc. cit.

412 L. E. Emmes and J. A. Riche.

Carpenter for control tests in a study of diabetes," the increase was
not so large, amounting on the average to a total increase in metabolism
of about 25 per cent. This increase is, however, more than twice as
great as that found in the experiments reported here. The natural
explanation for this is that the subjects of our experiments were ab-
solutely quiet aside from the change in body position, while in the
other experiments the muscular activity of the subject while sitting
in a chair inside the calorimeter was certainly very much greater than
that of either of the subjects in our sitting experiments.

A series of experiments on women, as yet unpublished, has re-
cently been carried out in this laboratory by Carpenter, in which the
chair calorimeter was first used, and then the bed calorimeter. These
experiments showed an increase in metabolism in the chair calorimeter
over that in the bed calorimeter amounting to 7 per cent. This value
agrees much more closely with those obtained in our experiments,
and the agreement is undoubtedly due to the fact that the series of
experiments with the women were made under such conditions that
extraneous muscular activity was minimized. The women, as a rule,
were noticeably less restless than the men, particularly while in the
chair calorimeter.

The values given by Benedict and Carpenter for the difference in
metabolism between persons lying asleep and sitting in a chair awake,
with a moderate degree of restlessness, may be of much practical value
in estimating the energy requirements of convalescents. For ex-
perimental purposes, however, when the metabolism at a given con-
dition of body rest is to be determined, it is also of value to know, as
a result of experiments with the respiration apparatus, that the
metabolism of a subject when sitting absolutely quiet in a chair with-
out extraneous muscular activity represents a metabolism 8 per cent
greater than that of a subject lying on a couch with similar mus-
cular rest. The difference in metabolism is then due, primarily, to
the difference in the internal muscular activity necessitated by the
sustaining of body parts. This is in conformity with the well-known
fact that the pulse rate of an individual when sitting is always notice-
ably higher than that when he is lying down.” From these tests

11 Cited in abstract by BENEDICT and Jositn: Loc. cit.
12 Guy reports the average pulse rate of roo men, averaging 27 years in age, as
70.1 per minute while s‘tting and 66.6 while lying. See “Cyclopedia of anatomy

Respiratory Exchange as Affected by Body Position. 413

we could infer that if it were possible to so support the body of the
subject in a sitting position that the pulse rate would be no greater
than when the subject was lying down, the metabolism would be
essentially the same in both positions.

”

and physiology,” 1852, iv, p. 186, cited by LEoNARD HILt in ScHAEFER’s Text-
book of physiology, 1900, ii, p. ror.

Dn et Ore re ie

4 — oe
- vias
7 L4
.
- ‘
.
z ,

* | \ - : Wr Al =
y eae is it; wiley facta rs oa

; ‘ to We

CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE
MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE
UNDER THE DIRECTION OF E. L. MARK. No. 220.

CONTRIBUTIONS TO THE PHYSIOLOGY OF REGENERA-
TION. — V. REGENERATION OF ISOLATED SEGMENTS
AND OF SMALL PIECES OF WORMS.

By SERGIUS MORGULIS.

HE question why an animal ceases to grow when it has once at-
tained a certain size is of fundamental importance in biology,
but thus far it has eluded every attempt at a solution. The question
why one organism grows faster or slower than another is a still more
difficult problem, a solution of which at the present time we cannot
even approach, because our knowledge of even the simpler processes
of growth is still very meagre. In the province of regeneration two
problems similar to those stated present themselves to the investigator
who deals with this phenomenon from a quantitative standpoint:
why does regeneration come to an end, and why does an organ regen-
erate at different rates under changed conditions?

It is an old fact, but recently studied anew and greatly emphasized,
that an amputated organ regenerates with varying degrees of inten-
sity according to the position, or level, of the cut surface, the intensity
diminishing in an antero-posterior direction. The tail of a salamander,
for instance, regenerates fastest from (7. e., when cut off at) the base
and slowest from the tip, and all gradations are found between the
two extremes. A phenomenon of such interest has naturally stimu-
lated scientific thought, and a number of more or less ingenious hy-
potheses have been offered to explain it. These hypotheses have
sought to interpret the different rates of regeneration on the basis of
the different amount of nutritive substance available for the regener-
ation of a lost organ, or on the basis of a nervous impulse expressed
in the form of functional activity, or on the basis of the varying de-
grees of pressure mutually exerted by developing parts, or, finally,

é 415

416 Sergius Morgulis.

even on the basis of an occult formative stimulus. It would take us
too far afield to discuss critically these various views, especially since
they have been discussed in the second part of these Contributions
(Morgulis, 1909"), to which the reader is referred for a detailed ac-
count. In the same publication was proposed an hypothesis suggested
by the fact that in the linear growth of an organ successive portions
are formed at gradually diminishing rates. The progressive diminu-
tion of the growth energy is undoubtedly due, it was there maintained,
to the diminished reproductive power of the cells, which lose their
original vitality and vigor with each successive generation. A linear
organ presents regionally throughout its length various degrees of
senility, the more senile segments being the more posterior ones; the
different rates of regeneration from the corresponding regions are
therefore explicable on the ground that the cells are endowed with
different degrees of vitality. This hypothesis, supported directly by
observations on regeneration from different levels (see Morgulis,
1909*, pp. 600-610), and indirectly by the fact that young animals as
a rule regenerate more vigorously than adult animals, presupposes
furthermore a regional differentiation in the organs. The assumption
of such a material basis seemed to be warranted by the experiments
on Lumbriculus, where it was shown that the relative potentiality of
regeneration is transmitted from the original piece to the regenerated
piece (Morgulis, 1907, p. 219), and also by the fact that when a circu-
lar strip of tissue is cut off from a jelly-fish the rate of regeneration is
the same both peripherally from the disc and centrally from the strip.
The assumption has received further confirmation in the experiments
on isolated segments of Podarke obscura, which will now be described.?

This worm, which is about three fourths of an inch long, is made
up of 4o to 50 segments, each bearing a pair of parapodia. The indi-
vidual segments are very small, but with the aid of a watchmaker’s
lens they can be easily distinguished. Between every two adjacent
segments there is a more or less deep constriction; the area of junction
of the segments is larger in the middle portion of the worm than in
the posterior. The operation of removing single segments increases
in difficulty as one passes from the tip of the tail towards the head. It
is a comparatively simple operation to obtain single segments from

1 The experiments were performed in the laboratory of the U. S. Bureau of
Fisheries, Woods Hole, Mass., in the summer of 1909.

Contributions to the Physiology of Regeneration. 417

the hinder third of the animal, the chief difficulty there being the small-
ness of the segments. In the middle of the worm the segments are
relatively large, but, the area of junction between segments being also
large, one is not always successful in isolating them without much
injury. It is practically impossible to obtain single segments from
the anterior third of the worm, because the digestive tube in that
region is rather large and muscular and therefore prevents the effec-
tive separation of two adjacent segments. In the few instances, how-
ever, where the isolation was accomplished, the segments soon died
because they were badly mutilated.

It is necessary, above all, to conduct experiments with large num-
bers of single segments, as only a small per cent of those successfully
removed survive. It is not easy, however, to fulfil the requisite condi-
tions. In the first place, for the survival of the segments it is abso-
lutely necessary that the wound should close as quickly as possible;
this occurs in those instances where no, or at any rate very little,
tissue from adjacent segments remains attached. A great many
isolated segments die in consequence of their failure to close up the
wound, either because the wound surface is too large or because a
piece of the gut protrudes, thus preventing the margins of the wound
from completely coalescing. Besides, attached fragments of tissue
become easily infected with bacteria and thus cause death to the iso-
lated segments. Secondly, on the day following the operation the
segments must be cleaned of every trace of cast-off epidermis and
adhering tissue; it is needless to say that, as a result of this delicate
operation, many isolated segments perish. Finally, the continual and
vigorous contraction of the segments is an important cause of mor-
tality, because the stiff bristles of the parapodia, being alternately
pulled in and out, break through the body wall and in fact frequently
tear the segment to pieces. The process of regeneration of the iso-
lated segments is in no manner different from that which I have al-
ready described for larger portions of the worm in the case of Podarke
(Morgulis, 1909", p. 601). Moreover, what was found previously with
regard to worms cut at different levels has now been fully corrobo-
rated by experiments on isolated segments belonging to different
regions of the worm; namely, that the more anterior the level the
sooner does the regeneration begin. The results, however, must be
judged with much caution, because, as pointed out before, the cut

418 Sergius Morgulis.

surface being smaller in posterior segments, the preliminary stages of
regeneration — the closure of the wound — are accomplished there
more easily. But in spite of that, as will be shown presently, isolated
segments from an anterior region regenerate faster than those from
a posterior region.

The results of the experiments with isolated segments agree among
themselves in every essential respect, except that more segments may
have survived in one instance than in another. Here are the results
of an experiment. Three days after operation 21 isolated single seg-
ments from the middle of the worm and 40 single segments from the
posterior portion remained alive. Of the former 7 (or 33.3 per cent)
commenced to regenerate; of the latter only 8 (or 20 per cent). Five
days after operation only 12 isolated segments from the middle of the
body were alive, and of these 58 per cent had regenerated; and of the
32 surviving single posterior segments 13 (41 per cent) were found re-
generating. Three days later still, 7. e., eight days after the segments
were isolated, the former had regenerated on the average 3.3 new
segments, while the latter had regenerated only 3 segments, the ex-
tremes in the number being 3 to 4 segments in the first case and 2 to 4
segments in the other. Although isolated segments were frequently
maintained alive for two weeks, they never regenerated more than
4 new segments. The isolated segments are bound to die soon for
lack of nutrition, because a head is not regenerated in this worm; but
it is very noteworthy that such an exceedingly small part of an organ-
ism — not more than a millimetre in length — is capable not only of
existing for the space of a fortnight without any food, but also of con-
tinually expending energy in its manifold motions and, in addition to
its other functions, of building new segments out of its own substance.
One who has watched these isolated fragments of the organism —
themselves complete organisms — convulsively contracting and thus
shifting about the bottom of the receptacle, who has seen the inces-
sant but vigorous lashing of their ciliated surfaces, and who has ob-
served the unfolding step by step of a new worm out of this clump of
tissue cannot evade the urgent question, Whence is all this energy?

Even parts of a segment can be maintained alive, and these will
proliferate new tissue. Of course, the difficulties here are so numer-
ous and so great that experiments ‘succeeded only rarely; therefore
the regeneration of fragments could not be investigated satisfactorily.

Contributions to the Physiology of Regeneration. 419

However, I am convinced from the few successful instances that not
only does the wound close over, but that also a growth of new tissue
occurs; but whether or not such fragments of a segment can complete
the missing portion is a question which, under the circumstances, must
remain unanswered. Another question of great theoretical weight,
namely, whether or not an isolated segment can regenerate again
when the regenerated tissue has been removed, must remain unsolved
on account of the impractibility of the experiment.

The regeneration of isolated segments is significant in several re-
spects. In the first of these Contributions (Morgulis, 1909) it was
shown that the process of posterior regeneration is divisible into four
distinct stages, each characterized by a different intensity of the re-
generative energy, and that this is true of all regenerating worms re-
gardless of the level of the cut. There is a difference between regener-
ation from an anterior and from a posterior level; for at no stage of
the process does the organ regenerate with as great intensity from
the posterior level as from the more anterior. Morgan (1906) thought
that the regenerative potentialities are the same at all levels, al-
though he foresaw that further researches might disclose a difference
- in the potentiality. The experiments on isolated segments do, I be-
lieve, prove the existence of a regional differentiation in the regenera-
tive potentiality. I may mention in this connection that isolated seg-
ments from the very tip of the tail did not regenerate at all. These
segments are extremely minute, but the smallness of their size has
nothing whatever to do with their failure to regenerate. Of course,
this point cannot be proven directly by experiment, but its cogency
is nevertheless obvious from a consideration of what is known gener-
ally on this subject. As I showed previously, segments only a few
times larger may regenerate even as many as four new segments;
furthermore, the fact that regeneration is largely independent of food
would make it highly doubtful whether the failure of the most poste-
rior segments to regenerate is due to a lack of formative material.
Besides, the greater regeneration from an anterior level completely
refutes such a possibility, because in this case the greatest regenera-
tion occurs where there is least material. According to the hypothesis
which I have suggested (Morgulis, 1909”, p. 435), the most posterior
segments of the worm, being the remotest descendants from the
original embryonic material, z.¢., the latest growth of the organism,

420 Sergius Morgults.

are also the most senile and therefore the least capable of regenera-
tion. Similarly, in some animals the power to regenerate diminishes
almost to complete disappearance as they grow from the embryonic
to the adult condition.

The regeneration of isolated segments gains particular interest
when compared with the regeneration of large portions of the worm.
A single segment does not begin to regenerate as quickly as a piece
embracing one half or two thirds of a worm, but it should also be re-
membered that in the single segments there are two wounds to be
closed, whereas in the larger pieces there is only the posterior wound.
Therefore the longer time required by isolated segments before the
tail commences to regenerate may be due to the greater complexity
of the preliminary stages of regeneration. The greatest number of
new segments which an isolated segment can regenerate in the course
of eight days is 4, or only two thirds of that which a piece 15 to 17
segments long regenerates within the same length of time; as has been
intimated in the foregoing, the isolated segments never regenerate
any more. When it is recalled, however, that large pieces of a worm,
even in the course of weeks, do not regenerate more than about 12 to
14 new segments, it becomes obvious that an isolated segment regen-
erates relatively more than a piece 15 to 17 segments long. The study
of the regeneration of isolated segments leaves no room for doubt that
practically every segment of the worm Podarke is capable of pro-
ducing out of its own substance a few new segments, and even as
many as 4. A group of 15 to 17 segments must, accordingly, have
sufficient formative material for some 40 to 60 new segments, of which,
however, they actually regenerate only a small fraction, therefore
the potential regenerative energy of each component segment is only
partially utilized. The proportionately greater regenerative effi-
ciency of isolated segments is of much importance in understanding
the physiology of regeneration, but before discussing this topic I
shall present some further results upon the regeneration of pieces
varying in initial size of another worm, Lumbriculus.

In a number of worms the posterior halves were removed, and the
anterior portions were divided into three groups. In the first group
(A) the half-worms were merely decapitated; in the second group (B)
the half-worms were cut into two parts, of nearly equal length, and in
the third group (C) they were cut into four nearly equal parts; in all

Contributions to the Physiology of Regeneration. 421

three groups only the most posterior pieces were experimented upon.
In other words, the level of the posterior cut is, roughly speaking, the
same in all three groups, but according to their size the pieces repre-
sent about one half, one fourth, and one eighth of the entire worm,

TABLE I.

Group ..

Segments .

N
ae
N
=
~
nm
=)
to
=
<

Ratio between old and
regenerated segments

t 1.271

as is shown diagrammatically in Fig. 1. When such pieces of Lumbri-
culus are allowed to regenerate for several days, one finds that the
difference in the number of regenerated segments is not commensur-
ate with the difference in the number of segments in the regenerating
pieces. An inspection of Table I will make this clear. From this
table it will be seen that the average number of old and of regenerated
segments in the A pieces was 42.5 and 54 respectively; in the B pieces
26.2 and 43.8, and in the C pieces 14.1 and 41, Even these figures
suffice to show the proportionately greater regeneration in pieces of

4220 Sergius Morgulis.

the smaller size; but the precise ratio is more clearly perceived when
the number of regenerated segments is divided by the number of seg-
ments in the regenerating piece, which is equivalent to determining
the number of segments regenerated on the average for each old seg-
ment; this number serves as the coefficient of regeneration for the
corresponding group of pieces. We find that the coefficient of regen-
eration in the A, B, and C pieces was respectively 1.271, 1.672, and
2.900.

<— Head <i Tail —~+>
I : 2.906

Ficure 1.— The three diagrams, A, B, C, represent the average number of segments in
the regenerating (heavy line) and in the regenerated (light line) portions of the worm
for the groups A, B, C, respectively. The numerals under the diagrams show the
ratio of the number of segments in the two portions.

In another experiment two groups of pieces, having the posterior
cut approximately in the middle of the worm, and embracing respec-
tively about one eighth (A) and one sixteenth (B) part of the worm,
were left to regenerate for several days. The results of this experi-
ment are recorded in Table II. Like those of the previous experi-
ment, these show that the number of regenerated segments does not
diminish in the same ratio as the number of segments in the regener-
ating pieces. Thus, in the A group the average number of old seg-
ments was 12.8 and of regenerated segments 34.5; in the B group it
was 7.6 and 27 segments respectively. The coefficient of regeneration
in group A is 2.706 and in group B 3.573; that is to say, there is mani-
fested relatively more regenerative potentiality in the small pieces of
the worm than in the larger ones.

We may infer from these experiments that two ‘lie pieces of a worm
regenerate considerably more than one 14 piece, and that two 4%
pieces regenerate more than one ]4 piece, etc. Furthermore, as the
coefficients of regeneration of the two experiments indicate, within
the same length of time, every segment of the 14 worm regenerates

Contributions to the Physiology of Regeneration. 423

1.271 new segments, of the 14 worm 1.672, of the 44 worm 2.906 (in
the second experiment 2.706), and, finally, of the '/;, worm 3.573 seg-
ments. Therefore the rate of regeneration in the small pieces is also

TABLE II.

Regenerated. : Regenerated.

35
33
30
27
27
27
26

\o
t=]
o

QQ
FS
oO

—
a,
oO

bo od

—)
Ww
=)
a0
=

20
18

Average : 5 : 27

Ratio between old and
regenerated eel . 2.706

greater than in the large pieces, the relative rate of output of the
regenerative energy increasing as the initial size of the regenerating

object decreases.
A

I: 2.706

<— Head —=—_—_———— Tail —>
I: 3-573

FicurE 2. — See explanation of Fig. 1.
Now, with the two chief results established, namely, that isolated

segments regenerate proportionately more than groups of segments (Po-
darke), and that the smaller the piece the greater the rate of its regenera-

424. Sergius Morgulis.

tion (Lumbriculus), we may proceed to discuss their bearing upon
some general problems.

Within the past fifteen years or so much biological discussion has
centred about the problem of the potentiality of the developing egg
and of the isolated blastomeres. ‘This discussion, which has stimu-
lated much thought and experimental work bearing fruitful results in
the province of developmental mechanics, concerned itself chiefly
with the qualitative study of the whole and partial egg. Unfortu-
nately, we know much less with regard to the quantitative differences .
between the development of an egg or of its component blastomeres,
and it would be highly desirable to fill this gap in our knowledge.
The general conclusion following from the study of isolated blasto-
meres is that “‘a part is capable of doing that which the whole is set
aside to do” (Morgan, 1896, p. 292). Moreover, as Morgan (1896,
p. 291) has said, the egg fragments divide ‘‘beyond their normal
cleavage limit’; this is very similar to the above results from the
regeneration of isolated segments.

The bearing of the study of regeneration of isolated segments and
of pieces of worms of different initial size upon the physiology of re-
generation will be more clearly seen if some of the already established
facts are recalled. It has been shown on’several occasions that an
organism is capable of regeneration many times in succession. Further-
more, the amount of regeneration after a simgle operation is consider-
ably less than the amount regenerated for the same length of time
following two or more operations (Morgulis, 1907*, 1909, 1909", 1911,
and 1911»). Another important fact recently brought to light is that
regeneration usually ceases before the lost portion has been completely
restored. Thus, in Podarke it was found (Morgulis, 1909, 1911°)
that only 0.4 of the number of removed segments regenerated in the
course of about eight weeks. As at that time the regenerative pro-
cess is practically at a standstill, it is quite certain that not more than
half as many segments regenerate as are removed. But the regenera-
tion does not come to an end because the regenerative capacity of
the organism has been exhausted, for by removing the new tissue the
regeneration may be started again. From the experiments on iso-
lated segments we may infer that had the sum of the regenerative
potentialities of all segments been utilized, it would have sufficed not
only to replace that which was lost, but to produce even an excess.

Contributions to the Physiology of Regeneration. 425

The explanation of this phenomenon, it seems to me, lies in the cir-
cumstance that the organism possesses a large regenerative potenti-
ality, of which, however, only a small fraction is actually utilized in
the process of regeneration. The results of the experiments with
pieces of worms of varying sizes suggests that there is a factor which
determines to what extent the regenerative potentiality may be uti-
lized. From the fact that, while other conditions remain equal, the
smaller the initial size of the regenerating object the more and the
faster proportionately it regenerates, we may infer that the organ-
ism presents a certain amount of inertia, due to a tendency to main-
tain a definite state of equilibrium and of functional adjustment.
_ This inertia, which forms a resistance to regeneration that must be
overcome, varies directly with the size of the regenerating object.

I do not wish to be understood as attempting to translate a bio-
logical fact into terms of pure mechanics, although a comparison with
the well-known behavior of inanimate objects is strongly suggested.
In conclusion I would, however, mention some other facts which admit
of the same interpretation. It was discovered by Chambers (1908),
when he assorted eggs of a single frog according to their dimensions,
that the smaller eggs developed faster; and Morgan (1906) also found
that the rate of growth of the salamander Diemyctylus viridescens
depends upon the size of the animal, being greater in the small indi-
viduals. We see in these instances the same relation between the
initial size of the object and the phenomena of growth and de-
velopment that has been established by the foregoing experiments
on regeneration.'

BIBLIOGRAPHY.
CHAMBERS, R.

1908. Einfluss der Eigrésse und der Temperatur auf das Wachstum und die
Grdsse des Frosches und dessen Zellen. Archiv fiir mikroskopische Anatomie,
Ixxii, pp. 607-661.

Moreay, T. H.

1896. The number of cells in larve from isolated blastomeres of Amphioxus.
Archiv fiir Entwicklungsmechanik, iii, pp. 269-294, Taf. 17.

1906. The physiology of regeneration. Journal of experimental zodlogy, iii,
PP- 457-500.

Morcutis, S.

1907*. Observations and experiments on regeneration in Lumbriculus. Journal

of experimental zodlogy, iv, pp. 549-574.

1 This paper was read at the meeting of the Eighth International Congress of Physi-
ologists in Vienna, September, 1910.

426 Sergius Morgulis.

Morcutts, S.

1907. Regeneration and inheritance. Ohio naturalist, viii, pp. 219-221.

1909*. Contributions to the physiology of regeneration. —I. Experiments on
Podarke obscura. Journal of experimental zodlogy, vii, pp. 595-642.

1909”. Contributions to the physiology of regeneration. — II. Experiments on
Lumbriculus. Archiv fiir Entwicklungsmechanik, xxviii, pp. 396-439.

1911. Contributions to the physiology of regeneration. — III. Further ex-
periments on Podarke obscura. Journal of experimental zodlogy, x, No. 1,
pp. 7-22.

ro11. Contributions to the physiology of regeneration. —IV. Regulation
of the water content in regeneration. Journal of experimental zodlogy, x,
No. 3 (in press).

Biologische Versuchsanstalt,
Wien, September 28, 1gro.

vet

i iF ite

THE INFLUENCE OF COLD BATHS ON THE
GLYCOGEN CONTENT OF MAN.

By GRAHAM LUSK.

[From the Physiological Laboratory of the Cornell University Medical College,
New York City.]

ULZ ! showed'that a dog which had fasted twenty days still con-
tained glycogen in the muscles, and he warned against the com-
mon use of the term ‘“‘carbohydrate-free animal.” He discovered
that strychnin tetanus was a real means of producing an animal which
was free from glycogen. Zuntz? demonstrated that fasting rabbits,
deprived of glycogen by strychnin convulsions and subsequently
narcotized with chloral hydrate, stored glycogen both in liver and
muscles during the interval of quiet.

These experiments of Zuntz illuminated certain respiratory quo-
tients obtained by him® as the result of exercise, in fasting men.
The period during which mechanical work was done was characterized
by a higher respiratory quotient than that obtained during a period
of rest. The following table shows this:

Respiratory quotient.
Rest previous to work. Period of work.

~ Fourth day of starvation . . . 0.63 0.78
Fifth day of starvation . . . . 0.66 0.74
Sixth day of starvation. . . . 0.69 0.74

Durig * has obtained low respiratory quotients (0.71) during periods
of rest in the Capanna Margherita, a hut on Monte Rosa (4560
metres), following periods of intensive exercise.

1 Kt1z: Lupwice’s Festschrift, Marburg, 1891, p. 109.
2 Zuntz: Archiv fiir Physiologie, 1893, p. 378.
3 LEHMANN, MULLER, Munk, SENATOR, ZuNTz: VircHow’s Archiv fiir pa-
thologische Anatomie, 1903, Supplement to vol. cxxxi, p. 197.
4 Duric: Denkschrift der mathematisch-naturwissenschaftlichen Klasse der
kaiserlichen Akademie der Wissenschaften, 1909, Ixxxvi, p. 107.
427

428 Graham Lusk.

Since 15 per cent of the total energy of a fasting man arises from
protein and 85 from fat, it has been calculated that the respiratory
quotient should be 0.722, were these materials oxidized in this
relationship. The low quotients during rest would indicate the re-
tention of a body rich in oxygen, perhaps glycogen, which could be
oxidized in case of further exercise. A quotient of 0.75 indicates that
about g per cent of the oxygen absorbed is utilized in the combus-
tion of carbohydrates. Benedict° finds that the respiratory quotient
of a fasting man varies between 0.74 and 0.75, from the second to
the seventh day of starvation.

Magnus-Levy ® has calculated that if a diabetic oxidizes 100 gm.
of protein and 250 gm. of fat, and if from the protein about 60 gm.
of dextrose arise (as first established by Reilly, Nolan, and Lusk”), the
respiratory quotient would then be 0.699 instead of 0.722. Magnus-
Levy also calls attention to the fact that owing to the loss of carbon
dioxide through the skin this quotient is reduced by from o.o1 to
0.015, when it is obtained by the measurement of the gaseous ex-
change through the lungs only.

Magnus-Levy ® himself has reported low quotients of 0.693, 0.697,
0.719, obtained from cases of severe diabetes. Similar quotients
have also been found in severe diabetes by Benedict and Joslin.®
Thus in case C. 19, Oct. 25, 1909, when the D: N ratio for the day
may be calculated at 3.5:1 (which indicates a complete intolerance
for carbohydrate), the quotients obtained during four intervals within
about an hour’s time were 0.70, 0.70, 0.68, 0.68, an average of 0.69.

Rubner 7° investigated the influence of baths upon a man and
found that a douche at 16° raised the oxygen absorption 110 per cent
above the normal, and increased the respiratory quotient from 0.87
to an average of 1.03, from which it was evident that the source of the
metabolism during the douche was glycogen.

5 BenepicT: Metabolism in inanition, Carnegie Nutrition Laboratory, 1907,
p. 451.

6 Macnus-Levy: Archiv fiir Physiologie, 1904, p. 381.

7 REILLY, NoLan, and Lusk: This journal, 1898, i, p. 395.

8 Macnus-Levy: Zeitschrift fiir klinische Medizin, 1905, lvi, p. 86.

® BENEDICT and Jostin: Metabolism in diabetes mellitus, Carnegie Institution
of Washington, rgro.

10 RuBNER: Archiv fiir Hygiene, 1903, xlvi, p. 390.

Influence of Cold Baths on Glycogen Content of Man. 429

The writer" has shown that cold baths administered to fasting
phlorhizinized dogs at first raised the D:N ratio, indicating a re-
moval of residual glycogen, but subsequent baths were without in-
fluence. Ringer ” has demonstrated that a dog so treated yielded no
extra sugar in the urine on administration of adrenalin, which is fur-
ther evidence of a complete exhaustion of the glycogen supply.

The question arises whether cold baths will not completely remove
the glycogen content of a man, and to that end two individuals who
had accustomed themselves during the summer to a swim of fifteen
minutes’ duration in the early morning in cold water, and who took
daily cold douches as a matter of routine at other times, submitted
themselves during the morning (without breakfast) to the action of
two baths, each of six to ten minutes’ duration. The water contained
floating ice blocks and had a temperature of 10° C., the baths being
given at intervals of about two hours. As will be seen below, the
treatment was effective in exhausting the glycogen supply.

EXPERIMENTAL PART.

The apparatus used in these experiments is called the ‘‘small Bene-
dict apparatus”* for studying the respiratory exchange.’ It was
constructed in this laboratory by Mr. J. A. Riche, who had had
large experience with its use in the Nutrition Laboratory at Boston.
Mr. Riche also had control of the apparatus in the determination of
the data here presented. The individual under investigation lies
comfortably on a bed and respires through a nose piece.

The following protocols may be given:

Case A.— October 14, 1910. 2 P.M., usual luncheon. Lively ride one and
a half hours on a horse, followed by cold douche. 7.30 p.M., dinner
of fish and steak, no carbohydrates.

October 15. 6 A.M. Rose and took a cold douche; 6.45, took a cup
of hot black coffee.

Period I. — 10.04 to 10.19 A.M. Normal resting period.

First Bath, 10.38 to 10.44. Immersed six minutes in bath with ice

1 Lusk: This journal, 1908, xxii, p. 163.
22 RINGER: Journal of experimental medicine, 1910, xii, p. 105.
13 BENEDICT: This journal, 1909, xxiv, p. 345.

430 Graham Lusk.

blocks; temperature, 10° C.; skin red, shivering. The knees and chest
could not be simultaneously immersed, and therefore a movement was
made which immersed the two alternately.

Period IT. — 10.48 to 11.03. Period with shivering. Shivering lasted during
eight minutes of the experimental period. Skin only partly dry, partly
covered with a sheet; room temperature, 23°, window open. No dis-
comfort, shivering seemed a normal compensation for heat loss.

11.05. Cup of hot coffee.

Period III. — 12.04 to 12.19 P.M. Normal resting period.

Second Bath, 12.29 to 12.37; temperature of bath, 12°. General
shivering during the eight minutes of immersion.

Period IV. — 12.41 to 12.56. Period with shivering. Shivering lasted six
to seven minutes. Urine showed neither sugar nor albumin as a result
of the second bath.

Period V. — 1.26 to 1.41. Normal resting period.

Food, 1.45 to 1.55. Food taken, tooo c.c. of milk with added cream,
milk-sugar, and four raw eggs. The whole contained 8.25 gm. of ni-
trogen and 1212 calories, divided as follows: protein 201, milk-sugar
282, and fat 729; the diet had been prepared for a tuberculous patient.

Period VI. — 3.30 to 3.45. Normal resting period after food ingestion.

The results of these experiments are found in Table I. The calorific
values are calculated on the basis of Zuntz’s tables as given by Magnus-
Levy,“ it being assumed that 15 per cent of the oxygen absorption
is utilized for protein oxidation.

During the shivering which followed the second cold bath the
metabolism as measured by the heat production was 63 per cent
higher than during the subsequent resting period, but the respiratory
quotient of 0.75 remained unchanged during both periods. It is
evident from this experiment that the influence of two successive cold
baths which cause shivering during a period when the intestine is free
from carbohydrate, is sufficient to change the metabolism from one main-
tained at the expense of carbohydrate (R. Q. = 0.99) to one maintained
essentially by the combustion of fat (R. Q. = 0.75). Hence the organism
of man may be quickly rid of glycogen by shivering.

Case B.— November 4, 1910. 2P.M., usual luncheon. 6.30, dinner of roast
beef without carbohydrate, appetite unappeased. 9.30, retired.

“4 Macnus-Levy: von NoorpEn’s Handbuch der Pathologie des Stoffwechsels,
1906, i, p. 207.

Influence of Cold Baths on Glycogen Content of Man. 431

November 5. 5 A.M. Rose and took a cold bath. Studied till 6.45.

7.30, took a large cup of black coffee.
Period I. — 10.18 to 10.33 A.M. Normal resting period. Blood pressure,
max. = 135 mm., min. = go mm.; pulse, 74.
First Bath, 10.44 to 10.53. Immersed nine minutes in bath with ice
blocks; temperature, 10° C. Commenced shivering after half a minute;

skin red.
Period IIT. — 10.59 to 11.14. Shivering continued throughout the period.
Skin partly dry, partly covered with a sheet.
11.05. Blood pressure, max. systolic = 150, min. diastolic = 100;
pulse, 92.
11.15. Urine is free from sugar and albumin.
Period IIT. — 11.56 A. M. to 12.11 P.M. Normal rest, covered with a warm

blanket.
Second Bath, 12.16 to 12.26; temperature, 10°. General shivering
throughout the period of immersion.
Period IV. — 12.32 to 12.47. Shivering during eight minutes of the period.
Period V. — 1.33 to 1.49. Normal resting period.
Food, 1.55. Took cup of coffee containing 50 gm. of dextrose.
Period VI. — 2.13 to 2.28. Resting period showing the influence of dextrose
ingested eighteen minutes before the commencement of the experiment.

During two periods of the experiment the respiratory quotient
sank to 0.67 and below, or near the diabetic quotient when even the
sugar arising from metabolized protein is not oxidized. On the as-
sumption that 250 gm. of fat and 1oo gm. of protein are oxidized by
the diabetic and that 60 gm. of protein-sugar appear in the urine
Magnus-Levy * has calculated the respiratory quotient to be 0.699.
Fat would here furnish 2325 calories, protein 185 calories (225 cal-
ories being eliminated as urinary dextrose). This oxidation would
require 549.3 litres of oxygen. Hence 1 litre of oxygen would have
a calorific value of 4.589. When fat alone is oxidized, the value is
4.686. The error in using the former figure given cannot therefore
be great. On account of tHe loss of carbon dioxide through the skin
the R. Q. of 0.67 may be raised to 0.685 in the present experiment.

No explanation. of the quotient 0.62 is available, although it has
been observed by other investigators, and no attempt has been made
to calculate the heat value of the metabolism of the period with
which it is associated. (See Table II).

15 Macnus-Levy: Archiv fiir Physiologie, 1904, p. 381.

432 _ Graham Lusk.

TABLE I.

Person A. Weicut = 75.95 xc. AREA OF Bopy SURFACE

Per minute.

Period.

10.04-10.19
10.48-11.03
12.04-12.19

12.41-12.56
1.26— 1.41
3.30— 3.45

TABLE II.

Person B. Wercut = 64.71 xc. AREA oF Bopy SURFACE

Per minute.
Period. as i R.Q.
COz Oz
c.c. c.c.
10.18-10.33 212 224 95
10.59-11.14 551 645 85
11.56-12.11 205 305 67
12.32-12.47 530 628 84
1.33- 1.49 188 302 62

2.13— 2.28 240 321 75

Influence of Cold Baths on Glycogen Content of Man. 433

TABLE, |e
= 2.206so. wm. Hetcut = 176.5 cm. Acer, 44 YEars.

Calories per hour. Calories per day. euleake
ae P 1 pee Remarks.
Per kgm. | “surface | Per kem. | “surface.
1.08 S12 25.9 893 5.000 Normal.
1.39 47.7 | 33.3 1145 4.794 After Bath I.
1.20 41.0 28.6 985 4.867 Normal.
1.60 55.4 38.6 1329 4,708 After Bath IT.
0.99 33.9 233 815 4.708 Normal.
1.23 42.4 29.6 1018 4.867 After food.
TABLE II.

= 1.982 sq. M. Hetcut = 170.5 cm. Acer, 26 YEARS.

EES a ee nn ee ge nn ea ae aT
Calories per hour. Calories per day. Calorific
value Remarks.
Per sq. m. Per sq.m. | 1 litre O2
Per kgm. surface. Per kgm. surface.
1.03 33.5 24.6 804 4,954 Normal.
2.89 94.4 69.4 2265 4.831 After Bath I.
1.30 42.4 Sled 1017 4,589 Normal.
2.81 91.7 67.4 2202 4.819 After Bath IT.
Normal.
After 50 gm.
1.40 45.8 So 1099 4,708 oes

434 ~~ Graham Lusk.

During the shivering which followed the first cold bath the metabo-
lism was increased 181 per cent, and the respiratory quotient was
0.85, which indicated that the oxygen distribution was divided as
follows: for protein 15 per cent, for carbohydrates 44 per cent, for
fat at per cent. During the next resting period the quotient was 0.67,
that found by Zuntz during complete rest in a fasting man and in-

TABLE III.

Person C. Weicut = 70.8 xc. AREA OF Bopy SURFACE

Per minute.

Period.

COz O2

c.c. c.c,

10.45-11.00 266

11.45-12.00 694 865
12.47— 1.02 314
1.46— 2.01 int
2.48— 3.03 320
~ 4,15- 4.30 297

terpreted by him to signify a storage of glycogen derived from pro-
tein. During the.period following the second bath the quotient
rose to 0.84, and this can only be interpreted as due to a further
oxidation of glycogen. During the subsequent period of rest the
quotient sank to 0.62. Such quotients have been found after exhaus-
tive exercise. Thus Durig!” states that he was forced to take sugar
on reaching the summit of a mountain, in order to obtain a respira-
tory quotient which was interpretable. In the present research it
will be noticed that the administration of 50 gm. of dextrose to
Case B caused the quotient to rise within half an hour from 0.62.
to 0.75.

It is apparent from this discussion, that the treatment of Case B
by cold baths had a much more pronounced effect than on Case A.
The difference lies in the relative difference in weight. If the normal

16 ZuNTZ: Loc. cit.
17 Duric: Loc. cit., p. 78.

Influence of Cold Baths on Glycogen Content of Man. 435

weight of a person in kilograms is the number of centimetres of his
height less one hundred, then Case A was of normal weight, whereas
Case B was nearly 6 kgm. under the normal. The difference in adipose
tissue represented by this difference in weight accounts for the
greater readiness of heat loss in Case B when subjected to cold
baths.

TABLE III.
=—2105 50. Ms Hercnr = 173.2 cu. AGE, 23 YEARS.

Calories per hour. Calories per day. Calorific

ee Ve Remarks.
Per kgm. | Persq.m. | Per kgm. | Per.sq. m. Saabs:
surface. surface.

1.26 42.4 30.2 1017 4.869 Normal.

3.50 117.9 84.0 2829 4.770 | After Bath I.

1.29 43.6 31.0 1046 4,844 Normal.

2.88 97.2 69.1 2332 4.783 After Bath II.

1st 44.3 31.4 1063 4.844 Normal.

1.22 41.3 29.3 991 4.856 piteet aeotei

food.

Case C.— December 11, 1910. Noon, usual dinner. 5 P. M. to 6 P. M.,
hard exercise. 6 P.M., supper of meat, fat, and a single piece of
toast. Sound sleep for eight hours during the night.

December 12, 7.30 A.M. Rose, drank some black coffee without
sugar, and came to the laboratory.

Period I. — 10.45 to 11.00 A.M. Normal resting period; pulse, 68.

First Bath, 11.28 to 11.40 A.M. Began to shiver at 11.44. Immer-
sion in bath with ice blocks for twelve minutes. Temperature of baths
rose from 8° to 10° during the experiment.

Period IT. — 11.45 to 12 noon. Subject shivered throughout the entire
period.

Period IIT, — 12.47 to 1.02 P.M. Normal resting period. Subject does not
feel as warm as before taking the bath.

Second Bath, 1.29 to 1.41 P.M. Immersed in water with ice blocks;
temperature of water, 8°. Shivering.

Period IV.— 1.46 to 2.01 P.M. Violent shivering throughout the entire
period. Pulse, 85; blood pressure, max. systolic = 144, min. diastolic
= 82.

436 Graham Lusk.

Intermission. Hot bath at 44° taken for fifteen minutes in order
to restore the feeling of warmth.
Period V.— 2.48 to 3.03 P.M. Normal resting period. Pulse, 80; blood

pressure, max. systolic = 92, min. diastolic = 60.
Food, 3.15 Pp. M. Large beefsteak, soup, and black coffee without
sugar. ,

Period VI. — 4.16 to 4.30 P.M. Normal resting period.

This individual had a well-developed musculature, and was a
highly trained and accomplished base-ball player. There was no sur-
plus fat on his body. It is evident, from the respiratory quotients
obtained, that in spite of intense shivering caused by the two cold
baths, the treatment was not sufficient to remove the glycogen from
the highly developed musculature of this individual. The results
are shown in Table III.

An interesting detail is that during the first hour following the in-
gestion of beefsteak no increase in the heat production was noted.

In all three individuals experimented on, the skin became in-
tensely red during the bath in ice water. This may be explained as
being due to paralysis of the contractile elements of the superficial
capillaries caused by the influence of cold on the protoplasm. This
does not necessarily indicate an increased blood flow to the skin area,
for the superficial arterioles are probably so constricted as to reduce
the blood flow to. the skin. The blood pressure rises considerably,
but its greatest height could not be obtained because the profound
shivering prevented taking the pulse.

Only these three experiments were made, for it was fully realized
that the ordeal of being immersed amid cracked ice was not to be
lightly undertaken. No ill effects whatever were noticed in any of
the men who served as subjects. The only after effect observed was
that of general muscular lassitude.

CONCLUSIONS.

1. Immersion of normal men in cold baths at a temperature of
10° when the intestine is free from carbohydrates, induces shivering
which causes a rapid utilization of body glycogen as determined by
a fall in the respiratory quotient to the fasting level. In one very
muscular individual this result could not be obtained.

Influence of Cold Baths on Glycogen Content of Man. 437

2. In one individual in whom the shivering had been severe, a
quotient of 0.67 and another of 0.62 were found during subsequent
periods of rest, which correspond to those observed during rest after
a period of exhaustive exercise (glycogen formation from protein).

3. The greatest increase in heat production which was brought
about by the cold baths was 181 per cent above the normal. The
urine remained free from albumin and from sugar.

ON NUCLEIN METABOLISM IN THE DOG.

By P. A. LEVENE anp F. MEDIGRECEANU.
[From the Rockefeller Institute for Medical Research.]

iC lens very large recent literature on nuclein metabolism abounds
in controversial statements and contains conclusions apparently
irreconcilable with one another. However, disregarding the personal
attitude of the writers to the results of their own experiments, and
viewing them in the light of impartial analysis, it is possible to formu-
late a very definite conception regarding certain phases of nuclein
metabolism. Particularly the final phases in the long cycle of nuclein
metamorphosis have been made obvious. It has been established,
principally through the work of Schittenhelm and his coworkers,
Wiechowski, and Mendel and his coworkers, that the end products of
purin metabolism are uric acid, allantoin, and urea. The intermediate
stages between allantoin and urea are not known. Whether or not
the end product of purin deterioration is only one for each animal or
for each species, is answered differently by individual writers. By
means of chemical manipulations it is possible to dismember the com-
plex nucleic acids into nucleotides, these further either into purins
and carbohydrate phosphoric acids or into phosphoric acid and nu-
cleosides. It is also possible to remove the purins from the nucleic acid
molecule before it suffers any other alteration in its composition.

It will always remain a very difficult task to ascertain all the exact
phases through which a nucleic acid passes in the organism on its way
to transformation into uric acid and into still simpler bodies. The meth-
ods which can be employed for the solution of this problem are many,
and perhaps alone none of them could lead to a decisive answer. It is
possible, therefore, that more than one method will have to be employed
before all the phases of the nuclein metamorphosis will become known.
One of the ways capable of bringing a certain amount of light on this
process consists in a comparison of the proportions of uric acid, allan-

438

My Paitin

a a ae Te

On Nuclein Metabolism in the Dog. 439

toin, or urea eliminated after administration of nucleic acid on one
hand, and its more or less complex components on the other hand.

It seems most probable that all organisms possess the capacity of
furnishing all the three named end products of purin metabolism, and
it is absolutely certain that in definite species of animals the predomi-
nating end product is uric acid, and in others allantoin, or still in others
urea. The relative proportion of these substances eliminated in the
urine is to a degree influenced by the condition of the animal and by
the character of the diet. However, the purins compose only one
component of the nucleins and even of nucleic acids. The phases
through which nucleic acid has to pass in order that its purin may be
liberated and further metamorphosed are not known. On the other
hand, owing to the progress in the knowledge of the chemistry of
nucleic acid, all the possible intermediate stages have become obvious.
Through the work of Jacobs and one of us, it has become known that
' purins enter the molecule of nucleic acid in form of nucleotides. These
are composed of phosphoric acid and a purin pentosid or a purin
glucosid.

‘Experiments with feeding nucleic acid and its derivatives of various
degrees of complexity have been performed by earlier investigators.
Unfortunately they were not carried out systematically, and many
of them were made at a time when the methods of analysis were im-
perfect and the knowledge of the constitution of nucleic acids was even
less perfect.

The general impression gained from this work is that after ad-
ministration of nucleic acid the output of uric acid or of allantoin is
greater than after ingestion of its decomposition products. However,
it is not possible to accept without further investigation the results
of the majority of workers, for the reason that in the analysis of their
results they failed to take into account the influence of the ingested
nuclein derivatives on the general metabolism. Even the most recent
investigators frequently omitted these considerations. And yet it
is evident that a definite estimate of the quantitative transformation
of any given purin derivative into uric acid, allantoin, or other final
decomposition product, cannot be obtained if the administration of
that derivative caused the output of nitrogen to rise above the intake.
In such experiments it is impossible to determine the part of the in-
creased nitrogen output which may be referred to the administered

440 P. A. Levene and F. Medigreceanu.

substance. Thus it is evident that only the experiments in which the
health of the animal was not affected by the administration of nuclein
derivative can be taken into consideration. The records of such ex-
periments are very few.

Another very frequent occurrence after the administration of purins
is the failure to produce any impression on the total nitrogen out-
put. This may be best illustrated by an experiment of Kruger and
Schmidt. It was performed on a man maintained on a purin-free
diet; 3.0 gm. of hypoxanthin containing 1.236 gm. of nitrogen, were ad-
ministered. ‘This raised the average uric acid nitrogen output from
0.1533 gm. to 0.346; or for four days the increase in nitrogen output
caused by the high uric acid output amounted to 4 (0.346 — 0.1533)
gm. = 0.7708 gm., or 62.3 per cent of the ingested hypoxanthin.
On the other hand, the total nitrogen output shows the following
values: the average normal nitrogen output was 10.89 gm.; after
hypoxanthin feeding, 10.94 gm. The excess in four days amounted to
4 (10.94 — 10.89) gm. = 0.2 gm. It is evident that in a similar ex-
periment it is impossible to establish the origin of the uric acid.
There is an abundance of similar records. In reality they should
not be considered when an attempt is made to establish the actual
process of purin metabolism.

In the present investigation an attempt was made to maintain the
animals in nitrogenous equilibrium between experiments. No new
experiments were performed before the animal returned to its normal
condition. It was noted in course of the experiments that adminis-
tration of sodium carbonate simultaneously with the nuclein deriva-
tives averted all undesirable influences.

The substances employed in the experiments were allantoin, uric
acid, hypoxanthin, inosin, and thymus gland.

The urine was analyzed for the following substances: total nitrogen,
uric acid, purin bases, ammonia, amino nitrogen, and allantoin. In
order to investigate the possibilities of the intermediate formation
of glycocol from purin bases, an attempt was made to ascertain the
output of the uric acid after the administration of sodium benzoate
before and during the experiment to be described. (See H. Wiener,”

1 KRUGER and Scumipt: Zeitschrift fiir physiologische Chemie, 1902, xxxiv,
p. 558.

7H. Wiener: Archiv fiir experimentelle Pathologie und Pharmakologie,
1897, Xi, Pp. 313.

On Nuclein Metabolism in the Dog. 4AT

Hirschstein,? Wiechowski,* Brugsch and Schittenhelm,? Abderhalden
and Guggenheim.®) The most favorable results were obtained on ad-
ministration of 4 gm. sodium benzoate and 2 gm. sodium bicarbonate.
It was noticed that after administration of the substances the nitrogen
output increased from o.2 to 0.3 gm. per day. The daily quantities
of urine were obtained by means of catheterization.

METHODS OF ESTIMATION.

Nitrogen . . . After Kjehldal-Gunning.

MBG el ie, ss By a modification of the method of Benedict and Gep-
dart.”

Uric acid . . . After Ludwig Salkowski.

Purin Bases . . Obtained from the filtrate of the uric acid precipitated

by means of mercuric sulphid. The nitrogen estima-
tion was made on the precipitate.
Amino Nitrogen Estimated by the method of Van Slyke.8
Ammonia... By the method of Folin.
Allantoin . . . After the new process of Wiechowski.°

Regarding allantoin estimation it should be noted that the esti-
mation was made on the crystallized substance dried at 100° C. The
substance was identified by the melting point, which varied between
223° and 225°. Pure allantoin obtained from uric acid was estimated
at a point of 225° C. The use of charcoal for purification of the sub-
stance was avoided, since such treatment leads to a loss of the sub-
stance. Pure allantoin was obtained by repeated precipitation with a
mixture of mercuric acetate and sodium acetate and by repeated
crystallization. In order to test the method, analyses were made on

3 HirscHsteEIn: Zeitschrift fiir experimentelle Pathologie und Therapie, 1907,
iV, pe 110:

4 WIECHOWSKI: HOFMEISTER’S Beitrige, 1906, vil, p. 204.

5 BrucscH and ScHITTENHELM: Zeitschrift fiir experimentelle Pathologie und
Therapie, 1907, iv, p. 540.

6 ABDERHALDEN and GUGGENHEIM: Zeitschrift fiir physiologische Chemie, 1909,
jix, p; 20.

7 LEVENE and Meyer: Journal of the American Chemical Society, 1900, xxxi,
Pp: 717.

8 VAN SLYKE: Proceedings of the Society for Experimental Biology and
Medicine, 1910, vii, pp. 46-48; Berichte der deutschen chemischen Gesellschaft,
1910, xliii, p. 3170.

* WiEcHOWSEI: Biochemische Zeitschrift, 1910, XXV, Pp. 431. »

442 P. A. Levene and F. Medigreceanu.

TABLE I.

Urine.

Total N Remarks.

gm. sodium benzoate.
gm. sodium bicarbonate.
.8 gm. monosodium urate
=10gm.N

.0 gm. sodium benzoate.
.0 gm. sodium bicarbonate.}

4.0 em. aaa benzoate.
2.0 gm. sodium bicarbonate.

re 841 “ay allantoin = 1.0

4.0 em. sodium benzoate.
2.0 gm. sodium bicarbonate.

{3 .0 gm. inosin = 0.83 gm.

The dog lost its appetite and
appeared ill.

4 gm. sodium carbonate.

gm. N instead of 14 gm.
plasmon = 1.61 gm. N.

_. 5S gms. Na2CO;

\ 2.514 gm. hypoxanthin =

) 1.0 gm. N.
5.0 gm. NazCOs. }

5 gm. Na2CO3.
4.784 gm. inosin

92 gm. veal thymus = 1.61

gm. N.
5 gm. Nae2COs3.

Datty DIET.

June 15-July 1: Plasmon, 14 gm. = 1.66 gm. N; cracker meal, 100 gm. = 1.59
gm.N; sugar, 20gm.; lard,10gm. Total N,3.25gm. Approximate calories, 700.

Sept. 5-Sept. 7: Plasmon, 14 gm. = 1.61 gm. N; cracker meal, 100 gm. = 1.77
gm. N; sugar, 40 gm.; lard, 10gm. Total N, 2.38 gm. Approximate calories = 800.

Sept. 20-Sept. 29: Plasmon, 14 gm. = 1.61 gm. N; cracker meal, 100 gm.
= 1.77 gm. N; sugar, 60 gm.; lard, 15 gm. Total N = 3.38 gm. Approximate.
calories = 900.

Date.

June.

On Nuclein Metabolism in the Dog.

Urinary Nirrocen Partition. NITROGEN IN GRAMS.

| Uric

443
TABLE II.
ete | Amino N. | Allantoin. pela

Traces 0.248 0.058 0.170
Traces 0.101 0.088 0.200
{ 0.232 0.152 0.219

0.005
0.111 0.120 0.162
{ 0.085 0.066 0.214

0.005
0.114 0.081 0.203
Traces 0.097 0.046 0.186
Traces 0.206 0.384 0.190
Traces 0.103 0.088 0.213
Traces 0.097 0.059 0.175
0.003 0.088 0.077 0.169
0.017 0.232 0.388 0.233
0.003 0.235 0.070 0.320
0.008 0.106 0.084 0.171
0.008 0.102 0.266 0.151
0.011 0.101 0.111 0.138
0.007 0.090 0.092 0.136
0.006 0.101 0.204 0.140
0.007 0.099 0.111 0.144
0.006 0.092 0.115 0.098
0.007 0.100 0.568 0.162
0.004 0.096 0.124 0.109
0.012 0.110 0.149 0.163
0.036 0.107 0.566 0.092
0.004 0.097 0.181

Urea! | Tea: acid.
2.66 0.175 Traces
2.64 0.092 Traces
305. | 0.192 i
0.029

2.95 0.130
2.40 | 0.132 \

: 0.014
2.48 0.144
2.28 0.169 Traces
3.45 0.192 0.009
2.53 0.133 0.008
Z.02 0.170 Traces
2.16 0.173 0.007
3.09 0.195 0.043
2.39 0.135 0.007
2.56 0.220 0.015
2.84 0.224 0.005
2.49 0.201 0.009
2.90 0.199 0.008
2.81 0.252 0.011
2.38 0.260 0.010
2.76 0.098 0.006
Sea, 0.094 0.017
2.64 0.185 0.006
2.38 0.126 0.009
2.18 | 0.090 | 0.085
2.48 0.182 0.006

1 The values for urea nitrogen include allantoin nitrogen.

444 P. A. Levene and F. Medigreceanu.

human urines to which o.2 gm. of pure allantoin was added. The
added allantoin was recovered nearly quantitatively. The loss seldom
exceeded 0.02 gm.

Percentage transformation of the fed purin was calves on the
basis of nitrogen eliminated in the feeding experiments in excess over
the nitrogen output in the normal periods.

RESULTS OF EXPERIMENTS.

J. Allantoin. — One gram of nitrogen fed in form of allantoin was
removed by the dog in the course of twenty-four hours; 31 per cent
of it was unchanged, and the rest oxidized to urea.

II. Sodium urate. — In the course of two days 60 per cent of the
nitrogen introduced in this form was eliminated by the urine, 15
per cent in form of allantoin, 2 per cent in form of the unchanged sub-
stance, and the rest in form of urea.

III. Hypoxanthin. — After the administration of 1 gm. of nitrogen
in form of hypoxanthin, 0.56 gm. were removed, 80 per cent as allan-
toin, 2 per cent as uric acid, and the remainder as urea.

IV. Inosin experiments. — In the course of twenty-four hours 0.83 gm.
of nitrogen ingested in form of inosin (4 gm.) were removed; 40 per
cent of it in form of allantoin, 4 per cent as uric acid, 2 per cent as
purin, 3 per cent as ammonia, 4 per cent as undetermined nitrogen,
and the rest as urea. It should be noted that the feeding of solutions
of hypoxanthin and inosin in water was frequently followed by dis-
turbances of nitrogenous equilibrium, lasting for a considerable time.
These disturbances were avoided when, simultaneously with the hy-
poxanthin, sodium carbonate was administered. After the adminis-
tration of hypoxanthin, inosin, and yeast nucleic acid, simultaneously

with sodium carbonate, there was always noted a retention of nitrogen.

Thus, after the administration of the 1 gm. of nitrogen in form of
inosin, 0.6 gm. were removed in course of the first twenty-four hours,
75 per cent in form of allantoin, 13 per cent as uric acid, 5 per cent
in form of purin bases, and 8 per cent in form of urea.

V. After the administration of nucleic acid in a quantity contain-
ing 0.6 gm. nitrogen, of which 0.4 were in form of purin nitrogen, there
reappeared in the urine in the course of the first twenty-four hours
0.3 gm. of nitrogen. Calculating on the basis of the nitrogen dis-

On Nuclein Metabolism in the Dog. 445

tribution in the nucleic acid, o.2 gm. of the total excessive output have
to be attributed to the purin nitrogen. Of this value 85 per cent were
removed in form of allantoin, the remainder in form of urea.

VI. After the administration of thymus containing 0.6 gm. purin
nitrogen there were removed in the course of twenty-four hours fol-
lowing the injection 17 per cent in form of allantoin, 5 per cent as
uric acid, and the rest as urea. There was no increase in the amino
nitrogen output after any one of the experiments.

From the results of these experiments it is apparent that the highest
proportion of allantoin output follows the administration of nucleic
acid and of hypoxanthin; the proportion is lower after the adminis-
tration of inosin. Thus it seems possible that the first step in the dis-
integration of nucleic acid in the organism is the liberation of purins
and not of inosin. Experiments of a totally different nature, which
will be published later, have made a similar conclusion suggestive.

We realize that further experiments wiil be necessary before this
conclusion can be definitely established.

THE EFFECTS OF VARIOUS FORMS OF EXERCISE ON
SYSTOLIC, DIASTOLIC, AND PULSE PRESSURES AND
PULSE RATE. .

By OSWALD S. LOWSLEY.
[From the Physiological Laboratory, Johns Hopkins Medical School.]

INTRODUCTION.

cue the discovery of reliable methods of determining blood

pressure in man, numerous attempts have been made to apply
these methods to the study of the effects of exercise.’ Most of these
investigations have agreed in finding that there is a rise in blood pres-
sure as well as an increase in heart rate during exercise, but that sub-
sequently the blood pressure may sink to a subnormal level. In the
recent paper by Barach, published after the results of the present
study were nearly completed, it is stated that a fall below normal of
about 20 per cent occurs in cases of violent exercise (Marathon racing)
immediately at the end of the exercise, at a time when the pulse rate
is still above normal. My results, on the contrary, indicate that even
after the most exhaustive exercise a rise of pressure, systolic as well
as diastolic, may still be observed at the end of the exercise, although a
long lasting fall of pressure ensues shortly afterward. In the present
paper advantage has been taken of an excellent opportunity for such
work to study the immediate and after effects of various forms of
exercise upon systolic and diastolic pressures, pulse pressure, and
heart rate.

The records were secured by the use of the Erlanger sphygmo-
manometer.”

1 McCurpy: This journal, 1901, v, p. 95; BowEN: Ibid., 1904, xi, p. 60; PEM-
BREY and Topp: Journal of physiology, 1908, xxxvii, p. Ixvi; EpGrecomBE and
BAIN: Ibid., 1899, xxiv, p. 48; GorDoN: Edinburgh medical journal, 1907, xxii,
p. 53; Krone: Miinchener medicinische Wochenschrift, lv, p. 69; PoTTreR and
HARRINGTON: Journal of the American Medical Association, 1909, lili, p. 1957;

Baracu: Archives of internal medicine, 1910, p. 382.
2 This journal, 1902, vi, Proceedings of the American Physiological Society,

p. Xxii.
446

The Effects of Various Forms of Exercise. 447

The subjects experimented upon were all healthy men who were in
the habit of exercising regularly. With the exception of a few indi-
viduals experimented upon in Group I of Part II, the subjects were
young athletes in the midst of a training season. Records were
taken at all times of the year, varying from the middle of summer
(twenty-mile race) to the dead of winter. Several practice runs
(thirteen miles, one hundred yards) and one hard race (ten miles)
were made in very cold weather with snow on the ground. An average
of about one-half minute elapsed between the completion of all exer-
cises considered in Part II and the determination of systolic pressure.

The records in Part I were made during exercise on the stationary
bicycle. The normal pressures and heart rate for each individual
were obtained by averaging a number of readings taken at different
times; none of these readings were made within the twenty-four hours
preceding a race, in order to avoid the stimulating effect of the excite-
ment attendant upon such contests.

PAR <1,

Tue Systoric, D1asToLic, AND PULSE PRESSURES AND PULSE RATE
DURING EXERCISE.

This part of the investigation includes records taken during short
fast sprints on a stationary bicycle, longer somewhat slower bicycle
riding, and finally during a slow ride upon a bicycle following upon a
walk of eighteen miles. In one such case there was an interval of
rest of ten minutes between the walk and the ride, in a second case
there was a rest of one hour.

Systolic pressure. — In all of the experiments except the last men-
tioned, there was an immediate and rather great rise in systolic pres-
sure, which reached a maximum in from five minutes or less to twenty-
five minutes, an average in nine cases of 14.33 minutes after the com-
mencement of exercise. The rapidity of this rise seems to depend on
the physical condition of the subject at the time of the experiment.
This fact is shown clearly in the cases of O.S.L.and R.A. K. O.S. L.
reached a maximum of systolic pressure in twenty minutes after the
commencement of work when in a perfectly fresh condition, but after
walking eighteen miles and then resting for ten minutes, allowing the

448 Oswald S. Lowsley.

systolic pressure to return to normal, and then riding the bicycle, a
record taken at the end of five minutes showed that the systolic pres-
sure had already reached its maximum. R. A. K. when in a perfectly
fresh condition reached a maximum of systolic pressure on one occa-
sion in seventeen minutes, and on another in twenty-five minutes.
After walking eighteen miles and then resting for one hour his systolic
pressure, which had dropped to normal, had already reached a maxi-
mum at the end of eight minutes of work.

Following the maximum rise there usually was a slight fall which
seemed to be due to fatigue, and a secondary rise was possible if the
efforts of the subject were increased.

The extent of the rise in systolic pressure was greatest in the case
of R. A. K., who showed an increase of 65 mm. Hg at the end of sev-
enteen and one-half minutes of riding. The smallest increase noted
was in the case of the same subject after a rest of one hour following
an eighteen-mile walk. This increase amounted to 10 mm. Hg eight
minutes after commencing the exercise. The average rise of systolic
pressure in sixteen observations was 32.7 mm. Hg.

Diastolic pressure. — The diastolic pressure showed a distinct rise
during exercise, although not nearly so much as the systolic. The
maximum diastolic pressure was reached rather later than the systolic
in some cases, but generally it occurred at the same time. The maxi-
mum diastolic pressure was reached at from five to sixty-five minutes
after the beginning of exercise, the average of eight experiments being
20.5 minutes.

The rise in diastolic pressure was greatest in the cases of O. S. L.,
who showed a rise of 40 mm. Hg thirty minutes after commencing
work, and C. R. S., who also showed a rise of 40 mm. Hg. The least
rise was in the case of R. A. K., who showed no change at all.
R. A. K., McL., and J. O. showed a rise of 1o mm. Hg. The average
rise of diastolic pressure in seventeen experiments was 22.9 mm. Hg.

Pulse pressure. — The pulse pressure in over 50 per cent of the
seventeen observations followed rather closely the fluctuations of the
systolic pressure. In the remaining experiments there was an initial
fall in pulse pressure followed by a subsequent rise to a maximum.

The greatest rise in pulse pressure occurred in the case of R. A. K.,
who showed an increase of 65 mm. Hg after seventeen and one-half
minutes of fast riding. Four experiments showed an increase of only

a eee

The Effects of Various Forms of Exercise. 449

5 mm. Hg. One of these was R. A. K. after the eighteen-mile walk
and one hour’s rest. Almost immediately after the cessation of work
there resulted a subnormal pulse pressure due to the fact that the sys-
tolic pressure dropped more rapidly than the diastolic. The average
rise of pulse pressure in seventeen experiments was 18.33 mm. Hg.

Pulse rate. — Immediately after work commenced there was a sudden
increase in the pulse rate, which reached a maximum in from five min-
utes or less to seventy-five minutes after starting. The average time
before the maximum rate occurred in eight observations was 35.4 minutes.

The greatest increase in pulse rate occurred in the case of McL., a
young strong athlete,, whose rate increased from 65 to 165, an increase
of 95 per minute five minutes after commencing the ride.

The smallest increase occurred in the case of R. A. K., who had a
pulse rate of 85 before walking for five hours. After resting for one
hour this rate had decreased to 78, and after fifty-three minutes of
slow riding it increased to 90 per minute. The average increase of pulse
rate in nine experiments was 51 per minute.

It is to be noted that after the primary increase in pulse rate there
was not a great deal of variation, even though the maximum was not
reached for some time. Immediately after the cessation of work there
was a rapid fall, which was more rapid after a short period of exertion
and usually became subnormal. After longer periods of exercise, par-
ticularly running, there was a much slower fall to normal but rarely
to a subnormal pulse. This interesting difference may be explained
perhaps by the fact that the acceleration of the heart rate is referable
to two factors: first, a reflex effect through the accelerator nerves, a
factor which it may be supposed is responsible for the initial effect of
exercise and which would fall away promptly upon the cessation of
the exercise; second, a metabolic effect due to the accumulation of
acid products in the blood which then react upon the accelerator
centre. We may assume that this effect is slow in developing and also
slow in disappearing. The long after acceleration noticed in prolonged
exercise may be attributed to this factor.

SUMMARY.

The conclusions to be deduced from these observations are: 1. Ex-
“ ercise causes an immediate rise in systolic pressure, but the maximum

450 Oswald S. Lowsley.

attained may occur some time after the exercise is begun. As fatigue
advances the systolic pressure falls, but it may be caused to rise
_ again by a greater effort on the part of the subject. Cessation of
activity causes a very rapid return to normal and in almost every
case to subnormal. A maximum systolic pressure is reached more
rapidly in the case of a fatigued individual but is not nearly so
extensive.

2. The maximum diastolic rise during exercise is generally reached
at the same time as the systolic maximum, although sometimes it
occurs later. The average rise in systolic pressure exceeds the average
rise in diastolic by about 1o mm. Hg. The diastolic pressure fluctuates
very little as a rule after its maximum has been reached. It returns to
normal rather more slowly than the systolic and invariably shows a
fall to subnormal after exercise. -

3.° The pulse pressure curve generally follows the contour of the
systolic curve, due to the fact that systolic pressure fluctuates more
than diastolic pressure.

4. The pulse rate increases rapidly at first, but usually does not
reach a maximum for some time. After the initial rise it does not
change greatly, although it is evidently influenced by the fatigue of,
and the effort expended by, the subject.

If a considerable time elapses between the completion of prolonged
fatiguing exercise and renewed effort, the new exercise causes but little
acceleration of the heart, while, if there is no interval or only a slight
interval of rest, there is a perfectly definite effect produced on the
heart (acceleration) and blood pressure (rise).

PART if.

AFTER EFFECTS OF VARIOUS FoRMS OF EXERCISE ON SYSTOLIC,
DIASTOLIC, AND PULSE PRESSURES AND PULSE RATE.

The subject here considered may be conveniently arranged under
five heads as follows:

I. Moderate exercise for a considerable time.
II. Rapid exercise for a short time.
III. Vigorous exercise.
IV. Fatiguing exercise.
V. Exhaustive exercise.

The Effects of Various Forms of Exercise. 451

I. MOopERATE EXERCISE FOR A CONSIDERABLE TIME.

The six men upon whom the records for this experiment were made
varied in age from nineteen to fifty-nine years, and the types of exer-
cise used were swimming and playing in the water, tennis, baseball,
shot putting, jumping, hammer and discus throwing. All of the exer-
cises were performed for the purpose of recreation, no match games
or attempts to make records entering into consideration. The physical
effort thus made was continued from about half an hour to two hours.

Systolic pressure. “%; Immediately after such exercise there was a
slight rise in systolic pressure, rarely over 10 mm. Hg, and only in one
case as much as 15 mm. Hg. Following this there was invariably a
drop to subnormal, the extent of which in a large number of cases was
20 mm. Hg and in one case 30 mm. Hg.

Diastolic pressure showed a more marked rise in three out of
six experiments than did systolic. In all three cases the systolic
pressure recorded immediately after exercise showed a rise of 10 mm.
Hg, while the diastolic showed a rise of 20 mm. Hg. In one of these
experiments the subject was a man of forty-nine years, and in another
a lad of seventeen.

Two of the other three cases showed a diastolic rise of to mm. Hg,
and the third showed no rise in diastolic pressure, although a systolic
rise of 15 mm. Hg was present after two hours of.shot putting, discus
throwing, and broad jumping.

Accompanying the fall in systolic to subnormal there occurred a fall
in diastolic pressure varying from 5 to 20 mm. Hg. The fall in dias-
tolic pressure was slower than systolic. One case in which after two
and one-half hours of moderate exercise, systolic pressure had dropped
to normal still showed a diastolic pressure of 5 mm. Hg above normal
for fifteen minutes. In another case a normal diastolic pressure was
reached twenty minutes after a swim of three quarters of an hour, at
which time the systolic pressure was 20 mm. Hg below normal.

Pulse pressure varied greatly after this type of exercise. Three
of the cases showed a drop in pulse pressure from 10 to 15 mm. Hg
below normal, two others gave no change, and another a rise of 15 mm.
Hg. During the subnormal period there was invariably a fall in pulse
. pressure ranging from 5 to 20 mm. Hg, due to the fact that systolic
pressure fell more rapidly than diastolic.

452 Oswald S. Lowsley.

Pulse rate increased from 16 to 36 per minute above normal
immediately after exercise, and during the period of subnormal pres-
sures was still above the normal rate in all, except one case, in which
the rate had dropped to normal five minutes after playing in the water
for three quarters of an hour.

Two young athletes in good condition worked for thirty-five and
forty minutes respectively in the gymnasium. They tossed a medicine
ball for about fifteen minutes and then did combination work on the
side horse and parallels very much as a regular gymnasium class
would do.

Systolic pressure in both cases showed a rise of 10 and 25 mm. Hg
respectively. Diastolic pressure at the end of exercise in the former
case (A) was 20 mm. Hg above normal, and in the latter (B) 15 mm.
Hg. Pulse pressure fell 1o mm. Hg in one case (A) and rose 10 mm.
Hg in the other (B). Pulse rate increased 15 in A and 35 in B.

The subnormal stage was not particularly marked in A, but was
more noticeable in B, reaching a systolic subnormal of 1o mm. Hg
thirty-five minutes after exercise. Diastolic pressure did not become
subnormal in either case. Pulse pressure was 10 mm. Hg subnormal
in A at the end of exercise; it returned almost immediately to normal
and again dropped to below after walking ten squares slowly. This
walk had no effect on systolic pressure in either case. In B the mini-
mum of pulse pressure (10 mm. Hg below normal) came thirty-five
minutes after the completion of exercise. Pulse rate became subnor-
mal in A one hour and twenty-two minutes after exercise and was not
marked by a subnormal stage in B.

Return to normal. — Systolic pressure returned to normal in A forty-
eight minutes after exercise, and in B one hour and twenty-seven
minutes after. Diastolic returned to normal at eight minutes in one
case and thirty-five minutes in the other. Pulse pressure was at nor-
mal in forty-eight minutes in A and only 5 below normal in fifty-five
minutes in B. Pulse rate returned to normal in one hour and forty-
two minutes in A and in forty-five minutes in B.

Four young athletes ran three miles to a swimming pool, swam and
played in the water twenty minutes, rested, then walked two miles
and finally ran the last mile to the club house. One of the four had
a supranormal systolic and also diastolic pressure of 10 mm. Hg
immediately after finishing and a pulse rate of 124 per minute. Forty

The Effects of Various Forms of Exercise. 453

minutes later his systolic pressure was 5 below normal and his diastolic
pressure was exactly normal, and the pulse rate had fallen to 96 per
minute. Two of the others had a subnormal systolic pressure of 35
mm. Hg and 5 mm. Hg respectively, and the diastolic pressure of the
former registered 5 below normal and the latter 5 above. The pulse
pressure in the former dropped from 35 to 5,? and the pulse rate changed
from 56 at normal to 84 after the test. The pulse pressure in the case
of the latter dropped from 25 to 15, and the pulse rate rose from 80
to 120. The fourth member of this group, whose record was taken five
minutes after finishing, showed a systolic pressure 20 mm. Hg below
normal, a diastolic pressure 5 mm. Hg below normal, a pulse pressure
15 mm. Hg below normal, and ‘his pulse rate had dropped from a
normal of 84 to 76.

II. Rapmp EXERCISE FOR A SHORT TIME.

The three young men performing these experiments ran 100 yards
as fast as they could, and the records, taken as quickly as possible
after the exercise, showed an average rise in systolic pressure of 45
mm. Hg and in diastolic pressure of 17 mm. Hg. (One rose 5 mm.
Hg, another 20, and the third 25.) The pulse pressure rose on an
average of 28.3 mm. Hg, and the average pulse rate increased 45 per
minute.

The subnormal phase reached its lowest ebb in ten minutes in one
case, thirty in another, and fifty in the third. The amount of depres-
sion of systolic pressure in the first two cases amounted to 15 mm. Hg
below normal, while in the third case the fall was 25 mm. Hg below
normal. The diastolic pressure in case I fell to 15 mm. Hg subnormal
twenty-seven minutes after exercise. In case II 15 mm. Hg sub-
normal was recorded forty-five minutes after exercise, and case III
showed 10 mm. Hg subnormal thirty-eight minutes after. Pulse
pressure in case I amounted to 5 mm.* Hg twelve minutes after the
race. In case II twenty minutes after the completion of exercise the
pulse pressure was 10 mm. Hg below normal. Case III showed a pulse
pressure of only 5 mm.* Hg sixty-five minutes after exercise, which

3 A pulse pressure of 5 mm. Hg is very low. Probably the circulatory conditions
changed during the observations.

4 There is a possibility of pressure conditions having changed during these
observations, but there is no doubt that pulse-pressure was extremely low.

454

was 25 below normal.
any of these experiments.

Oswald S. Lowsley.

The pulse rate did not become subnormal in

The return to normal systolic pressure occurred in case I in one hour

and ten minutes.

120

PULSE RATE

Cases II and III were 10 mm. Hg below normal

one hour and ten minutes after ex-
ercise. Diastolic pressure in case I
was 10 mm. Hg below normal one
hour and ten minutes after exer-
cise, and ro mm. Hg below normal

in case II one hour and twenty-five
minutes after exercise. Normal was
reached in case III one hour and
five minutes after exercise. Pulse
pressure fluctuated in cases I and IT,
but in III was subnormal for over
one hour and thirty-five minutes
after exercise. Pulse rate returned
to 10 above normal one hour and
ten minutes after exercise in case I,
SYSTOLIC PRESSURE and in cases II and III returned
a to normal in one hour and one hour
and twenty minutes respectively.
Records taken on sixteen other
men and averaged with those just
considered yield the results given in
the table.

PULSE PRESSURE

DIASTOLIC PRESSURE

= & 10 30 50 70 90
ge
Ficure 1 (G 100 yds).— Shows the effect III. Vicorous EXERCcIsE.
of a short hard run upon the blood
pressures and heart rate. Ordinates =
mm. Hg pressure and rate per min-
ute. Abscissa = time in minutes.

The types of muscular activity
classed in this group include a hard
wrestling bout lasting eight min-
utes and runs varying from one and one-half miles to five miles.

The systolic pressure of both of the wrestlers rose 40 mm. Hg after
exercise (No. 1 ran until the record of No. 2 was completed). Dias-
tolic pressure rose 25 mm. Hg above normal in the case of No. 1 and
15.mm. Hg in that of No. 2. Pulse pressure accordingly rose 15 mm.
Hg in the first case and 25 mm. Hg in the second case. Pulse rate
increased 65 per minute in one case and 8o in the other.

The Effects of Various Forms of Exercise. 455

The subnormal phase was marked in No. 1 by a systolic drop to 15 mm.
Hg below normal thirty-six minutes after exercise. This was fol-
lowed by a rise to normal twenty minutes later, and there was then a
drop to about 20 mm. Hg. below normal which lasted for nearly an
hour. The chart of No. 2 demonstrated a similar fall in systolic pres-
sure of from 15 to 20 mm. Hg
lasting from one hour and ten min-
utes after exercise to one hour and
sixty-five minutes. Diastolic pres-
sure also showed a fall which came
on a little more slowly and was
less extensive. In No. rit dropped
to 15 mm. Hg below normal at one
hour and twenty-eight minutes
and again at two hours and
twenty-eight minutes after exer-
cise. In No. 2 its lowest drop
was to 1o mm. Hg. below normal
after one hour and forty-three
minutes. The pulse pressure was Fyicurr 2.— Shows the effect of vigorous
very irregular in No. 1, but in No. exercise sustained for several minutes
2 it remained considerably below *Testling 8 minutes) upon the blood

pressures. DW, drank four glasses
normal for two hours and twenty of water. B, bath. DB, drank one
minutes. The pulse rate did not bottle of beer. Ordinates = mm. Hg
show any considerable subnormal Pressure and rate per minute. Ab-
stage. | scissa = time in minutes.

The return to normal.— Systolic pressure returned to normal in
about two hours and forty minutes in one case and three hours and
five minutes in the other. Diastolic pressure became normal in No. 1
one hour and twelve minutes after exercise, but this was followed by
an extensive subnormal stage which had disappeared two hours and
forty-three minutes later. In No. 2 the diastolic pressure was irreg-
ular, but tended to remain below the normal for two hours.

Records of sixteen men who ran from 1.5 miles to 4 miles showed
an average systolic rise of 35 mm. Hg immediately after exercise.
The smallest rise was 10 mm. Hg and the greatest was 75 mm. Hg.
The magnitude of the rise appeared to be associated directly with the
effort of the final sprint. The average diastolic rise was 25 mm. Hg

456

Oswald S. Lowsley.

and ranged from 5 mm. Hg to 4omm. Hg. Pulse pressure in ten cases
increased on an average of 17.5 mm. Hg and in six cases decreased an

120

PULSE RATE

PULSE PRESSURE

120

SYSTOLIC PRESSURE

100

DIASTOLIC PRESSURE
80

ut
S
pra

BE
AFTER

FicureE 3.— Shows the effect of fatiguing
exercise (6-mile run) upon*the blood
pressure and heart rate. Ordinates =
mm. Hg pressure and rate per minute.
Abscissa = time in minutes.

average of 10.8 mm. Hg. Pulse
rate varied from no increase to
an increase of 70 per minute and

averaged for the sixteen cases an

increase of 39 per minute. The
case that showed no increase at
the end of a three-mile run was
recorded thirty minutes later and
at that time an increase of 36 per
minute was noted.
Subnormal. — Systolic pressure
always showed a subnormal stage
after this form of exercise, and in
thirteen of the sixteen subjects the
subnormal phase had developed
between thirty and forty minutes
after work. The average fall in
these cases at this time was
noted to be 18 mm. Hg, the great-
est being 30 mm. Hg and the
least 5 mm. Hg. From a study
of six of these cases, it is noted
that the minimum systolic pressure
occurred in runs of this distance
from thirty-five minutes to one
hour and ten minutes after the
completion of the exercise, averag-
ing in the six cases 62.5 minutes.
Subnormal diastolic pressure was
noted in ten out of sixteen cases,
but its extent was noteworthy
only in two cases in which the
fall was 15 mm. Hg below normal.

The average for the ten cases was 8.5. mm. Hg. Pulse pressure showed
a subnormal stage in fifteen cases, varying from 5 to 30 mm. Hg and
averaging 15 mm. Hg. Three cases did not show a subnormal pulse

The Effects of Various Forms of Exercise. 457

pressure during the time for which records were taken, but if the
time had been extended they probably would have done so, as this
period extends from ten minutes to one hour and fifteen minutes,
although in most cases it seems to come between twenty and forty
minutes after work. Pulse rate did not become subnormal to any
noteworthy extent.

Return to normal. — Systolic pressure seemed to return to normal
in about one hour and forty minutes in a well-trained man, but in
persons out of condition it took much longer. Diastolic pressure was
not so greatly disturbed, and although its decline was slower it returned
to normal more quickly than systolic. Pulse pressure followed rather
closely systolic pressure in this sort of exercise, and consequently
returned to normal at about the same time. Pulse rate in well-trained
men reached normal in a little over one hour, but in untrained men
remained rather high for considerably over an hour. The above results
are collected in the table.

IV. Faticuinc EXERCISE.

The eight young men experimented upon in this test ran from five
to nine miles and were all training for a Marathon race of twenty miles.

Systolic pressure. — The rise in systolic pressure noted in the man
who ran five miles was 25 mm. Hg; the three who ran six miles showed
a rise varying from 30 to 65 mm. Hg, with an average of 43. In the
seven-mile run a rise of 25 mm. Hg was noted, and the three men who
ran eight miles varied from 15 to 35 mm. Hg above normal, averag-
ing 27 mm. Hg. The average increase for all eight runners was 32.5
mm. Hg.

Diastolic pressure. — Diastolic pressure showed the following rises:

5 miles 15 mm. Hg
i ike)

6 miles i: 35 average, 30 mm. Hg
3: 45

7 miles 25 mm. Hg
ts 20

8 miles {: 20 average, 15 mm. Hg
3: 5

Average rise for 8 runners, 20.6 mm. Hg

458 Oswald S. Lowsley.

No pulse pressure rise was noted in two of the cases, the others
varied from an increase of 5 (six-mile run) to 20 (six-mile run). The
average increase for the runners was 9 mm. Hg. ‘The increase in pulse
rate varied from 20 to 62 per minute, with an average increase of 44.6.

The subnormal systolic pressure varied from 5 to 25 mm. Hg below
normal, with an average of about 20 mm. Hg. The time at which this
stage was most pronounced occurred between thirty-five minutes and
one hour and ten minutes after running, although systolic pressure
was always subnormal within ten minutes after the end of a long run.
Diastolic pressure dropped much more slowly than systolic and did
not show so marked a subnormal phase. Pulse pressure showed
an average drop of 19 mm. Hg below normal, and this might last
from fifteen minutes to one hour and ten minutes after a run of
this distance. Pulse rate did not become subnormal.

Return to normal. — Systolic pressure might return to normal in
about an hour, but in case the sprint at the end of the race was hard
one it took considerably longer. Diastolic pressure seemed to reach
a normal condition in about thirty-five minutes. Pulse pressure was
observed to follow in some cases the fluctuations of systolic pressure.

The above results are collected in the table.

V. EXHAUSTIVE EXERCISE.

This test was made on the same group of men (members of the Cross
Country Club of Baltimore) who furnished records for the previous
tests on runners. Records were made after a ten-mile race and two
twenty-mile races and after several practice runs varying from ten to
thirteen miles.

In the ten-mile handicap race the average rise in systolic pressure
was 40.8 mm. Hg in six runners. The greatest rise, 60 mm. Hg, was
noted in the case of the winner, and the least rise occurred in the case
of the scratch man, who made the best time; it amounted to 25 mm.
Hg. However, it is noteworthy that this man rarely had a rise of
more than 25 mm. Hg in any run, the single exception noted being a
rise of 30 mm. Hg after a hundred-yard dash, which is below the av-
erage for the cases examined, 7.e., 45 mm. Hg. The diastolic rise
averaged 22.5 mm. Hg, varying from 15 to 30. Pulse pressure rose
on an average about 18.3 mm. Hg, the extremes varying from 5 to 40.

The Effects of Various Forms of Exercise. A459

Pulse rate increased an average of 66.3 per minute, the greatest in-
crease being 80, a figure more than double that usually observed. In
fact, nearly all of the pulse rates were about doubled at the end of the
race.

The subnormal phase. — Systolic pressure was observed to reach its
lowest point in from forty-two minutes to four hours and seven min-
utes. It is to be noted that even after dinner, which always sends
systolic pressure up, as shown by Erlanger and Hooker, the pressure
returned to 20 mm. Hg below normal after the immediate effects of
the dinner had worn off. The average fall of systolic pressure in eight
tests was 14.5 mm. Hg, varying from o to 25 mm. Hg. Diastolic
pressure averaged a fall of g mm. Hg below normal, and came on
much more slowly than in the case of the systolic pressure. Pulse
pressure had a corresponding fall, averaging 13.3 mm. Hg, which was
in most cases coincident with the systolic fall. Pulse rate did not fall
below normal.

Return to normal. — Systolic pressure in the two cases least affected
by other influences than the race had not returned to normal in one
case until after dinner (four hours and thirty-five minutes), and in the
other at one hour and thirty minutes after the race, but in this case
it dropped again, and reached normal the second time in five hours
and five minutes. Diastolic pressure reached normal in these experi-
ments rather rapidly. Pulse pressures followed the fluctuations of
the systolic pressure very closely. Pulse rate did not show a tendency
to return to normal until a long time after the completion of the exer-
cise (four hours in one case, six hours in another, and after the sixth
hour in a third.

Thirteen-mile practice run. — This run, which was a hard one, was
similar in its effects to the ten-mile race already discussed.

The rise in systolic pressure in four cases averaged 36.25 mm. Hg.
Diastolic pressure showed an average rise of 15 mm. Hg. Pulse
pressure rose an average of 20 mm. Hg. Pulse rate decreased in one
case and in the other three showed an average increase of 29 per
minute.

The subnormal phase was quite pronounced. Systolic pressure fell
on the average 22 mm. Hg below normal in five cases. Diastolic
pressure fell on the average to 17 mm. Hg below normal. Pulse pres-
sure did not fall below normal in one case, but in the other four gave

460 Oswald S. Lowsley.

an average fall below normal of 15 mm. Hg. Pulse rate showed a
secondary rise in four of the curves after about twenty minutes.
Return to normal.— In the case whose record is given in Fig. 4,
the systolic pressure showed a tendency to remain below normal for
three and one-half hours until caused to rise by the effect of smoking,
aus ane bearing out the experiments of
(Senate Bruce, Miller, and Hooker,> who
40 found that smoking caused vaso-
2

constriction and rise of blood

SYSTOLIC PRESSURE, pressure.

The diastolic pressure in this
ent case had a general tendency

Ze0 0 0 @ Maw za ~=—- downward for three hours and ten
FIGURE 4.— Shows the effect of exhaus- minutes, and was brought up to
tive exercise (13-mile run) upon the 10 mm.Hg below normal by smok-
blood pressures and heart rate. WD, ing. Pulse pressure returned to
no dinner. S, smoking. Ordinates = normal or nearly to normal in

mm. Hg pressure and rate per minute. ‘ £

IN = eet three hours and thirty-five min-

utes. Pulse rate returned to nor-
mal more slowly than in the other experiments recorded.

Twenty-mile races. — The contestants in two races of this length
were examined. It was impossible to make observations immediately
after the first race, hence only the subnormal phase was studied. When
the runners were brought to the examination room, they were prac-
tically in a fainting condition and one or two of them were delirious.
The pulse pressures in most cases were so low that the instrument
would not record the fluctuations and it was impossible to get the
diastolic pressure.

Records taken from four to twenty minutes (average twelve min-
utes) after the completion of the race on twelve runners showed a
variation in systolic pressure of from 10 mm. Hg to 30 mm. Hg below
normal, averaging 24 mm. Hg for the twelve men. In 50 per cent of
the cases it was impossible to record the diastolic pressure on account
of the greatly diminished pulse pressure. The six diastolic pressures
obtained at this time varied from 1o mm. Hg subnormal to 10 mm. Hg
above normal. Two of the cases showed a normal pressure, and two

5 Bruce, Mitrer, and Hooxer: This journal, xxiv, p. 104.

The Effects of Various Forms of Exercise. 401

5 mm. Hg above normal. Pulse pressure in the six cases recorded
varied from 15 to 35 mm. Hg below normal, averaging 27.5 mm. Hg.
The average pulse pressure was 14 mm. Hg.

One and one-half hours after the completion of the race the average
systolic pressure was 20.5 mm. Hg below normal. Diastolic pressure
averaged 8.4 mm. Hg below normal, one case only being 5 mm. Hg

PULSE RATE

20 4 60 8 00 120 40 0 80 440 460 48 500

BEFORE
AFTER

FicurE 5.— Shows the effect of completely exhaustive exercise (20-mile run) on the
blood pressure and heart rate. Note the prolonged subnormal stage exhibited
in all the curves. D, broken lines. G, unbroken lines. LD, lying down. B,
bath. S, sitting. DR, dinner and recreation. Ordinates = mm. Hg pressure
and rate per minute. Abscissa = time in minutes.

above. Pulse pressure varied from 5 above normal in one case to 30
below, and averaged 13 mm. Hg below normal. The average pulse
pressure at this time was 20.5 mm. Hg.

Four of these runners were examined eight hours after the comple-
tion of the race. Two of them recorded normal systolic pressures, and
the other two recorded subnormal systolic pressures of 25 and 30 mm.
Hg respectively. One of the runners who had a normal systolic pres-
sure recorded a diastolic pressure 1o mm. Hg above normal. Pulse
pressure was subnormal in three of the cases, 10, 25, and 25 mm. Hg
respectively, while the fourth case was 5 mm. Hg above normal.

462 Oswald S. Lowsley.

On account of the fact that it was impossible to learn exactly what
these four men had been doing in the time between the race and the
last examination, it was not deemed proper to average this record
with the others, the figures of which are collected in the table.

Conditions were proper for taking observations at the finish of the
second twenty-mile race, and the contestants ran to the examination
chair, so that less than thirty seconds elapsed between the cessation
of muscular activity and the determination of systolic pressure.

All six of the men examined had systolic pressures varying from 15
to 25 mm. Hg above normal, and averaging 17 mm. Hg. Diastolic
pressure showed rises of to mm. Hg above normal in all cases. Pulse
pressure showed an average rise of 8.33 mm. Hg. Pulse rate showed
an average increase of 29.5 in the six cases.

Two hours after the completion of the race systolic pressure showed
an average fall of 24 mm. Hg below normal. Diastolic pressure had
become practically normal. Pulse pressure averaged 26 mm. Hg below
normal. Pulse rate averaged 18 per minute above normal except one
case which was 18 per minute below normal.

Tests made between seven and eight hours after the race showed
two of the contestants_to be still in the subnormal state (see D,
Fig. 5), while a third individual had returned to normal.

Relation of exercise to albuminuria. Albuminuria was first observed
to follow athletic exercise by Dunhill and Patterson in 1902,° and
has since been commonly noted. The present research gave an
opportunity to test further the hypothesis advanced from this
laboratory 7 that albuminuria in otherwise healthy individuals is
dependent upon a relative decrease in the magnitude of the pulse
pressure. Accordingly the urine was examined in seven of the cases
here studied. When the period of subnormal pressure came on accom-
panied by low pulse pressure, protein was found in the urine even
after the short runs, and the more extensive the period of low pulse
pressure the greater seemed to be the amount of protein present.

For convenience in comparing the results obtained for the different
forms of exercise, the following general table has been prepared sum-
marizing the figures previously given:

6 DuNHILL and Patterson: Intercolonial medical journal of Australasia, 1902,
Vil, 334.

7 ERLANGER and Hooker: Johns Hopkins Hospital reports, 1904, xii, pp.
145-378.

463

The Effects of Various Forms of Exercise.

*9SINIOXO OYE ][CT 1

eee wae oar eee ‘even eee eee eee 4 oe
(2) eageyen <0) || (2c) ey al eye (ey Gaz) |) Ce)S) Te) 5). ey ey ep Tt ® Syne ey aeltd
£T~ OL $9 09 ie = + + Lp eal Oe TS OPP tr cee 97
ae (€) |
(By leet es Eo] 1 ey ey). eS) ny) 4) 3) | or) |) 4) 1 -@n. |G) |. @® Se aa oe eae
cSt OL 88 Os Poe POL — | Obese is a A Sebi) sip 6 Ly o'OT 6
1) aia ee ey s)he) | en) @ 4 xy] x) | 1) 1) @z)-| © | CD Spy “urw
OT |) “Se 4 Fe g9 Iz |€O0I—| £— |76—| 6— |S'9—| GF | 907] TZ | FOZ | TT ur ainssord pooyq IpoIseIC
| 1) ©) | Ce) (1) We) || $e) 1) |) a1 -(6z)_| (oz) |, GY ) Ez)) ED. | z) yy "wrt
tho. sROne| Per): 28 TE. | GI | OC) SP) OT ets eer) See | ee | oe I ur aanssard Poojd I1]0}S4S
El ts Ss ts Se ry is
oe 5 0q° HO mH © non a oe gg" HO Bg m O. = 4 0Q ne) a mH
oho) Geel Ge ao. | Go| Gh | Go| Ga eee) Gee Be |) Be |) a) eo Be
| ee Be) Pe Re PS | ee he | ee Re ee OE) ee ee
RR | elas ® Cul 8 a o 8 o *yuNTHOINSvA TAT
*‘[euLtou 0} UINJOI = *o8v1S “OSTOIOXO I9}Je
Jo} pormbos soynurypy [euliouqns juonbosqns AJO}VIPOWIUUL ISTYT

‘ASIONTXY TO ANY AHL Wows
GALVWILSA SI ,,TVWYON OL NYALAY,, “INAWIAAdX*, HOVA NI GYAWASHO SIVAGIAIGNY JO ATIAWAN AHL OL YAIAN SASAHINAYVY NI
AdAJ, OVIVIT NI SAUNNOIY AH, ‘“ASIONTX A ADAILSAVHXY GNV ‘ONINOILVY ‘SNOMODIA ‘dldvy ‘aLVNAGO|. YAAIAV NOILVANASAO
UAINA SASVD AHL TIV YOU ALYY ASTAG ANV SaVASSaAXT ASTAG ANV ‘OITOISVICT ‘OITOLSAG NI SHONVHD AOVATAWY AHL ONIMOHS ATAV I,

464 Oswald S. Lowsley.

SUMMARY OF RESULTS.

A. Rise above normal after exercise. — 1. Systolic pressure rises more
in rapid and exhaustive than in other forms of exercise. The ex-
tent of this seems to depend partly on individual characteristics
and partly on the amount of energy put into the sprint at the end of
exercise. ;

2. Diastolic pressure rises about the same for all types of exercise
except the moderate, in which it shows about one-half the rise noted
in other types of exercise.

3. The pulse pressure rise is greatest after rapid and exhaustive
exercises. .

4. Pulse rate increases in all exercises, and the more vigorous the
exercise the greater the increase.

B. Fall below normal due to the effects of exercise. — rf. Systolic pres-
sure falls about equally below normal for all kinds of exercise, although
in the more exhaustive types there is a slightly greater fall.

2. The diastolic pressure shows a similar fall below normal follow-
ing all types of exercise.

3. Pulse pressure falls below normal in all cases almost equally.

4. Pulse rate fell slightly below normal in only four cases out of
sixty.

C. Return to normal.— 1. Systolic pressure returns to normal after
the subnormal phase more slowly the more exhaustive the exercise.

2. Diastolic pressure returns to normal later in the cases of rapid
and exhaustive exercise than in the others.

3. Pulse pressure returns to normal more slowly the more exhaustive
the exercise.

4. The pulse rate returns to normal more slowly the more exhaustive
the nature of the exercise.

CONCLUSIONS.

The conclusions to be drawn from these experiments are as follows:
1. All types of prolonged exercise which cause an increase in pulse
rate cause also a rise in systolic and diastolic pressures. The systolic
pressure shows the greater rise; hence there is an increase in pulse

The Effects of Various Forms of Exercise. 405

pressure, which may be interpreted to mean that the heart beats are
augmented as well as accelerated.

2. After all types of exercise here studied the systolic, diastolic,
and pulse pressures invariably fall below normal and remain in this
subnormal condition for a considerable time. The more exhaustive
the nature of the exercise the longer will be the subnormal period which
follows. Systolic pressure invariably falls more rapidly than diastolic,
and hence the pulse pressure becomes weaker. ‘The presence of al-
bumin in the urine coincident with low pulse pressure was observed in
accordance with the results reported by Erlanger and Hooker.’ It
seems to add another factor to the possible injurious results of long-
continued exhaustive exercise.

3. Pulse rate, which always increases during exercise, decreases
rapidly after its completion. This drop in the curve of the pulse rate
is frequently followed by a secondary rise which is possibly a reflex
effect due to the low blood pressure of the subnormal stage. In no
case was it observed that this secondary rise was accompanied by a
rise in blood pressure.

4. Rapid exercises (vigorous, fatiguing, and exhausting) are fol-
lowed by a fall of pressure below normal which lasts longer than after
moderate exercise, even if the former is continued for a very short
period and the latter for quite a long period of time. If we consider
the subnormal phase as indicative of an overstrain following upon
the great reflex excitation of the heart and vaso-motor centre, then it
would seem that after these so-called rapid exercises the strain is
more serious, as is shown by the much longer time required before
the conditions return to their normal level.

5. If our interpretation of the subnormal phase is correct, it would
follow that the so-called field events, consisting of jumping; shot
putting, discus and hammer throwing and baseball, gymnasium ap-
paratus work, and exercises of a similar nature, are preferable to
rapid exercises such as basket ball, football, and running races. This
is particularly true in the case of the rapidly growing youth, whose
heart is under the additional demand of keeping pace with an increase
in the tissue mass of the body.

6. There is less strain put upon the circulatory system by walking
a number of miles at a moderate rate than by sprinting 100 yards at

8 ERLANGER and HooKER: Loc. cit.

406 Oswald S. Lowsley.

top speed. This conclusion follows from the fact that blood pressure
returns to normal after moderate exercise in about thirty minutes, while
after short sprints the subnormal stage continues about three times
as long.

7. Long-distance running races and similar forms of exhaustive
exercise give rise to a serious strain on the heart, as is indicated by the
long period of subnormal blood pressure.

8. It would seem probable that in individual cases the beneficial
or injurious effect of any given form or amount of exercise might be
determined by observations upon the subnormal phase following the
exercise. When the subnormal phase returns to normal within sixty
minutes, the exercise may be considered as lying well within hygienic
limits for that individual, while a return that is delayed beyond one
hundred and twenty minutes may be regarded as exceeding these
limits.

This investigation was suggested by Dr. T. A. Storey and directed
by Dr. W. H. Howell, to whom the author is very greatly indebted.
Great appreciation is expressed to R. A. Kocher for his valuable assist-
ance in the experiments concerned with Part I, and to Dr. P. M. Daw-
son, who directed that part of the work. The members of the Mary-
land Swimming Club, the Baltimore Athletic Club, and the Baltimore
Cross Country Club were the subjects of these experiments, and par-
ticular gratitude is felt towards the members of the last-mentioned
organization, who were most generous in submitting to all the demands
made upon them by the author.

ON THE QUESTION WHETHER DEXTROSE ARISES
FROM CELLULOSE IN DIGESTION.

By GRAHAM LUSK.

[From the Physiological Laboratory of the University and Bellevue Hos pital
Medical College.|

A PUBLICATION by Hoffmann,! which shows that although the

hemicellulose of agar agar and cellulose of cabbage is digested
and absorbed by the rabbit, it does not increase the sugar in the urine
if the rabbit be phlorhizinized, recalled to the writer two experiments
briefly reported by him ? at a meeting of the American Physiological
Society in 1gor1 and never fully published.

The great sugar elimination which followed the ingestion of fat
and cauliflower in the phlorhizinized dogs of Hartogh and Schumm,?
suggested that the sugar might in part have been derived from the cau-
liflower ingested. Twenty grams of cauliflower were therefore cooked
and given to a fasting dog phlorhizinized according to the author’s
method.* The results of the urinary analyses were as follows:

1901. Condition. Dextrose. Nitrogen, D.N.
Nov. 8... Fasting 33.34 9.69 3.34
Nov. 9. . .20g. cauliflower 29.18 8.82 3.30
Mov.t0..°-. . Fasting 28.20 8.30 3.40

It is evident from this, that the cauliflower did not increase the
sugar output as it would have done had dextrose arisen from it in
the intestine.

Another experiment was performed upon a phlorhizinized goat
weighing 34.2 kgm. ‘The cellulose ingested consisted of 4 gm. of

1 HoFFMANN: Inaugural-Dissertation, Halle-Wittenberg, r1o1o0.
2 Lusk: This journal, 1902, vi, p. xiii.
3 HARTOGH and Scuumm: Archiv fiir experimentelle Pathologie, 1900, xlv, p. 2.
4 Stizes and Lusk: This journal, 1903, x, p. 67, and other papers.
467

468 Graham Lusk.

filter paper soaked in sodium chloride, given at the beginning of the
twenty-four-hour period; six hours later 6 gm. of thick brown wrap-
ping paper were given. This was all taken voluntarily. Two grams
of phlorhizin were administered to the goat subcutaneously three times
daily, beginning the day before the experiment and continuing through-
out. The urinary analyses showed the following results:

1901. Condition. Dextrose. Nitrogen. Dz N.
Deciuy. . .. Fasting 16.56 7.27 2.28
Dec: 15°. +.‘ To'g. paper 11.62 5-04 2.30

Fragments of cellulose fibres were found in the stools indicating their
distribution throughout the intestinal tract.

Tappeiner ® was the first to point out that fatty acids arose from
cellulose in the intestinal canal of herbivora, and it is probable that
the nutritive value of cellulose has its origin in these fatty acids.®

These isolated experiments are now published because they sup-
port the careful work of Hoffmann in showing that sugar does not
arise from the digestion of cellulose.

5 TAPPEINER: Zeitschrift fiir Biologie, 1888, xxiv, p. 105.
6 Consult the criticisms of Lonriscu: Zentralblatt fiir die gesamte Physiologie
und Pathologie des Stoffwechsels, 1907, N. F. ii, p. 301.

vise Deal ys aR CAR eRe a iS A II,

Pax. TOeVOLEy SAV I.

peo, relation to shock, 152.
Adrenalin produces glycosuria after

thyroidectomy, 331!

Alcohol affects nitrogenous metabolism, 1.

Allantoin excretions in monkey, xv.

Anesthesia, local, relation to
changes in skin, 45.

Anaphylaxis, xxiv.

Antagonism of salts, xxxii.

Asphyxia, rise of blood pressure, xxii.

AvER, J. Acute anaphylactic death in
rabbits, xxiv.

sensory

ARBOUR, G. F., and P. G. STmILEs.
On localized contraction in skeletal
muscle, xi.

BaArRRIncER, B. S.
BARRINGER, IIQ.

BARRINGER, T. B., Jr., and B. S. Bar-
RINGER. A comparison of the total
nitrogen excretion of either kidney in
normal individuals during varying periods
of time, 1109.

Baths, cold, influence on glycogen content,
427.

Becat, F.C. See LuckHARDT and BEcHT,
Xvi.

Becat, F. C., and A. B. LuckHarpt. The
source of the immune bodies in the
lymphs, xi. :

BENEDICT, F. G., L. E. Emmes, and J. A.
Ricue. The influence of the preceding
diet on the respiratory quotient after
active digestion has ceased, 383.

Blood pressure, asphyxial rise in spinal
animal, xxii.

——,, pulmonary circuit, xxi.

, relation to secretion in kidney, 24.

Brooks, C. The effect of lesions of the
dorsal nerve roots on the reflex excitabil-
ity of the spinal cord, 212.

See BARRINGER and

469

ANNON, W. B. Some observations on
the nature of gastric peristalsis, xii.

Cannon, W. B., and C. W. Lies. The re-
ceptive relaxation of the stomach, xiii.

Cartson, A. J. The effects of stretching
the nerve on the rate of conduction of
the nervous impulse, 323.

Cartson, A. J., J. F. Rooxs, and J. F.
McKie. Attempts to produce experi-
mental hyperthyroidism, xiii.

Cell division, dynamics of, 240.

, relation to calcium salts, 280.

Cerebrospinal fluid, modified by infundib-
ular secretion, 60.

Circulation, in acapnia and shock, 152.

——, influenced by exercise, 446.

, in kidney, 24.

——., measurement of blood flow, xx.

——,, migration of liquids in the absence of
the heart, xxix.

Conductivity, of nerve, affected by stretch-
ing, 323.

, of nerve and muscle, affected by pres-
sure, 308.

Corpus luteum, xxii.

CusHING, H., and E. Gortscn. Concern-
ing the secretion of the infundibular lobe
of the pituitary body and its presence in
the cerebrospinal fluid, 60.

pe origin from cellulose, 467.
Diet, relation to respiratory quotient,
8

393:
Digestion, of cellulose, 467.
, relation to respiratory quotient, 383.
Duodenum, inhibition of, during contrac-
tion of stomach, xxxi.

|B ae L. E., and J. A. Ricue. The
respiratory exchange as affected by
body position, 406.

470 Index.

Emumes, L. E. See BENEpictT, Emmes, and
RICHE, 383.

Equilibrium, nerve control of, 207.

ERLANGER, J. Observations on auricular
strips of the cat’s heart, 87.

Exercise, influence on circulation, 446.

Eye, constants of pupillary reaction, xxviii.

——, pupillometer experiments, xxviii.

——, shadow pupillometer, xiv.

Eyster, J. A. E., and H. E. Jorpan. Ef-
fect of intravenous injection of extracts
of the pineal body, xxiii.

ARADIC stimulation, 226.
Firefly, production of light by, 122.

Fishes, olfactory sense of, xix.

Firz, G. W. A shadow pupillometer for
the accurate study of pupillary reactions,
xiv.

Firz, G. W. The constants of pupillary
reaction. (A preliminary report of ex-
perimentation with the shadow pupil-
lometer), xxviii.

Frawz, S. I., and W. C. RuepicEer. Sen-
sory changes in the skin following the
application of local anesthetics and other
agents. — I. Ethyl chloride, 45.

(Ge. M. H. See Hunter and
GIVENS, xv.

Glycogen, distribution over the liver, 341.

, removal from human subject, xxii.

Glycogenolysis, 341.

Glycosuria, by adrenalin, after thyroidec-
tomy, 331.

GoetscH, E. See Cusninc and GOETSCH,
60.

GREER, J. R. See Muriin and GREER,
xviii.

ATTREM, W. M., and P. B. Hawk.

On intestinal putrefaction during

copious and moderate water drinking
with meals, xxv.

Hawk, P. B. The activity of the pan-
creatic function under the influence of
copious water drinking with meals, xxvi.

Hawk, P. B. See HAtrrem and HAwkE,
XXV.

Hawk, P. B. See Writs and Hawk, xxxii.

Heart, auricular strips, rhythmical con-
tractions of, 87.

, relation to respiratory metabolism,

XViil.

HENDERSON, Y. Acapnia and shock. —
VII. Failure of the circulation, 152.
Hitpircu, W. W. See MENDEL and Hit-
DITCH, I.

Hooxer, D. R. The influence of pulse
pressure upon renal function, 24.

Hunter, A., and M. H. Givens. The
allantoin-purine excretion of the monkey,
XV.

Hypophysis, action of extracts of, xvii.

MMUNITY relation to lymph, xi.

, role of spleen, xvi.

Innervation, contrary, law of, xxxi.

Intestine, putrefaction in, affected by drink-
ing water with meals, xxv.

, resection of, influence on metabolism,

3006.

ACKSON, D. E. An automatic shel-
lacking device, xxx.

Jorpan, H. E. See Eyster and JORDAN,
Xxiil.

Josrepu, D. R., and S. J. Mretrzer. Inhi-
bition of the duodenum coincident with
the movements of the pyloric part of the
stomach, xxxi.

lenene J. K., and A. McDERmotTT.
Some observations on the produc-
tion of light by the firefly, 122.

Kidney, isolated, pulse pressure related to
secretion, 24.

, secretion of right compared with that

of left, 119. ‘
EAPER, W. E. See MEEK and LEAPER,
3c8.

LEVENE, P. A., and F. MEDIGRECEANU. On
nuclein metabolism in the dog, 438.

Lewis, D. D. See Mrtter, Lewis, and
MATTHEWS, xvii.

Lies, C. W. See CANNON and LIEB, xiii.

Liu, R. S. The physiology of cell divi-
sion. —III. The action of calcium salts
in preventing the initiation of cell divi-
sion in unfertilized eggs through isotonic
solutions of sodium salts, 289.

Light of firefly, 122.

Loes, J. Further experiments on the
antagonistic action of salts, xxxii.

Lores, L. The function of the corpus
lutem, xxii.

Lowstey, O. S. The effects of various
forms of exercise on systolic, diastolic,
and pulse pressures and pulse rate, 446.

OS ee ee

Index.

LuckxHarpt, A. B. See Becut and Luck-
HARDT, Xi.

Lucxuarnt, A. B., and F. C. Becut. The
part played by the spleen in the forma-
tion of immune bodies, xvi.

Lusk, G. A method of removing glycogen
from the human subject, xxii.

Lusk, G. On the question whether dex-
trose arises from cellulose in digestion,
467.

Lusk, G. The influence of cold baths on
the glycogen content of man, 427.

Lymph, immune bodies in, xi.

NM ACLEOD, J. J. R., and R. G. PEARCE.
Studies in experimental glycosuria.
—VI. The distribution of glycogen
over the liver under various conditions.
Post mortem glycogenolysis, 341.

Martin, E. G. A quantitative study of
faradic stimulation. — V. The influence
of tissue resistance and of kathode sur-
face on stimulating effectiveness, 226.

MattTHews, S.A. See Mirter, Lewis, and
MATTHEWS, xvii.

McCtenpon, J. F. On the dynamics of cell
division. —II. Changes in permeability
of developing eggs to electrolytes, 240.

McDermott, A. See KastLe and McDErR-
MOTT, 122.

McK, J. F. See Cartson, Rooks, and
McKr, xiii.

MEDIGRECEANU, F.
MEDIGRECEANU, 438.

MEEK, W. J., and W. E. LEAPER. Effects
of pressure on conductivity in nerve and
muscle, 308.

Metcs, E. B. The osmotic properties of
smooth muscle, xvii.

Me tzer, S. J. The migration of solu-

ions in animal bodies deprived of their
cardiac circulation, xxix.

MeEtrzer, S. J. See JosrepH and MELTZER,
XXXxi.

MENDEL, L. B. See OSBORNE and MENDEL,
XXVi.

MENnpEIL, L. B., and W. W. Hitpircu. The
influence of alcohol upon nitrogenous
metabolism in men and animals, r.

Menstruation, metabolism of nitrogen in,
77

Metabolism, glycogen, xxiii.

, nitrogen balance during pregnancy
and menstruation, 177.

——,, nuclein, 438.

See LEVENE and

471

Metabolism of glycogen influenced by cold
baths, 427.

, proteid, affected by alcohol, r.

——.,, purin, xv.

——.,, relation to sitting and lying, 406.

, respiratory, relation to heart action,

XViii.

, with resected intestine, 366.

Murr, J. L., D. D. Lewis, and S. A.
Mattruews. The effects of extracts of
the different parts of the hypophysis,
XVii.

Moore, A. R. On the nervous mechanism
of the righting movements of the starfish,
207.

Morcutts, S. Contributions to the physi-
ology of regeneration. — V. Regenera-
tion of isolated segments and of small
pieces of worms, 415.

Morin, J. R., and J. R. Greer. The
heart action in relation to the respira-
tory metabolism, xviii.

Morir, J. R. Metabolism of develop-
ment.—II. Nitrogen balance during
pregnancy and menstruation of the dog,
E77

Muscle, conductivity, affected by pressure,
308.

, localized contraction, xi.

, smooth, osmotic properties, xvii.

ERVE, conductivity, affected by pres-
sure, 308.

, mechanism of equilibrium in starfish,
207.

Nuclein, metabolism in dog, 438.

LFACTORY sense of fishes, xix.
OsBoRNE, S. B., and L. B. MENDEL.
Feeding experiments with mixtures of
isolated food substances, xxvi.

ANCREAS, activity affected by drink-
ing water with meals, xxvi.

ParKER, G. H. The olfactory sense of
fishes, xix.

PEARCE, R. G. See MACLEOD and PEARCE,
341.

Peristalsis, gastric, xii.

Permeability, affected by calcium salts, 289.

, relation to electrolytes, 240.

Pree, F. H. The mechanism of the as-
phyxial rise of blood pressure in the
spinal animal, xxii.

Pineal body, effect of extracts, xxiii.

472

Pituitary body, secretion of infundibular
lobe, 60.

Porter, W. T. The relation of afferent
impulses to the vasomotor centres, 276.

Pregnancy, metabolism of nitrogen in, 177.

Proteins, individual, significance in nu-
trition, xxvi.

Pulse pressure, in pulmonary circuit, xxi.

Pupillometer, xiv.

EGENERATION of segments, 415.
Respiratory exchange, affected by

body position, 406.

Respiratory quotient, related to diet, 383.

RicuHe, J. A. See BENEDICT, Emmes, and
RIcHE, 383.

RicHE, J. A. See Emmes and RIcHE, 406.

Rooks, J. F. See Cartson, Rooks, and
McK, xiii.

RUEDIGER, W. C. See Franz and RUueE-
DIGER, 45.

ee eb antagonistic action, xxxil.
ScHREINER, O., and M. X. SULLIVAN.
Biological analogies in soil oxidation, xxv.

Secretion, relation to blood pressure, 24.

Shellacking device, xxx.

Shock, in acapnia, 152.

Smrpson, S. Are the parathyroids capable
of replacing the thyroids functionally,
XXVil.

Skin, effects of local anesthesia, 45.

Soil oxidation, xxv.

Spinal cord, reflexes affected by lesions of
dorsal roots, 212.

——,, transection, effect of, on asphyxial
rise of blood pressure, xxii.

Spleen, its part in immunity, xvi.

Stewart, G. N. Measurement of the
blood flow in man, xx.

Index.

Stites, P.G. See BARBour and STILES, xi.

Stimulation, faradic, relation to tissue re-
sistance and kathode surface, 226.

Stomach, movements of, xxxi.

, receptive relaxation, xiii.

, secretion affected by drinking water

with meals, xxxii.

Stretching nerves affects conductivity, 323.

Sullivan, M. X. See Schreiner and Sulli-
van, XXV.

HYROID, removal, relation to adren-
alin glycosuria, 331.
Thyroidism, xiii:
Thyroids, replacement by parathyroids,
XXVii.

NDERHILL, F. P. The metabolism
of dogs with functionally restricted
small intestine, 366.
UNDERHILL, F. P. The production of
glycosuria by adrenalin in thyroidec-
tomized dogs, 331.

ASOMOTOR centre, relation to af-
ferent impulses, 276.
Vasomotor reflexes, 276.

on intestinal putrefaction, xxv.
, effect on pancreatic function,

XXVi.

, ——, stimulation of gastric secretion,
XXxil.

WicceErs, C. J. Pulse pressure variations
in the pulmonary circuit, xxi.

Wits, F., and P. B. Hawk. The stimula-
tion of the gastric secretion under the
influence of water drinking with meals,
XXxil.

\\/ ATER drinking, with meals, effect

665 ~-¢%

—27° \~ 2
hd

= ie \ A
a
,

i i
‘ Mire

“¢

ite
a > |
:

ui ae Pel

BINDING SECT. MAR 4 1966

QP American Journal of Physiology
iL

A5
Vier
COPee

Biological
& Medical

Serials

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY

Wt de be eae
\,

be viehar go a0
leit Se

4 Sekar te
ahs cere cen iy
eens ays

aosri> Sha erm Wy Baty
er ak pemn es
nal Deere)

aebeiy sy ah
Sadia ae aaa tea Ms
Bearer

SP tay othe
SUN errs
sy t

338, ab aS
s d ae
ran) Mae
ee

he
Be aie ay
SM sire bass ie

Pater t
. ae

vetah . $4

ab sit shy 95 gh

fe se sit
‘

Senge

art

raat

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "American journal of physiology"

See other formats