r e
ont
Pa a

4
a

Det tel

if

ait

4
a
ye GSE Pere 4s thas
pal 4 . :
‘ "

ao ote
Vo yhe ty
fist

reels,

or lMy
se terint
a
ri
oteqre dese
at RVR SE STC ETN
Pay yet eon

preset
‘

8
sna t i f

as

Yaar i

l be
VU

THE AMERICAN JOURNAL

OF

PHYSIOLOGY.

EDITED FOR

Che American Physiological Society

BY

H. P. BOWDITCH, M.D., BOSTON FREDERIC S. LEE, PH.D., NEW YORK
R. H. CHITTENDEN, Pu.D., NEW HAVEN JACQUES LOEB, M.D., BERKELEY
W. H. HOWELL, M.D., BALTIMORE W. P. LOMBARD, M.D., ANN ARBOR

W. T. PORTER, M.D., BOSTON

/

AMERICAN JOURNAL

Py SiG OCY

VOLUME. [xe

BOSTON, UsS.%
GINN AND COMPANY
1903

Copyright, 1903

ee By GINN AND COMPANY
rk
.
——- Gniversity Press
JoHN WiLson AND Son, CAMBRIDGE, U.S.A.
mn —,
_ * aN

CONE NES:

No. I, MaRcH 2, 1903.

PAGE

STUDIES ON THE INFLUENCE OF ARTIFICIAL RESPIRATION UPON STRYCH-
NINE SPASMS AND RESPIRATORY MOVEMENTS. By William J. Gtes
and S. J. Meltzer :

ON THE NEGATIVE AND POSITIVE PHOTOTROPISM OF THE EARTHWORM
ALLOLOBOPHORA FcTIDA (SAV.) AS DETERMINED BY LIGHT OF
DIFFERENT INTENSITIES. By George P. Adams

AN EXPERIMENTAL STUDY OF THE SUGAR CONTENT AND EXTRAVAS-
CULAR COAGULATION OF THE BLOOD AFTER ADMINISTRATION OF
ADRENALIN. By Charles H. Vosburgh and A. N. Richards

THE IMMEDIATE INFLUENCE OF EXERCISE UPON THE IRRITABILITY OF
HuMAN VOLUNTARY MuSCLE. By Thomas Andrew Storey

No. IJ, APRIL 1, 1903.

A STUDY OF THE VASOMOTOR NERVES OF THE RABBIT’S EAR CON-
TAINED IN THE THIRD CERVICAL AND IN THE CERVICAL SYMPA-
THETIC NERVES. By S /. Meltzer and Clara Meltzer .

THE SPECIFIC ROTATION OF THE NUCLEIC ACID OF THE WHEAT
Embryo. By Thomas B. Osborne .

THE INFLUENCE OF COLD ON THE ACTION OF SOME HMOLYTIC
AGENTS. By G. WV. Stewart .

QuIckK METHODS FOR CRYSTALLIZING OXYHMOGLOBIN: INHIBITORY
AND ACCELERATOR PHENOMENA, ETC.: CHANGES IN THE FORM OF
CRYSESMMIZATION: By Ldward Todeichert 2 . . . + .

ARTIFICIAL PARTHENOGENESIS IN NEREIS. By Martin H. Fischer

™N

Lal

I10O

vl Contents.

PAGE
No. III, May 1, 3903.
THE IMMUNITY OF FUNDULUS EGGS AND EMBRYOS TO ELECTRICAL
STIMULATION. . By Orville H. Brown <1. . 2 Ge

ON THE INFLUENCE OF VARYING INTENSITIES AND QUALITIES OF VISUAL
STIMULATION UPON THE RAPIDITY’ OF REACTIONS TO AUDITORY
StTimuLit. By Robert MacDougall’... 0) ws ere

ON THE RELATION OF EYE MOVEMENTS TO LIMITING VISUAL STIMULI.

By Robert MacDougall . : 2 ek etme

ON THE IRRITABILITY OF THE BRAIN DURING ANEMIA. By William
Fs GOS i eg

ON THE FORMATION OF GLYCOGEN FROM GLYCOPROTEIDS AND OTHER
PROTEIDS. By Lyman Brumbaugh Stookey . . .-. += eae

THE SHARE OF THE CENTRAL VASOMOTOR INNERVATION IN THE VASO-
CONSTRICTION CAUSED BY INTRAVENOUS INJECTION OF SUPRARENAL
Extract: By S. J. Melizer and Clara Meltzer. .. . ~ Go

MuscuLark CONTRACTION AND THE VENOUS BLOoop-FLow. By R. Bur-
ton-Opite ww ws be de

No. LV, JUNE, 15 1903.

THE INFLUENCE OF FORMALDEHYDE ON THE ACTION OF CERTAIN
LAKING AGENTS AND ON COAGULATION OF BLooD. Sy Charles
Claude Guthrie |... an & Mig Sh ree 9)

Venous Pressures. By Russell Burton-Opitz «. . « « = ja) ee

A STUDY OF THE PHYSIOLOGICAL ACTION AND TOXICOLOGY OF C@SIUM
CHLORIDE. By G. A. Hanford eee ne ete et oy

SOME FACTS CONCERNING GEOTROPIC GATHERINGS OF PARAMECIA.

By Anne Moore... wis 6 aon 2 ee se
No: V; JULY a, roou

SOME OBSERVATIONS ON THE EFFECTS OF AGITATION UPON ARBACIA
Eccs. By S. /. Meltzer Reet weer) 2:15

ON THE EFFECTS OF SUBCUTANEOUS INJECTION OF THE EXTRACT OF
THE SUPRARENAL CAPSULE UPON THE BLOOD-VESSELS OF THE
Raspit’s Ear. By S. J. Meltzer and Clara Meltzer . . . . . « 252

DIFFERENCES OF POTENTIAL BETWEEN BLOOD AND SERUM AND BE-
TWEEN NORMAL AND LAKED BLoop. By GIN. Stewart eee

THE Acipiry oF Urine. | By Otto Folin -. 1. 2 2 . | ee

A STUDY OF THE REACTIONS AND REACTION TIME OF THE MEDUSA
GONIONEMA MurBACHIT TO PHOTIC STIMULI. By Robert M. Yerkes 279

Contents. Vil

PAGE
EXPERIMENTS IN ARTIFICIAL PARTHENOGENESIS. By EZ. P. Lyon. . . 308
THE RELATIONSHIP BETWEEN THE FREEZING POINT DEPRESSION AND
SPECIFIC GRAVITY OF URINE, UNDER VARYING CONDITIONS OF
METABOLISM, AND ITS CLINICAL VALUE IN THE ESTIMATION OF
SUGAR AND ALBUMIN. “By G. HW. A Clowes, PHD. 2) eae B19
PROCEEDINGS OF THE AMERICAN PHYSIOLOGICAL SOCIETY . ... .« iX-XViil
Nox Vly VAvGusr 1, 1903.
NEw EXPERIMENTS ON THE PHYSIOLOGICAL ACTION OF THE PROTEOSES.
MC CI AMO MMCR MAL 8s wh em, of ee eens i oe 2 a Coe a AS
SitmeintGok MWORTIS: 2yO1o Mola 2 ss. 2 Be ce se SF
ON THE FORMATION OF DEXTROSE IN METABOLISM FROM THE END-
PRODUCTS OF A PANCREATIC DiGEst OF MEAT. By Percy G. Stiles
HUE RONOMDIGWS PA” <i ss Ue fee et. a eee Gee aase os 380

ON THE QUESTION OF PROTEID SYNTHESIS IN THE ANIMAL Bopy. By
Wandell Pienderson and Arihur L. Dean. <9... as = = = 386

OBSERVATIONS ON THE URINE OF THE MUSKRAT (FIBER ZIBETHICUS).
By Robert Banks Gibson

SALIVARY DIGESTION IN THE STOMACH. By W. 2B. Cannon and H.F. Day 396
NUCLEIN METABOLISM IN LyMPHATIC LEUK&MIA. By Vandell Henderson

GHONGUSIO} LT. FEAWAarAS. i.) cs tea te et Me ae og) ee ee AL
THE EFFECT OF DIURETICS, NEPHRITIC POISONS, AND OTHER AGEN-

CIES ON THE CHLORIDES OF THE URINE. By Torald Sollmann . . 425
THE COMPARATIVE DIURETIC EFFECT OF SALINE SOLUTIONS. By Torald

SOM LAT re Oa a I Fey ME ek ie ce ae ts erie iy lls
Pam. IKESPONSE OF THE FROG To LignT. Ay Ellen Torelle..-. ..-: 466
INDEX

eet OO)

mere rOAN PHYSIOLOGICAL. SOCIETY.

SIXTH ANNUAL MEETING

IN CONNECTION WITH THE SIXTH CONGRESS OF AMERICAN
PHYSICIANS AND SURGEONS.

WASHINGTON, D.C. May 12, 1903.

PROCEEDINGS OF THE AMERICAN PHYSIOLOGICAL
SOCIETY,

THE EFFECT OF ALCOHOL AND ALCOHOLIC FLUIDS UPON
THE EXCRETION OF URIC ACID IN MAN.

By R. H. CHITTENDEN (For S, P. BEEBE),

Two kinds of experiments were tried: metabolism experiments, cov-
ering from one to three weeks; and hourly period experiments, in
which the urine was collected each hour during the day. The sub-
ject of the metabolism experiments was a healthy young man of 65
kilos weight, not accustomed to the use of alcohol in any form. A
fixed diet was eaten. Alcohol in four forms was used: absolute
alcohol suitably diluted, whiskey, beer, and port wine. In most cases
the quantity of the alcoholic fluid used in twenty-four hours contained
between 75 and 80 c.c. of absolute alcohol. The most decided effect
was obtained by the use of beer and port wine. There was but little
difference between the effect of pure alcohol and whiskey. Typical
results are as follows:

WHISKEY EXPERIMENT.

Nitrogen. Ammonia N. Uric acid.

grams. per cent.

gram. 5
Fore period? . . 15.61 0.741 ee 89.6
Alcohol period. . 15:73 0.873 ; 87.8

After period. . . 16.53 0.717 ; 90.8

1 Per cent of total nitrogen.
2 These periods were of 6 days duration in each case, and the figures given are
the daily averages for the periods.

xii Proceedings of the American Physiological Society.

WINE EXPERIMENT.

Nitrogen. Uric acid.

grams.

Fore period . . . 16.07
Alcohol period . . 16:8
After period. . . 16.64

Since the amount of alcohol in the different experiments is very
nearly the same, the larger effect of beer and wine must be due to
something other than the alcohol they contain.

In the hourly period experiments, the subjects fasted from 6 P.M.
until 12 M. or 1 P.M. the following day, at which time a test meal was
eaten; this served as a control day, while for the alcohol day the
same program was followed, with the exception of adding varying
quantities of alcohol to the test meal. In all but one individual the
results showed a marked increase in the uric acid excretion. This
increase began in most cases in the second hour after the meal, and
reached the highest point at the fifth hour. The total quantity of
uric acid excreted during the twenty-four hours of the alcohol day
was greater than that of the control day, showing that the increase
immediately following the alcohol meal was not due to a mere hasten-
ing of the normal output.

It may be noted that in one experiment where 50 c.c. absolute
alcohol, suitably diluted with water, were taken without any food
after the usual period of fasting, a very marked diuresis was pro-
duced, but a decrease in uric acid was noted. This indicates that the
effect produced is due to a disturbance in the metabolism of the purin
bases of the food.

THE EFFECT OF SALINE INJECTIONS, DIURETICS, AND
NEPHRITIC POISONS ON THE CHLORIDE-CONTENT
OF THE URINE IN THE DOG.

By TORALD SOLLMANN.

Tue factors studied can be arranged in four classes:

Class I; Those which diminish the per cent of chlorine (to about
0.020 per cent), but which cause considerable diuresis and conse-
quently an increase in the absolute quantity of chlorine: Intravenous

Proceedings of the American Physiological Society. xiii

injections of solutions of urea, glucose, alcohol, and sodium acetate,
ferrocyanide, phosphate and sulphate.

Class [7 ; diminishing both the per cent and the absolute amount
of chlorine: water, salt starvation.

Class IIT : No effect upon the chlorine: nephritic poisons, caffein,
phlorhizin, laking, degree of diuresis, and (within the limits of the
experiments) the quantity or concentration of the injected fluid.

Class IV: Increasing the per cent of chlorine, if this has been pre-
viously low: solutions of sodium nitrate, sulfocyanide, iodide, and
presumably bromide.

The study of these factors shows that the essential cause of the
diminished chloride-content of the urine is not an increased diuresis,
nor the presence of a foreign salt, nor the dilution of the serum; but
the only essential fact is a lowered per cent of chlorine in the serum.

Analyses of the serum show, however, that the chlorine of the urine
is not proportional to the total chlorine of the serum. To reconcile
these facts it is necessary to assume that two forms of sodium chloride
exist in the serum, and that the one form (perhaps in combination
with proteid?) is not secreted by the living kidneys, whilst the other
form (free NaCl) is readily excreted. This conclusion is confirmed
by the perfusion of excised kidneys with diluted blood and with
saline solutions; whereas, in the latter case, the urine contains the
same per cent chlorine as the perfusing solution, it only contains a
fraction of the chlorine of the serum if blood is perfused.

The effect of the nitrate, sulfocyanide, and iodide is most readily
explained by assuming that these anions displace the chlorine ion
from the unfiltrable compound.

THE CAUSE OF THE GREATER DIURETIC ACTION OF
HYPERISOTONIC SALT-SOLUTIONS.

By TORALD SOLLMANN.

Ir is found that the diuretic effect of equimolecular salt solutions
is (with a few exceptions) proportional to their osmotic pressure.
It is shown that this is not sufficiently explained by the greater
hydrzemia,

Experiments on excised kidneys show that hyperisotonic solutions

xiv Proceedings of the American Physiological Society.

increase, whilst hypoisotonic. solutions decrease or arrest, the blood
flow and urine filtration in excised and dead kidneys. The superior
diuretic effect of hyperisotonic solutions is therefore, at least in large
part, purely physical, and is explained by the dehydration and shrink-
age of the renal tissues, and the more rapid circulation which this
produced. The diuretic effect of saline injection is also partly ex-
plained by the dilution and lessened viscidity of the blood, and
the resulting quickened renal circulation.

ON THE PHARMACOLOGICAL ACTION OF OPTICAL
ISOMERS.

ARTHUR R. CUSHNY.

Ir is well known that living protoplasm is capable of differentiating
between optical isomers, such as the sugars, and tartaric acids, and
Fischer has recently shown that the ferments are also possessed of
this power. The mammalian tissues also oxidize one of the isomers
more rapidly than the other. My experiments were performed with
lazevo-hyoscyamine and the racemic form, atropine. It was found that
these acted equally strongly on the terminations of the motor nerves
in striated muscle in the frog, on the frog’s heart muscle and on the
central nervous system in mammals, while atropine had a more stimu-
lant action on the central nervous system of the frog. Hyoscyamine
acted almost exactly twice as strongly as atropine on the nerve ter-
minations in the salivary gland, heart, and pupil in mammals, from
which it was inferred that the racemic atropine liberates the two
optically active forms when dissolved, and that the dextro-hyoscya-
mine is practically devoid of action in these organs. This was con-
firmed by examination of some dextro-hyoscyamine obtained from
its discoverer, Dr. Gadamer. It was found that laevo-hyoscyamine
acts 12-16 times as strongly on the salivary secretion as dextro-
hyoscyamine, and about 12-14 times as strongly on the terminations
of the inhibitory cardiac fibres. The terminations of the nerves in
the salivary glands, heart, and pupil can therefore differentiate be-
tween these optical isomers.

Proceedings of the American Physiological Society.

ON SECRETIN AND LYMPH-FLOW.

By LAFAYETTE B. MENDEL (with HENRY C. THACHER).

XV

Mr. THACHER has sought to ascertain whether the metabolic activity
of the glands stimulated by secretin may give rise to an increased
flow of lymph from the thoracic duct comparable with that induced

11.08-11.38

11.47-12.17, injection of 50 c.c.

{He Ie)
11.57-12.17
12.19-12.29
ZION SD
12.39-12.59

Lymph-flow in Total solids
10 minutes. in lymph.

Ash in
lymph.

-c, per cent.

c
Ls 5.46

2.4
4.6
a5 |

Zl

21]

per cent.

0.87

of secretin solution.

1.03-1.33, second injection of 50 c.c. of secretin solution.

1.02-1.12
LVZ-1.22,
1.22-1.32
1.32-1.42
1.48-1.58
1.58-2.08
2.08-2.18

|

AP |

4.5 |
3.0

so]

After a third injection of secretin the animal was killed by

blowing air into its veins.

was noted, as follows:

3.00-3.20
3.20-3.45
3.45-4.10

2.9 |

1.6 ¢
0.1

A post-mortem lymph-flow

by other lymphagogues and independent of any changes in arterial

pressure.

Fifty cubic centimetres of a secretin solution (prepared as directed

1 Read by title.

xvi Proceedings of the American Physiological Society.

by Bayliss and Starling) were injected into the facial vein of a dog
of 16.5 kilos in thirty minutes. The arterial pressure remained un-
changed at 76-78 mm. of mercury during the entire time, with the
exception of a few seconds after the inadvertent rapid injection of
two cubic centimetres of the solution. Similar injections were sub-
sequently made with the same results. Extracts from the protocols
follow:

The injection thus gave rise to an increased flow of lymph some-
what richer in solids, —a feature characteristic of the action of cer-
tain lymphagogues. The absence of the typical blood-pressure
effects which accompany albumose injections would lead one to attrib-
ute the lymphagogic action to the secretin. Since this experiment
was performed, however, Falloise has published a series of observa-
tions carried out with similar purpose. He concludes that purified
and albumose-free secretin fails to accelerate lymph-flow, although
it may still induce the typical stimulation of the pancreatic and
hepatic cells. We have not followed this experimentally. An in-
creased flow of lymph as observed above entirely independent of
alterations in general blood-pressure is, however, of interest in
itself.

THE INFLUENCE OF CHINIC ACID ON THE ELIMINATION
OF URIC ACID.

By W. A. TALTAVALL anv WILLIAM J. GIES.

Our work thus far has shown that the uric acid output in the urine
of dogs is not materially affected by the administration of chinic acid.
We observed only a slight lowering of the small amounts of uric acid
present in the urine to begin with. This result was obtained when
the animal was in approximate nitrogenous equilibrium on a mixed
diet consisting of hashed meat, cracker meal, lard, bone ash, and
water, and after daily doses, for ten days, of chinic acid in amounts
varying from 1 to 20 grams. These results were obtained before the
recent publication of the data of Hupper’s experiments on himself.
They agree with this observer’s conclusions that the therapeutic
deductions of Weiss, Blumenthal, and others, in this connection, are
without foundation.

Proceedings of the American Phystological Society. xvii

PEPTIC PROTEOLYSIS IN ACID SOLUTIONS OF EQUAL
CONDUCTIVITY.

By WILLIAM J. GIES.

NuMEROUsS digestive experiments with various equidissociated acids,
and with fibrin as the indicator, have invariably given results lack-
ing quantitative agreement. Undigested residue, neutralization pre-
cipitate, and uncoagulable products were determined gravimetrically.
With all conditions exactly the same for each mixture in a series,
except the character of the acid, the digestive products differed not
only in the rate of their formation, but also in their amounts. The
digestive results were particularly discordant in mixtures containing
relatively small amounts of pepsin acting for comparatively short
periods of time. That the anions greatly modified the action of the
common cation seems certain, the SO, anion being especially antago-
nistic in its influence.

The temperature of the digestive mixtures in each experiment was
kept steadily at 25°C. The acids used thus far were of the same
conductivity as a 0.2 percent solution of hydrochloric acid.

I am much indebted to Mr. C. W. Kanolt, of the Department of
Physical Chemistry of Columbia University, not only for the acid
solutions already used, but for others about to be employed in additional
experiments,

ON NUCLEIC ACID.

By Bo As EEVENE-Z

FURTHER investigation into the composition of nucleic acids of dif-
ferent origin resulted in the author improving the process of obtain-
ing the pyrimidin bases so that no silver need be used in order to
obtain thymin and cytosin. On decomposition of the acids of the
spleen and of the pancreas there were found three bases, namely:
thymin, cytosin, and uracil. The acid derived from yeast yielded on
hydrolysis, only two of the bases: uracil and cytosin.

On THE TRUE ELEMENTARY COMPOSITION OF PURIFIED ADRENALIN AND
THE RELATION OF THIS SUBSTANCE TO EPINEPHRIN. By J. J. ABEL.

xviii Proceedings of the American Physiological Society.

On A DIFFERENCE IN THE EFFECT BETWEEN THE SIMPLE CUTTING OF THE
CERVICAL SYMPATHETIC NERVE AND THE REMOVAL OF THE SUPERIOR
GANGLION. By S. J. MELTZER and CLARA MELTZER.

THE CHEMISTRY OF BACTERIAL CELLS, WITH A DEMONSTRATION OF THE
AppaARATUS USED IN OBTAINING THE CELLULAR SUBSTANCE IN LARGE
Amount. By V. C. VAUGHAN.

RESPIRATION EXPERIMENTS IN PHLORHIZIN DiABETES. By G. Lusk and
A. R. MANDEL.

THE FORMATION OF DEXTROSE IN METABOLISM, FROM THE END-PRODUCTS
oF A PANCREATIC DiGEsT OF Meat. By P. G. STILEs.
By invitation.
Dr. Hatat’s OBSERVATIONS ON THE EFFECT OF LECITHIN ON THE GROWTH
OF THE NERVOUS SYSTEM OF THE WHITE Rat. By H. H. DoNnaLpson.
Read by title.
New CRYSTALLINE FORMS OF OXYH/MOGLOBIN ARTIFICIALLY PRODUCED.
By E. T. REICHERT.
Read by title.
OBSERVATIONS ON THE EFFECTS OF VIOLENT AGITATION UPON ARBACIA
Eccs. By S. J. MELTZER.
Read by title.
This journal, 1903, ix, p. 245.
ON THE EXCRETION OF STRONTIUM. By L. B. MENDEL (for H. C. THACHER).
Read by title.

THE

American Journal of Physiology.

MOE. LX. MARCH. 2,5. 1903. NO. I.

SLUDIES: ON THE INFLUENCE OF ARTIFICIAL
RESPIRATION “UPON .StTRYCHNINE- SPASMS
AND RESPIRATORY MOVEMENTS.

By WILLIAM «J. GIES, ann; S., J, MELIEZER.

[From the Laboratory of Physiological Chemistry of Columbia University, at the College of
Physicians and Surgeons, New York.]

I. HISTORICAL.

| Leama to the discovery of the effect of artificial respiration
upon strychnine convulsions, the observation was made by
Meissner and Richter (1) that artificial respiration in curarized
animals will prevent the outbreak of strychnine convulsions even
after the paralyzing influence of curare has worn off. These authors
did not ascribe this favorable result to the effect of artificial respi-
ration, but assumed that, during the period of rest enforced by
curare, the strychnine was partly eliminated from the body and
partly neutralized within the body.

Leube (2), however, came to a different conclusion. Under the
direction of I’ Rosenthal, Leube studied the alleged immunity of the
chicken to strychnine. He found that if artificial respiration be in-
stituted during a strychnine tetanus, the tetanus will soon give way.
If the dose of strychnine be too large, or the artificial respiration
last only a short time, the convulsions may return.

Uspensky (3), also working under the direction of Rosenthal, a year
later studied the influence of artificial respiration upon the spasms

brought on by other poisons. He found that the convulsions which
I

2 William J. Gies and S. J. Meltzer.

followed poisoning with brucin, thebain or caffein may be inhibited
by artificial respiration, but that artificial respiration has.no influence
upon convulsions following poisoning with nicotin or picrotoxin. The
poisons of the latter group, though capable of producing spasms, do
not increase reflex irritability, while those poisons, the convulsions of
which are affected by artificial respiration, have the common charac-
teristic that they do increase reflex irritability. It appears evident,
therefore, that artificial respiration inhibits only such spasms as are
of reflex origin.

From the results of the preceding experiments, Rosenthal (4) con-
cluded that artificial respiration exerts its influence upon the spasms
by means of the increased oxygenation of the blood. He compared
this influence with the effects which artificial respiration exerts upon
the mechanism of respiration itself, in the production of apnoea. In
both cases the oxygen reduces the irritability of the central organs;
in respiration it is the natural irritability of the respiratory centre in
the medulla oblongata, whereas in strychnine poisoning it is the
exaggerated irritability of the spinal cord.

Shortly after the experiments by Rosenthal and his pupils had
been published, Schiff (5) obtained essentially the same results.
Schiff observed, also, that after prolonged artificial respiration a few
animals survived very large doses of strychnine.

During the succeeding interval a few publications have dealt with
the facts and views presented by Rosenthal and his pupils. There
is One assertion on record which is in contradiction to previous state-
ments of fact. This was made by Rossbach and Jochelsohn (7) in
a brief preliminary communication, which was never supplemented
by a full publication of their experiments. They claim that artificial
respiration has no soothing influence whatsoever upon strychnine
spasms. These observers make additional statements in this con-
nection which are not in conformity with the general experience,
but which need not be discussed here. All other investigators
confirm, unreservedly, the fact that artificial respiration exerts an in-

' According to this view the favorable results in the experiments of MEISSNER
and RICHTER were due solely to the artificial respiration. We should like to call
attention, however, to the experiments of CH. RicHET (6) in this connection.
RicHeT found that after poisoning with large doses of strychnine, the life of the
animal is greatly prolonged, if, in addition to the artificial respiration, curare is
also administered. RiCHET makes no reference to the work of either MEISSNER
and RICHTER, or to that of ROSENTHAL and his pupils.

Artificial Respiration and Strychnine Spasms. 3

hibitory influence upon the strychnine convulsions. There is, how-
ever, a divergence of opinion regarding the nature of this inhibitory
influence.

Ebner (8) and Buchheim (9) stated that they were able to induce the
same soothing effect by simple movements of the body and extremi-
ties of the animal, and denied, therefore, that oxygenation has any-
thing to do with the favorable action. They believe that the muscular
movements are the favorable factors in the relaxation of the spasms.
L. Pauschinger (10), however, working under Rosenthal’s direction,
could easily dismiss this contention by showing that the authors sim-
ply employed the now well-known Schultze’s method of instituting
artificial respiration without opening the trachea.

Brown-Séquard (11), after confirming the fact that the convulsions
may be relieved by artificial respiration, denied that the favorable
effect is due, as Rosenthal believed, to a greater charging of the blood
with oxygen. He was of the opinion that artificial respiration causes
a mechanical stimulation of the nerves of the lungs, thorax, and dia-
phragm, and thus affects an inhibition of the reflex centres.

The statements of Brown-Séquard were contradicted by Filehne
(12). Later we shall have occasion to return to the works of both
these observers.

Rosenthal’s view was supported by Ananoff (13), who, in a brief
communication, reported that animals breathing pure oxygen show
a greater resistance to the effects of strychnine.

From 1878 to 1900 there is no publication to be found bearing
on this subject. In the last-named year this question was studied
by Osterwald (14) in the Pharmacological Institute of Gottingen.
Osterwald, like Ananoff, put animals under glass bell jars through
which a stream of oxygen was conducted. Experiments with mice
did not yield striking results, but a few positive results with guinea-
pigs led Osterwald to the unreserved support of the opinion that the
favorable influence of artificial respiration is due to the greater intro-
duction of oxygen into the blood.

Similar experiments were made last year by Von Czyhlarz (15)
with guinea-pigs as well as with rabbits. His experimental results
may be more appropriately discussed farther on.

The present status of this subject is, then, as follows: It is now
a well-established fact that artificial respiration may prevent the out-
break of convulsions due to strychnine poisoning, or may inhibit

4 William J. Gres and S. J. Meltzer.

them if already present, provided the dose of strychnine be not too
large. The striking feature of its action is the perfect relaxation of
the convulsed muscles, the absence of any muscular rigidity or any
kind of tremor.! Artificial respiration here, apparently, inhibits the
artificial increase of reflex-irritability. It is now the consensus of
opinion that this inhibition is produced by increased introduction of
oxygen into the blood, and that the mechanical effect of the expan-
sion of the lungs, suggested by Brown-Séquard, has no share in the
result.

This position, did not appear to us to be entirely satisfactory, for
the following reasons: the soothing influence of artificial respiration
upon the increased reflex-irritability due to strychnine is apparently
identical with its soothing influence upon respiration itself, z. e., with
the production of apnoea.2, We have already stated above that Ros-
enthal, who may be said to be the discoverer of these phenomena,
looked upon both as processes of identical character — the inhibition
of the normal reflex-irritability in one and inhibition of the increased
reflex-irritability in the other. As regards the causation of apnoea it
now seems to be a settled conviction that this state can be brought
about by hyperoxygenation as well as by the mechanical distention
of the lungs. Why, then, should it be different with the inhibitory
effect of artificial respiration upon the strychnine spasms?

Furthermore, the evidence upon which the present prevailing
view is based does not appear to us to be entirely conclusive. The
chief points in the evidence are: 1. Filehne’s work in disproving
Brown-Séquard’s claims of the disappearance of the effect of arti-
ficial respiration after section of the cord or the vagi; 2. Ananoff’s,

* Some writers, when speaking of the favorable effect of artificial respiration,
mean simply that it prolongs life. Life can be prolonged by artificial respiration,
however, even if the administered strychnine dose is very large: but then the
tonic and clonic convulsions continue even during the most energetic artificial
respiration,

* The inhibitory effect of artificial respiration upon the complex mechanism

of respiration may show itself in several ways: 1. The animal stops its normal

abdominal and thoracic respiratory movements. 2. The concomitant respiratory

movements of mouth and nose stop during artificial respiration. 3. All respiratory
movements remain quiet for some time immediately after discontinuation of the
artificial respiration. Usually only the last form of inhibition is termed apnoea.
It is obvious, however, that the arrest of the respiratory movements during arti-

ficial respiration also belongs to the inhibition phenomena, and ought to be
included in the term apnea.

Artificial Respiration and Strychnine Spasms. 5

Osterwald’s, and Von Czyhlarz’s experiments in producing the same
favorable effect by simple normal inhalation of pure oxygen.

Considering the last line of evidence first, we have to exclude at
the outset the testimony of Ananoff. In his short communication
Ananoff speaks only of artificial respiration prolonging life, and does
not mention the absence of spasms during the process which, as
remarked above, is the essential point.

Osterwald’s successful experiments were made on guinea-pigs
(animals which are very resistant to strychnine), and were few in
number. These strictures appear the more important when we read
them in the light of the results reported by Von Czyhlarz. This
last-named observer made nine experiments with guinea-pigs. In
each experiment one animal inhaled pure oxygen and the other (con-
trol) inhaled air. In four of these experiments both animals had only
marked hyperesthesia. In one both animals had tetanus and survived.
‘In the remaining four experiments the oxygen animals had marked
hyperzesthesia, whereas the control animals had non-fatal convulsions.
In the majority of the experiments, therefore, there were hardly any
differences between the oxygen-breathing animals and the control
animals, while the differences observed in the minority of the exper-
iments were only of a minor character.

Von Czyhlarz’s experiments on rabbits are still more instructive.
Here he records eight experiments. In three experiments both
animals had non-fatal tetani. In two both had fatal tetani. In one
the oxygen animal had a fatal tetanus, and the control survived. In
the remaining two experiments the oxygen animals died, while the
controls survived their tetani. We fail to see in any of these experi-
ments with rabbits even the shadow of proof that the inhalation of
oxygen can suppress in these animals the increased reflex-irritability
due to strychnine poisoning. Now, all the successful experiments
ever made with artificial respiration were upon rabbits! If, however,
it be admitted that the results of the above experiments with pure
oxygen do prove that the oxygenation of the blood can neutralize
to some degree the effect of strychnine, they surely do not prove that
the mechanical distention of the lungs has no share in the effects
produced by the classical method of artificial respiration.

There remains the work of Filehne, which was conducted in con-
travention of the claims put forward by Brown-Séquard. The latter
observer, as stated above, was of the opinion that the arrest of res-
piratory movements of the animal (apnoea), as well as the arrest

6 William J. Gies and S. J. Meltzer.

of the spasms in strychnine poisoning, both of which artificial respi-
ration is capable of effecting, are not due to hyperoxygenation of the
blood. The arrest in each case was attributed to the mechanical
irritation of the branches of the vagus, the phrenic “or other dia-
phragmatic nerves,’ caused by the forced insufflation of air into the
lungs. In support of his view Brown-Séquard states that transverse
section of the spinal cord above the origin of the phrenic nerves or
below their origin, or even section of the vagi, removes the arresting
influence which artificial respiration exerts upon the respiratory
movements.

In contravention of these statements Filehne reports that he
tested these claims in a series of experiments and could not confirm
them. An analysis of Filehne’s experiments reveals the fact, how-
ever, that no experiment was made in which he studied the arrest of
strychnine spasms by artificial respiration in animals whose spinal
cord was cut. His attempts in this line were confined to the demon-
stration of the presence of apnoea after the severance of the cord.
Even in these he succeeded in cutting the cervical cord in only one
experiment. Ina few other experiments he tried to crush the cord
in young animals by forcibly constricting the cervical column with a
string. The crudity of such a method hardly inspires confidence in
the results attained by it.

Filehne further records a few experiments in which artificial res-
piration arrested strychnine spasms after cutting the vagi. These
experiments, however, as was pointed out by Filehne himself, seem
also to demonstrate that the cutting of the vagi visibly impairs the’
favorable effect of the artificial respiration.

We see, therefore, that the experiments to show the effect of inha-
lation of pure oxygen are still far from being decisively in favor of the
exclusive oxygen theory. We find, further, that Filehne’s work can-
not be considered a sufficient refutation of Brown-Séquard’s mechan-
ical theory. Apparently, much more work must be done before the
questions raised here can be satisfactorily answered.

Il. Our Own EXPERIMENTS WITH SECTIONS OF CORD AND VAGI.

From the above analysis it is evident that the claims of Brown-
Séquard have not yet been properly tested, and that they deserve,
therefore, to be investigated anew.

Brown-Séquard believed that the insufflation of air into the lungs

Artificial Respiration and Strychnine Spasms. 7

irritates the endings of the vagus as well as of the phrenic and
‘other diaphragmatic nerves,” whatever the latter may be. He
might as well have said, also, that the sensory nerves of the thorax
wall and the pleura might be stimulated by the rhythmical pressure
of the artificial respiration. A more suggestive conjecture would be
that the rhythmical pressure upon the contents below the diaphragm
irritates the splanchnic nerves. We now know that stimulation of
the central ends of the splanchnic nerves causes inhibition of inspi-
ration (16). Experiments which exclude only one set of nerves, while
the other paths of innervation remain intact, afford inconclusive
evidence that the mechanical irritation of the nerves has no share
in the inhibiting effect of artificial respiration. Our experiments
were therefore directed in the first place toward the study of the
action of artificial respiration on animals in which the vagi were
cut, and, at the same time, the spinal cord severed at one place or
another.

General method. —The experiments were made on rabbits, which
were kept stretched on a holder, and were under ether anesthesia
during the operations. The artificial respiration was administered
by bellows through a tracheal tube. The bellows were fastened to
the table on which we operated, and were manipulated by hand. An
average of thirty uniform strokes per minute was maintained, which
caused a pressure rarely exceeding 36 mm. Hg. Each stroke with
the bellows caused a distinct jar of the table upon which the animal
was resting, a fact of importance in our experiments. We employed
strychnine nitrate. An extensive experience has taught us that
white rabbits are more sensitive to strychnine than colored ones.
We have found that 0.45 mgm. of strychnine nitrate per kilo is a
surely toxic dose for a white rabbit, and 0.5 mgm. for a gray one.
Although this knowledge might have sufficed, we employed controls
in almost every experiment. While our main object was the study
of the influence of artificial respiration upon the spasms of strych-
nine, we also made note of the relation of artificial respiration to
apnoea under these conditions.

Abbreviated protocols of our various experiments are given below:

Experiment J, — Gray and white male rabbit, 1920 grams. ‘Tracheotomy.
5-30 P.M. Cord cut between third and fourth vertebra.
5-34. Strychnine injected, 0.6 mgm. per kilo.
5-35. Artificial respiration started, 30 to 35 mm. pressure.

8 William J. Gres and S. J. Meltzer

Experiment I— (continued).

5-37. Both vagi cut.

6.07. Artificial respiration discontinued. Animal was watched for ten
minutes longer, and then was removed from board. During the forty-
seven minutes after the injection of the fatal dose of strychnine, the animal
did not show even any hyperesthesia due to strychnine, although the table
was jarred, the rabbit untied and its paws squeezed in testing for reflexes
of the paralyzed hind limbs, and the animal even removed from the table.

With the dose administered in Experiment I a normal animal
would have succumbed to a fatal tetanus in less than thirty minutes!
The inhibitory effect of the artificial respiration was distinctly man-
ifest, although the nervous paths of the vagi and the splanchnici
were cut off. However, during the entire period of artificial respi-
ration, there was in this experiment no full suppression of the
animal’s own breathing. Further, the concomitant respiratory move-
ments of mouth and nose continued, and became very pronounced
after the vagi were cut. There was no sign of apnoea after stopping
the artificial respiration. In short, artificial respiration produced no
apnoea. Possibly the artificial respiration with only 30 to 35 mm.
pressure was not strong enough to cause apnoea in this large animal.
But it remains a noteworthy fact that in an animal in which the
paths through the vagi and splanchnici were blocked, a certain
degree of artificial respiration was sufficient to influence the strych-
nine spasms, but not to cause apncea!

Later the same animal was given another injection of strychnine —
o.5 mgm. per kilo. It had distinct convulsions after sixteen minutes.
The noteworthy fact was observed that the convulsions did not appear
simultaneously in the anterior and the posterior parts of the animal,
but occurred in the part above the section of the cord usually before
the part below. Subsequently, when the animal recovered from these
convulsions, it was killed by asphyxia. It again had convulsions,
which appeared in the hind part later than in the front, both sets of
convulsions apparently continuing independently of one another.

Experiment I. — Gray rabbit, 1030 grams. ‘T'racheotomy.
5-37 P.M. Cord cut opposite third dorsal vertebra.
5.42. Injected strychnine, 0.67 mgm. per kilo.
5.44. Artificial respiration started.
5.46. Artificial respiration slackened, signs of convulsive movements
appeared. Artificial respiration immediately increased, perfect rest again.

Artificial Respiration and Strychnine Spasms. 9

Experiment [1 — (continued).

6.00. Both vagi cut, no independent respirations, but concomitant
breathing appears and remains throughout artificial respiration.

6.08. Artificial respiration stopped, soon vibration in upper part, and
gasps.

6.09. Artificial respiration resumed, followed by rest again.

6.14. Artificial respiration discontinued, apnoea for a few seconds.

6.15. Convulsions in front parts, not in hind parts.

6.16. Convulsion in hind legs, none in front ; soon, however, opisthot-
onus and death.

In this experiment, with a still larger dose of strychnine, the arti-
ficial respiration could not abolish the convulsions permanently, but
while it was continued, arrested them for thirty-three minutes, the
animal being perfectly relaxed, and even without hyperesthesia
during the entire period. In this smaller animal artificial respira-
tion suppressed the independent respirations of the animal, and
even caused a very brief period of apnoea, but it had no effect upon
the concomitant respiratory movements of mouth and nose after the
vagi were cut.

Lixperiment I[[. — White rabbit, 1400 grams. Tracheotomy.

5:03 P.M. Cervical cord cut opposite fifth vertebra, paralysis of hind
and fore legs; no voluntary breathing. Artificial respiration begun.

5-11. Strychnine injected 0.6 mgm. per kilo.

5-31. While handled, slight and brief spasms ( ?).

5-33. Both vagi cut, gasping and other concomitant respiratory move-
ments cannot be suppressed ; pinching of any part brings out a tetanic
convulsion confined to that part and lasting only as long as the part is
handled ; reflexes.

5-45. Artificial respiration discontinued, brief apnoea, then attempts to
breathe ; convulsions in upper part alone, later in lower part alone.

5-46. Rabbit dead.

In white rabbits 0.6 mgm. strychnine per kilo is a rapidly fatal
dose. For thirty-five minutes, while the artificial respiration lasted,
there were no real convulsions, but throughout the entire experiment
there was a marked reflex hyperzesthesia, upon which the artificial
respiration had apparently only a moderate inhibiting influence.
After the vagi were cut the artificial respiration could not longer
arrest the strong concomitant respiratory movements.

IO Witham J. Gies and S. J. Meltzer.

Experiment IV a. — Gray rabbit, 1840 grams. ‘Tracheotomy.

5-17 P.M. Cord cut opposite fifth cervical vertebra. Rabbit collapsed,
no voluntary respiration, and only faint heart-beat. Artificial respiration,
elevation of rear end of rabbit holder, and compression of abdomen.

5-46. Animal fully recovered.

6.o1. Strychnine injected (0.06 mgm. per kilo strychnine sulphate
+ 0.06 per kilo strychnine nitrate).

6.33. Both vagi cut. Extremities squeezed or pulled, board hit, table
thumped, but no convulsions, and even no hyperzsthesia. Independent
voluntary respiration continually present.

6.41. Artificial respiration discontinued, no apncea, breathes well.

6.46 to 6.49. Short tetanic convulsions either in front extremities alone,
with legs upward, or in the four extremities with front legs downward.

6.50. Tetanus, opisthotonus, and death.

Lixperiment LV 6. — Control, gray rabbit, 1760 grams.
4.44 P.M. Cord cut between third and fourth dorsal vertebre.

6.03. Injected strychnine (0.06 mgm. per kilo strychnine nitrate + 0.06
per kilo strychnine sulphate).

6.30. On striking table tetanus in all parts at once, opisthotonus and
death.

Although the animal in Experiment IVa received a fatal dose of
strychnine and was subjected to all sorts of irritations, it manifested
neither convulsions nor hyperzesthesia during the entire time it re-
ceived artificial respiration. The control animal, Experiment IVb,
on the other hand, had a fatal tetanus when the table was struck,
twenty-seven minutes after receiving the strychnine. Six minutes
after discontinuance of artificial respiration the strychnine poisoning
became manifest also in Experiment IVa. The artificial respiration
apparently only inhibited an increase of reflex-irritability but did not
destroy the poison in the body; neither was the strychnine sufficiently
eliminated from the body during the period of artificial respiration to
prevent tetanus.

In this experiment artificial respiration did not suppress the volun-
tary breathing, nor did it produce any apnoea after its discontinuance.

Experiment Va.— Gray and white rabbit, 1240 grams. ‘Tracheotomy. Ar-
tificial respiration for three minutes, voluntary and concomitant breathing
suppressed. Artificial respiration stopped, apncea only two seconds. Ar-
tificial respiration resumed, voluntary and concomitant breathing sup-

pressed in one minute. Artificial respiration stopped after two minutes,
apnoea eight seconds.

Artificial Respiration and Strychnine Spasms. 11

Experiment Va— (continued).

5.45 P.M. Cervical cord cut between fifth and sixth vertebree, animal in
good condition. Artificial respiration resumed, voluntary and concomi-
tant respiration suppressed only after seven minutes. Artificial respiration
stopped, apnoea eight seconds. Artificial respiration immediately begun
again, no voluntary and concomitant movements.

5.57. Strychnine nitrate injected, 0.65 mgm. per kilo.

6.07. Vagi cut; voluntary and labored concomitant breathing set in,
each suppressed after seven minutes.

6.34. Artificial respiration stopped, apncea eight seconds. Animal
observed until 6.52. Although extensively handled during the fifty-five
minutes since strychnine was injected, no sign of hyperesthesia.

Later on asphyxia was caused by inhalation of hydrogen, and again by
clamping of the trachea. Tetanic convulsions appeared only in the an-
terior part ; the legs were directed towards the head.

Experiment Vb. — Control, gray and white rabbit, 1250 grams.
6.03 p. M. Injected strychnine nitrate 0.63 mgm. per kilo.
6.27. Animal stiff.
6.29. Tetanic convulsion.
6.33. When lifted there was a convulsion which the animal survived.

In this experiment the effect of artificial respiration upon strych-
nine spasms was very plain. They were entirely suppressed during
the fifty-five minutes of observation, although the control animal
began to show a distinct strychnine effect even twenty-four minutes
after injection. The inhibitory effect upon the respiration was
retarded by section of the cord as well as by section of the vagi, but
finally a distinct apnoea was attained. :

In the last three experiments the vagi and splanchnici, as well as
the sensory fibres of the pleura and thoracic wall, at least most
of them, were separated from the respiratory centre, etc. The
roots of the brachial plexus were apparently divided in two parts, for
when convulsive movements occurred in the upper part alone, the
front legs were directed toward the head.

Experiment VI a. — Gray and white male rabbit, 1550 grams. ‘Tracheotomy.

Artificial respiration (25-30 mms. Hg) for three minutes. Voluntary and

concomitant breathing soon suppressed. Artificial respiration stopped,
apnoea three seconds.

4.43 P.M. Cord cut “between fourth and fifth cervical vertebr,” breath-

ing stopped. Artificial respiration begun. Heart, lid reflex, etc., all right.

Concomitant breathing soon suppressed. Artificial respiration stopped,

12 Witham J. Gies and S. /. Meltzer.

Experiment VI a— (continued).

apnoea fifteen seconds, soon ‘head breathing.” Artificial respiration
resumed again, head breathing soon suppressed.

4.57. Injected strychnine nitrate o.7 mgm. per kilo.

5.16. Both vagi tied off, concomitant breathing appeared, but disap-
peared again after four minutes:

5.32. Artificial respiration stopped, apnoea fifteen seconds, then “head
dyspncea”’ ; no hypereesthesia otherwise. Artificial respiration again.

5.40. Artificial respiration stopped, apnoea fifteen seconds, then gradual
development of head dyspnoea and asphyxia.
5-43. Heart stopped. No convulsions.

Experiment VT 6. — Control, gray and white rabbit, 1050 grams.
gray » 10508
5.01 Pp. M. Injected strychnine nitrate o.7 mgm. per kilo.
5-07. Convulsions, did not survive.

Ir Experiment VlIa when artificial respiration was stopped there
appeared now and then a very faint indication of thoracic movement.
Possibly it was produced passively by the dyspnoeic contraction of
the cervical muscles. The autopsy showed that the cord was severed
at the lower border of the fourth cervical vertebra, but the cut was
diagonal and possibly a few fibres of the phrenic escaped. At all
events this experiment is a strong demonstration of the efficiency of
artificial respiration in suppressing strychnine spasms, and in pro-
ducing apnoea, even after the vagi, splanchnici, brachial plexus,
and almost all of the phrenic nerves are excluded. Although the
control animal had a convulsion six minutes after injection (0.7
mgm. per kilo), the animal in Experiment Vla manifested no
sign of strychnine spasms either during the forty-three minutes of
artificial respiration or during final asphxyia. Furthermore, there
was no concomitant breathing during the artificial respiration, and
an apnceic pause was present after each interruption.

Experiment VII a. — Gray female rabbit, 1120 grams. ‘Tracheotomy.

5-20 P.M. Cord cut near upper border of fifth cervical vertebra, animal
breathes normally. Artificial respiration for a few minutes, voluntary res-
piration persistent. Artificial respiration stopped, no distinct apnoea.
Artificial respiration begun again.

5.30. Both vagi cut, labored, concomitant breathing; voluntary and

5?
concomitant breathing subsiding slowly. °
5-35- Injected strychnine nitrate, 0.7 mgm. per kilo.
6.05. Artificial respiration discontinued, no apnoea. Animal observed

five hours longer. Had no sign of strychnine poisoning. When then

Artificial Respiration and Strychnine Spasms. 1g
Experiment Vila— (continued ‘
given a comparatively large dose of strychnine, it had a number of short
convulsions in either of the two parts, independently of one another.

Experiment VII b.— Control, gray female rabbit, ro50 grams. For better
comparison had ether anzesthesia for a few minutes.
5.40. Injected strychnine nitrate o.7 mgm. per kilo.
5-59. Convulsions, succumbed in six minutes.

This experiment again is a classical demonstration of the inhibitory
effect of artificial respiration upon the strychnine spasms even after
section of cord and vagi. The effect upon respiration was less
pronounced.

Experiment VITI a. — Gray female rabbit, 1750 grams. ‘Tracheotomy.

3.59 P.M. Artificial respiration begun, only very slight concomitant res-
piratory movements.

4.01. Cord cut between second and third cervical vertebrae, concomitant
breathing greatly increased, after a few minutes decreased again.

4.09. Artificial respiration discontinued, apnoea for a few seconds, then
dyspnoea. Artificial respiration again.

4.15. Animal recovered from anesthesia. Injected strychnine nitrate,
0.7 mgm. per kilo. Heart-beat, lid reflex, etc., normal until 4.30, when
heart-beats became slower and labored concomitant respiratory movements
reappeared.

4.35. The respiratory movements rapidly diminished ; no lid reflex.

4-37. Heart-beats faint.

4-39. Death.

During the fifteen minutes after injection the animal was perfectly nor-
mal, but there was no sign of hypereesthesia.

Lxperiment VILT 6. — Control, gray rabbit, 1700 grams.
4.18 P.M. Etherized (for comparison) and kept under ether until 4.23.
4.19. Injected strychnine nitrate o.7 mgm. per kilo.
4.33- Tetanic dance.
4.36. Convulsion terminating fatally at once. Although this animal
was under the influence of ether while strychnine was injected, fourteen
minutes after injection it manifested the unmistakable effects of this drug.

In Experiment VIII a the vagi, splanchnici, and all thoracic nerves,
including the phrenici, were excluded. Although the animal died
early, it lived long enough, and was normal long enough, to demon-
strate that the strychnine had no effect so long as the artificial respi-
ration was continued. There was once also a distinct apnceic pause.

14 Witham J. Gies and S. /. Meltzer.

Experiment IX a. — Gray female rabbit, 1750 grams. Tracheotomy.

4.00 P. M. Artificial respiration begun.

4.02. Injected strychnine nitrate, 0.7 mgm. per kilo. The voluntary
respirations were completely suppressed. Six minutes of artificial respir-
ation, concomitant breathing not completely suppressed.

4.07. Cord cut at third cervical vertebra. Heart-beat, lid reflex, etc.,
normal, concomitant respiratory movements increased ; remained unsup-
pressed throughout entire experiment.

4.12. Vagi cut. Animal normal throughout the remainder of the ex-
periment. ‘There was apparently a hyperesthesia in the lower part,
pressing or pulling legs was followed by contraction or tremor in legs, but
these continued only as long as pull or pressure lasted. Blowing on
animal, hitting table, no effect. Tremor in abdominal muscles ; they even
seem to contract synchronously with artificial respiration, simulating
superficial independent voluntary breathing.

4.42. Artificial respiration stopped, all contraction and tremor disappear
immediately (are due apparently to the local stimulus of the artificial res-
piration); head dyspnoea appears. Artificial respiration resumed again.
Extremities and tail repeatedly pinched, pulled, etc., response with local,
short reflexes, either during stimulation or immediately after. Pinching
ear or other parts of head produce no reflex, but voluntary motion, moving
away.

5-03. Artificial respiration discontinued; head dyspnoea, but no other
movement of body.

5-04. Slight movement, and later, vibration only in front legs.

5-05. Sudden tetanus in all four extremities, followed by clonic con-
vulsions.

5.06. Artificial respiration resumed. Lid reflex and heart-beat soon
normal again, no more convulsions.

5-10. Artificial respiration discontinued again.

5-12. Sudden tetanus. Artificial respiration at once, and tetanus
stopped suddenly. This procedure was repeated several times with same
result, but sometimes tetanus stopped even while there was no artificial
respiration.

Experiment LX b. — Control, gray female rabbit, 1600 grams.
4-22. Strychnine nitrate, o.7 mgm. per kilo.
4.35. Tetanic convulsions.
4-45. Blown on, immediately violent convulsion, succumbs.

In Experiment IX a the section of the cord was above the phrenici,
and the influence of the vagi and all other nerves concerned was pos-
itively excluded. The animal had a dose of strychnine which proved

Artificial Respiration and Strychnine Spasms. 15

fatal to the control rabbit in less than half ‘an hour. Although the
animal in Experiment IX a was continually handled, and the reflexes,
etc., tested, for an hour, while artificial respiration lasted, there was
no reaction which could be ascribed to the effect of strychnine.
Pounding the table or blowing on the animal had no effect at all.
Pinching or pulling a leg brought out a local reflex which was
apparently due only to the increased reflex-irritability caused by the
section of the cord. Pressing one hind leg, for instance, would bring
out a short flexion or extension of the opposite, or of a front leg.
The artificial respiration caused short contractions of the abdomi-
nal muscles. But pinching any part of the head caused no reflex-
response. The strychnine, however, was not destroyed within the
body, nor sufficiently eliminated from it. Soon after stopping the
artificial respiration there appeared convulsions and tetani, which by
their entire character were apparently due essentially to the strych-
nine and not to asphyxia, or at least not to asphyxia alone. But these
convulsions also could be stopped instantly by artificial respiration.

The influence of artificial respiration upon apnoea was not carefully
noted in this experiment, but the concomitant respiratory movements
continued during the hour while the artificial respiration lasted,
although their intensity gradually decreased.

Our first series of experiments brought out one positive result.
The claim of Brown-Séquard, that section of the cord or of the vagi
abolished the arresting influence which artificial respiration exerts
upon strychnine spasms, is entirely unfounded. Not only does sec-
tion of the vagi alone, or of the cord alone, fail to impair this influ-
ence, but even cutting the vagi, combined with such section of the
cord as excludes a!l influences of the splanchnic, diaphragmatic, and
thoracic nerves, apparently does not interfere with the inhibitory
influence of artificial respiration upon strychnine spasms. There
were no convulsions in any of our experiments as long as suffi-
ciently strong artificial respiration was administered. In many
experiments no convulsions appeared even after the artificial res-
piration was stopped, although in all cases doses of strychnine
were employed which by control experiments were proved to be
surely toxic and mostly fatal. In some experiments artificial respi-
rations arrested instantly the tetanic convulsions which were per-
mitted to break out.

The doses of strychnine which we employed were, however, not

16 Witham J. Gies and S. J. Meltzer.

much above the toxic or fatal minimum. Possibly section of the cord
or vagi does interfere somewhat with the degree of favorable influ-
ence which artificial respiration might exert under such conditions.
Filehne, who admits some impairment due to the section of the vagi,
does not state the weight of his animals. Possibly, however, the —
doses which he employed were a trifle too large. Overdosage might
also explain the claims put forward’ by Brown-Séquard. But the
description of his experiments is too brief to permit any very definite
interpretation. In fact it is not even evident that Brown-Séquard’s
conclusions regarding the relations of the sections of cord, or vagi,
to the arresting influence of artificial respiration upon strychnine
spasms were derived from actual experiments, and that they were not
mere inferences from the experiments he made on the production of
apnea.

Regarding the latter, our own experiments have indeed demon-
strated that section of the cord and the vagi impairs more or less the
production of apnoea by artificial respiration. In some cases after
section of the cord, and especially after additional section of the vagi,
neither the voluntary respirations nor the concomitant respiratory
movements could be suppressed. This was observed in some of the
larger animals. Possibly the degree of ventilation employed in our
experiments was not sufficient to accomplish this end in an animal
with a comparatively large thorax. However, in all the experiments,
section of the cord, or of the vagi, even during artificial respiration,
immediately brought out again the voluntary breathing of the animal
and especially the concomitant respiratory movements. It invariably
took a much longer time to suppress the latter after section than
before it.

Our experiments also showed that while artificial respiration com-
pletely suppresses the increased reflex-irritability due to strychnine-
poison, it does not interfere, at least not strikingly, with the increased
reflex-irritability induced by section of the cord. In all cases we were
able, with little or no difficulty, to produce distinct reflexive move-
ments by pinching a leg, touching an eye, etc., the posterior extrem-
ities responding more readily than the anterior ones. In one case,
with section above the phrenici, each blow of the bellows brought
out a contraction of the abdominal muscles simulating spontaneous
breathing, which ceased on stopping the artificial respiration.

We noticed also, in the cases in which mild convulsions appeared
after artificial respiration was stopped, that the parts lying above the

Artificial Respiration and Strychnine Spasms. 7

line of section of the cord and those lying below it had their convul-
sions independently of one another. They were mostly insynchro-
nous. In the experiments in which section of the cord was near the
fifth cervical vertebra, the interesting observation was made that
when the convulsions occurred in the anterior part, the anterior legs
took part in it by moving toward the head, and that when the pos-
terior part was convulsed the anterior legs moved toward the tail,
pressing against the body. When, however, a violent tetanus broke
out, the spasm convulsed all parts nearly simultaneously.

Thus it is evident that our experiments have established the fact
contended for, but not proved by Filehne, namely, that section of the
cord and vagi does not interfere with the inhibitory influence which
artificial respiration exerts on strychnine spasms. But does this fact
prove that the inhibitory influence of artificial respiration is due to
the chemical influence of the oxygenation of the blood and to this
alone? Does this fact indicate that the mechanical act of rhythmical
insufflation has no share in the inhibitory influence ?

The persistence of the favorable influence observed after section
of cord and vagi could only then serve as an irrefutable proof if the
claim for the mechanical share had been restricted to a hypothesis
that the inhibition acts either through the agency of the respira-
tory centre or through the inhibitory mechanisms of the brain. If
this is what Brown-Séquard meant, his theory is surely disproved
by our experiments. The favorable influence of artificial respiration
against the increased irritability of the spinal cord continues even
after the cord has been severed from the controlling parts above it.
But why restrict our hypothesis? We know that any reflex may be
inhibited within the spinal cord by any mechanical stimulation of any
part of the body. We have also seen in our experiments that, in an
animal with a severed cord, artificial respiration caused rhythmical
contraction of the abdominal muscles. This fact shows that the
insufflations into the lungs, and the consequent abrupt increase of
pressure upon the organs within the thoracic cavity, result in stimu-
lating also the dorsal nerves imbedded in the abdominal section of the
body. Furthermore we know that this insufflation causes an inhibi-
tion of centres lying within the medulla (respiratory, vaso-motor,
cardio-inhibitory centres). Why then should it not be assumed that
the rhythmical insufflations into the lungs stimulate all nerves within
the thoracic and abdominal regions and thus inhibit increased reflex-
irritability in all parts of the cord ?

18 Witham J. Gres and S. J. Meltzer.

The hypothesis formulated by Brown-Séquard is certainly untenable.
That the arrest of the spasms can be due to the mechanical stimu-
lation of the endings of the vagi, the phrenic and “ other diaphragmatic
nerves ” alone, our experiments with section of the vagi and the cord
have proven conclusively. But no cutting of the cord is capable of
disproving the hypothesis that rhythmical insufflation is a mechanical
stimulus for all the nerves within the trunk, by means of which an
inhibition is caused in every section of the spinal cord above a cut as
well as below it.

The question, therefore, is still open: Does the mechanical element
involved in artificial respiration have a share in the arrest of the
strychnine spasms, just as it is now generally assumed that it has a
share in the production of apnoea ?

III. ARTIFICIAL RESPIRATION WITH HYDROGEN,

For the solution of this question a method presents itself which at
first sight appears to be quite simple. Previous investigators who
desired to prove that it is the chemical factor which causes the arrest
of the spasms have tried to introduce the oxygen without the compli-
cation of the mechanical element. Desiring to test the efficiency of
the mechanical factor, we sought to determine the effect of artificial
respiration with its chemical factor removed ; 2.¢., artificial respira-
tion with an indifferent gas. It was partly by this method, indeed,
that the value of the mechanical element in the production of apnoea
was ascertained. We have, therefore, endeavored to study the effect
of artificial respiration with pure hydrogen upon the strychnine
spasms.

General method.— The method we employed was comparatively
simple. Bellows were connected on one side with a gasometer con-
taining pure hydrogen, and on the other side with the trachea of the
animal. The tube connecting the bellows with the gasometer con-
tained a valve which permitted the entrance of the gas into the bel-
lows, but prevented it from going back to the gasometer. The tube
connecting the bellows with the trachea contained a valve permitting
the escape of the gas in the direction of the trachea, but preventing
its return to the bellows. The expiratory air escaped through a lat-
eral tube submerged under water (Miiller’s water valve), by which
arrangement air was prevented from entering into the trachea through
the expiratory aperture during a voluntary inspiration. All the con-

Artificial Respiration and Strychnine Spasms. 19

nections were carefully made air tight. Each suction of the bel-
lows brought hydrogen into it, and each compression drove the
hydrogen into the lungs. The pressure was regulated by means of
a stop-cock carried by the expiratory tube, and it was registered by
a manometer connected by a T tube with the bellows-trachea tube.

We had, of course, no expectation of being able to continue the
exclusive inhalation of hydrogen long enough to prevent the develop-
ment of the strychnine poisoning, in the same manner as we suc-
ceeded in preventing it by the artificial respiration of air. We had
observed that when once a tetanus broke out in our experiments it
could be suppressed instantaneously by artificial respiration. In fact
this instantaneous effect appeared to us to be in favor of the theory
of a mechanical effect, since an effect due to a sufficient increase of
oxygen in the blood could hardly develop so promptly after the first
few strokes with the bellows. We therefore had reasonable expecta-
tions of witnessing the same instantaneous effect when pure hydrogen
would be insufflated, or at least of observing it, long before the
unavoidable asphyxia would finally compel the discontinuation of
this gas. However, the first preliminary experiment, to determine
the effect of insufflation of pure hydrogen upon the production of
apnoea, brought us a surprise.

Experiment X.— White rabbit, 1700 grams. Tracheotomy, connected with
bellows and gasometer, expiratory tube submerged. Insufflation of hy-
drogen for a brief period, apnoea for a few seconds. Repeated a few
times with same result. Encouraged by the absence of asphyxia, the
insufflation was continued consecutively for eighteen minutes, during
which time there was no voluntary breathing, no concomitant respiratory
movements, and no perceptible cyanosis of visible mucous membranes.
After discontinuing the insufflation of hydrogen an apnoea of fifteen
seconds appeared, but this was followed immediately by rapid superficial
breathing and very rapid, faint heart-beats. Artificial respiration with air
improved this condition, but the animal soon died through an accident.

Eighteen minutes’ insufflation of pure hydrogen without asphyxia !
That was surely an unexpected result. Before discussing it, how-
ever, we should quote a few of these hydrogen experiments in which
also toxic doses of strychnine were injected.

Experiment XT a. — White rabbit, 1240 grams. Tracheotomy.
4.30 P.M. Injected strychnine nitrate, 0.6 mgm. per kilo.
4.33. Trachea connected with bellows, etc. Continued insufflation
without incident till 4.58, when tetanic convulsions set in. Continued

20 William J. Gies and S. J. Meltzer.

Experiment XIa— (continued).

insufflation until 5.or without favorable effect. Insufflation stopped,
animal thoroughly asphyxiated.

5-05. Artificial respiration with air.

5.07. Discontinued, no apnoea, immediately rapid breathing, a minute
later convulsions, which continued for a few minutes. Animal killed.

Experiment XT b. — Control, gray and white rabbit.
5-17. Injected strychnine nitrate, o.5 mgm. per kilo.
5-30. Convulsions broke out.

The animal in Experiment XI a was a white rabbit which, as men-
tioned above, was more susceptible to strychnine than the gray and
white one. It received a larger dose than the gray control animal.
Nevertheless, the convulsions did not break out until twenty-eight
minutes after the injection, while the control had convulsions thir-
teen minutes after the injection. In this experiment the insufflation,
however, could not put off asphyxia longer than twenty-five minutes,
and with the onset of asphyxia the convulsions broke out.

Lexperiment XII a.— White rabbit, 2600 grams. ‘Tracheotomy.
, S ¥

4-51 P.M. Injected strychnine nitrate, 0.53 mgm. per kilo.

4-55. Trachea connected with bellows, etc.

5.11. Some spasmodic twitching (beginning dyspnoea?). Increased
the number and energy of the ventilations, animal quiet again.

5-15. Both vagi cut, ‘* head dyspncea ” sets in.

5-25. Voluntary breathing of the animal appears and gradually in-
creases.

5.28. Insufflation of hydrogen stopped, no apnoea, very labored dysp-
neeic breathing.

5.42. Trachea clamped, death. No strychnine effect at any time.

Lixperiment XIf 6. — Control, white rabbit, 1970 grams.
4.29. Injected strychnine nitrate, 0.45 mgm. per kilo.
5-01. Convulsions, died in two minutes.

In Experiment XII a, the animal received more strychnine than the
control, which succumbed thirty-four minutes after injection, but had
no convulsions for forty-seven minutes; z.e., during the time it was
under observation. The slight twitchings which appeared sixteen
minutes after injection were promptly suppressed by the increase in
ventilation with hydrogen. The inhibitory effect upon respiration,
however, was greatly diminished by the section of the vagi.. The

Artificial Respiration and Strychnine Spasms. 21

concomitant breathing set in immediately, and the voluntary breath-
ing appeared soon also, and apparently would have terminated in
asphyxia, if the hydrogen insufflation had not been discontinued.

Experiment XIII a. —White rabbit, 1660 grams. Tracheotomy.
4.48 p.M. Injected strychnine nitrate, 0.54 mgm. per kilo.
4.50. Trachea connected with bellows, etc. At no time voluntary or
concomitant breathing, no sign of hyperzesthesia.
5.21. Insufflation of hydrogen stopped, brief apnoea, then normal
breathing. Observed till 5.42, no convulsions.
Lixperiment XIII b. — Control, white rabbit, 1420 grams.
5:27. Injected strychnine nitrate, o.4g mgm. per kilo.
5-53- Had convulsions, and died in about two minutes.

The control had a fatal tetanus in twenty-six minutes, while animal
XIiILa, with a larger dose, showed no strychnine effect for the fifty-
five minutes it was kept under observation. The insufflation lasted
for thirty-one minutes and exerted a distinct inhibitory effect upon
the respiration.

The results we obtained in these experiments were extraordinary
indeed. Not only could the effects of fatal doses of strychnine be
completely prevented by insufflation of pure hydrogen, but the animal
could be kept by such an uninterrupted insufflation, as was seen in
Experiment XII] a, for thirty-one minutes without manifesting any
signs of asphyxia, dyspnoea, or cyanosis.

We all know very well that spontaneous inhalation of hydrogen
alone results in asphyxia almost immediately. This is an old, well-
established fact, and we have tested it ourselves by the following
direct experiments. When the trachea of an animal was connected
directly with the gasometer, without the intervention of the bel-
lows, the animal thus surely inhaling, spontaneously, pure hydrogen,
asphyxia set in after thirty to forty-five seconds, and rarely as late
as after sixty seconds. Apparently, then, it was the action of the
bellows which deferred asphyxia so long.

The first thought which occurs is that the bellows were, after all,
not perfectly air tight. We have tested them by letting the animal
spontaneously inhale the hydrogen from the gasometer through the
expanded bellows without ventilating them. The asphyxia was then,
indeed, deferred a little longer than when the inhalation occurred
without the intervening bellows. However, the gain was at the
utmost a minute or two, and therefore the amount of air which
could have found access to the bellows must have been at most

22 Witham J. Gies and S. J. Meltzer.

very small. But even granting that during the sudden and forcible
expansion of the bellows more air was sucked into them than during
the voluntary breathing, the amount of air which was able to pene-
trate the pores must under all circumstances necessarily have been
very small in proportion to the quantity of hydrogen which, under
constant pressure, had free access through the open lumen of a wide
tube. It must also be remembered that the animal not only had no
asphyxia under these conditions, but also that it was constantly in
a state of apnoea, —a state which occurs only, it is assumed, when
the animal receives more air than normally.

We may add, also, that, according to Osterwald (17), a diminution
of oxygen favors the outbreak of strychnine spasms. In our experi-
ments, with surely diminished oxygen there was no sign of convul-
sions even with fatal doses of strychnine.

These experiments brought us more than we looked for. It was
now no longer a simple question whether the mechanical factor of
artificial respiration has a share in the inhibition of strychnine
spasms. The question which confronted us was whether one of the
fundamental and apparently definitely settled principles in the theory
of respiration did not require revision.

Searching through earlier literature on the subject of respiration
we discovered that we had touched upon a long-forgotten chapter in
the discussion whether the absence of oxygen or the presence of
carbon dioxide is the cause of inspiration.

In 1862 L. Traube (18) made experiments with insufflation of hydro-
gen on dogs, in the same manner as we have made them on rabbits,
and found, as we did, that artificial respiration with pure hydrogen
may be carried on for a long period (forty-six minutes in one experi-
ment), the animal remaining all the while in a state of apnoea. On
the other hand the addition of carbon dioxide to the air rapidly
caused dyspnoea. Traube, in consequence of these observations, gave
up his original idea, that the absence of oxygen is the stimulus for
inspiration, and accepted the view that the real cause of respiration
is the presence of carbon dioxide. Heidenhain and Krause (19) soon
contradicted Traube’s statement, and explained his conclusion by
assuming that his bellows were not air tight.

Traube (20) repeated his experiments, oiled his bellows, and took
all precautions, as he states, to prevent the entrance of air, and in-
sisted on the correctness of his former results, attributing the failure
of Heidenhain and Krause to some fault in their technique.

Artificial Respiration and Strychnine Spasms. 23

Traube was contradicted also by Thiry (21),! and finally by
I. Rosenthal (23). Rosenthal did not repeat Traube’s experiments,
but connected the trachea of the animal with a gasometer of special
construction containing pure hydrogen, and found that the animals
became rapidly asphyxiated. By special calculations Rosenthal
arrived at the conclusion that air which contains only 1 per cent of
oxygen is sufficient for the maintenance of the animal, an amount
which presumably found its way into the bellows in Traube’s experi-
ment. That was the last word, at least the last we found recorded
in this discussion.

We may add that Traube’s technique suffers from still another
objection. In his experiments the opening for expiration had no
valve. The animal, therefore, could obtain, during inspiration, suffi-
cient air through this opening, even if it were made very small. As
long as it was large enough for expiration it was also sufficient for
inspiration. We have established this fact by experiment. The
trachea was connected directly with the gasometer while the expi-
ratory tube was submerged: asphyxia in forty-five seconds. The
expiratory tube was left free in the air, and the stop-cock turned
so as to make the lumen permissibly narrow: the animal went on
breathing without noteworthy impediment for some time.

Rosenthal’s paper appeared in 1864, and at that time there had
not yet arisen the question whether the mechanical distention of
the Jungs can cause inhibition of inspiration. The only question in
the minds of the earlier investigators was whether absence of oxygen
or presence of carbon dioxide is the stimulus of respiration. And as
the simple inhalation of hydrogen caused asphyxia, this appeared
to prove that it is the absence of oxygen which causes respiration.
Traube’s experiments, therefore, seemed to have no further object.
The value of the mechanical element which distinguishes artificial
respiration from spontaneous breathing had not yet been recognized.
We now know, from the studies of Hering and Breuer, Head, Gad,
Meltzer, and others, that the mechanical effect of the distention of
the lungs has a distinct inhibitory influence upon respiration.

It is now, furthermore, the general consensus of opinion that both

1 That, at least, is what THIRY states in his paper in the Zeitschrift fiir ration-
elle Medizin (iii), xxi, p. 25. It is not stated on what grounds the opinion is based.
MIESCHER-RUSCH (22), however, quotes THIRY from a French paper as saying
that artificial respiration with air and hydrogen causes apnoea. This paper was
not accessible’ to us.

24 \ William J. Gries and S. J. Meltzer.

the presence of carbon dioxide, as well as the absence of oxygen, act
as stimuli to the respiratory mechanism. But it is surely not the
actual immediate need of oxygen for metabolic purposes which in
the latter case is the stimulus. The blood and lymph and tissues
are provided with a surplus of oxygen for actual oxidative necessities.
It is the first intimation of a deficit in this sinking fund which acts
as a warning signal,—as a stimulus for increased provision of oxygen,
increased inspiration. Is it, then, inadmissible to assume that this
warning, this stimulus resulting from diminution in the body’s in-
come of oxygen, could be overcome for some time by the inhibitory
influence of the rhythmical mechanical effect of distention of the
lungs, if sufficient provision were made for the full escape of the
carbon dioxide? Our experiments do not, of course, warrant such a
positive conclusion. The bellows permitted the entrance of air to
some degree, but the amount of air which entered was surely com-
paratively smail. If, therefore, our experiments, as well as those of
Traube, do not yet permit positive conclusions in this regard, they
are at least suggestive enough to urge the necessity of a reinvestiga-
tion of this particular question with more favorable methods. In
this relation the necessity of avoiding suction apparatus in the execu-
tion of artificial respiration with indifferent gases seems important.

In this connection, also, we wish to call attention to the statement
of Head (24) that he caused apncea by insufflation of hydrogen. His
conclusion was that the apnoea was due to mechanical effects. He
used bellows, and does not mention any precaution taken.to guard
against the entrance of air into the bellows. Could not the conten-
tion be made against his conclusions also, as it was raised against
Traube’s, that it was the air which entered through the pores of the
leather into the bellows that brought about the observed result?

Regarding the arrest of the strychnine spasm, which we observed,
with hydrogen insufflations, it appears very probable that it is due
largely to the mechanical effect of the insufflation, and that it is not
essentially a result of the admixture of small amounts of air. Here
also additional experiments, and by other methods, will have to be
made before the question can be definitely settled.

BIBLIOGRAPHY.

I. MEISSNER und RICHTER: Géttinger Gelehrten Anzeiger, 1862, p. 165;
Zeitschrift fiir rationelle Medizin (3), xviii, p. 76.
2. LrusBe: Archiv fiir Anatomie und Physiologie, 1867, p. 629.

wb

Artificial Respiration and Strychnine Spasms. 25

Usprensky: Archiv fiir Anatomie und Fhysiologie, 1868, p. 522.

ROSENTHAL: Comptes rendus, 1867, !xiv, p. 1142.

ScHIFF: Beitrage zur Physiologie, 1°96, iii, p. 211; reprinted from I’Imparziale,
1867.

RICHET: Comptes rendus, 1880, xci, p. 131.

RossBacH und JOCHELSOHN: Centralblatt fiir die medicinische Wissen-
schaften, 1873, No. 24, p. 369.

ExBNER: Inaugural dissertation, Giessen, 1870.

BucHHEIM: Archiv fiir experimentelle Pathologie, 1875, iv, p. 137.

PAUSCHINGER: Archiv fiir Physiologie, 1878, p. 401.

BROWN-SEQUARD: Archives de physiologie, 1872.

FILEHNE: Archiv fiir Physiologie, 1873, p. 361.

ANANOFF: Centralblatt fiir die medicinische Wissenschaften, 1874, No. 27,
Pp 417.

OsTERWALD: Archiv fiir experimentelle Pathologie, 1899, xliv, p. 451.

V. CzyHLARZ: Zeitschrift fiir Heilkunde, 1901, p. 160.

GRAHAM: Archiv fiir die gesammte Physiologie, 1881, xxv.

OSTERWALD: Loc. cit.

TRAUBE: Allgemeine medicinische Central-Zeitzung, 1862, p. 296.

KRAUSE: Studien aus dem physiologischem Laboratorium zu Breslau, 1863,
i, p- 31:

TRAUBE: Allgemeine medicinische Central-Zeitung, 1863, p. 776.

Tuiry: Zeitschrift fiir rationelle Medizin (3), xxi, p. 25.

MiescHEer-Rtscu: Archiv fiir Physiologie, 1885, p. 355.

ROSENTHAL: Archiv fiir Physiologie, 1864, p. 456.

HEAD: Journal of physiology, 1889, x, p. 40.

CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE
MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE.
E. L. MARK, Direcror. No. 140.

ON THE NEGATIVE AND POSITIVE PHOTO?R@re®
OF THE EARTHWORM ALLOLOBOPHORA PGi
(SAV.) AS DETERMINED BY LIGHT OF DIFFERER

INPENSITIES:
By GEORGE, P. ADAMS.

| Bios earthworms are negatively phototropic, that is, creep away

from a source of light, has been frequently demonstrated, and
the extent to which this is true has already been roughly estimated
for Allolobophora foetida. Parker and Arkin (:01, p. 154) found
that when this worm was illuminated on the whole of one side by a
light of 3.3 to 4.0 candle-metres intensity, 30.2 per cent of the head
movements were away from the light. Of these, however, 4.2 per
cent were due to other deflecting causes than light, and hence 26 per
cent may be taken as a measure of the deflecting effect of the light.
Miss Smith (: 02, p. 469) carried out similar experiments, but by
slightly different methods, and found that 61 per cent of the reactions
were in a generally negative direction. If Miss Smith’s results are
interpreted in the way in which Parker and Arkin’s were, it would
appear that 61 per cent less 22 per cent, or 39 per cent, is the propor-
tion of reactions to be attributed to the deflecting effect of the light.
Thus, there is a disagreement in the two sets of records amounting to
the difference between 39 per cent and 26 percent. This disagree-
ment may be due to differences in the intensities of the light used in
the two sets of experiments. Miss Smith carried on her experiments
in diffuse daylight, but as she gives no data from which the intensity
can be calculated, this assumption may or may not be correct.

The present investigation was made under the direction of Dr. G.
H. Parker, to ascertain the relation between the intensity of the light
to which the earthworm Allolobophora foetida (Sav.) was exposed,
and its negative phototropic reactions ; and to find out under what
conditions, if any, the reactions would become positive. Inversions

in the sense of the -phototropism have already been recorded for a
20

Negative and Positive Phototropism of the Earthworm. 27

number of animals. Thus, Wilson (91, p. 414) has shown that
Hydra is negative to bright light, and positive to dim light; and
the same has been demonstrated for Polygordius larvae by Loeb
(93, p. 89), for Limax by Frandsen (: 01, p. 206), and for the
females of the Copepod Labidocera estiva by Parker (: 02, p. 114).
In these instances negative phototropism is associated with an
increased intensity of light, and this seems to be a general rule,
though there is at least one exception to it, namely, in Orchestia agilis.
This crustacean, as reported by Holmes (: O1, p. 216), becomes posi-
tive in bright light, but is subsequently negative when transferred to
diffuse daylight.

In testing Allolobophora foetida, twelve different intensities of
light were used, ranging from 192 candle-metres to 0.001 candle-
metre. The first set of experiments was made with one piece of
apparatus, using eleven different intensities. The second set was
carried on with another apparatus, using the lowest intensity of light,
0.001 candle-metre. A third set of experiments, devised to check
any disagreement between the results of the other two sets, due to
a difference in the apparatus, was carried on with both pieces of
apparatus, using the same light intensity for each.

The first apparatus (Fig. 1), which may be called the high-intensity
apparatus, was essentially the same as that used by Parker and Arkin
(: OI, pp. 151-152). It consisted of a dark chamber (C) illuminated
only through a horizontal oblong opening by a distant incandescent
electric lamp (f). A glass jar (A), with parallel sides 6.25 cm.
apart, and containing water, was placed immediately in front of the
lamp to cut off heat. Within the dark chamber, directly in the path
of the beam of light coming through the oblong opening, was a glass
plate (G), suspended so that it was horizontal and could be easily
rotated about its vertical axis. The glass was covered with wet filter
paper, which was frequently changed. The light intensity is given
for each experiment in candle-metres, and is the intensity of the
beam of light at the centre of the plate. In all cases, the candle-
power of the incandescent lamp was determined by a Lummer-
Brodhun photometer.

In the first set of experiments, for each of the eleven light
intensities, twenty fresh worms were used. Each worm was headed
across the filter paper on the glass plate in a direction at right angles
to the rays of light. As the worm crept, the plate was rotated, so
that the axis of the worm was kept constantly at right angles to the

28 George P. Adams.

direction of the light. Ten readings were taken on each worm; five
with the light falling on the worm’s right side, and five with the light
on the left side. Thus, for each intensity, there were two hundred
readings. As is well known, the worm creeps by projecting the head
forward, and then drawing the, body toward the head. When the
worm projected the head in a line with the axis of the posterior part
of the body, the reaction was called indifferent (0); when the head
was moved to one side, if it was toward the light, the reaction was
called positive (+); if away from the light, negative (—). For each
of the eleven intensities, there was a total excess of negative over
positive reactions. By subtracting the positive from the negative

FicurEe 1.— Ground plan of high-intensity
apparatus. 4, glass vessel containing
water to intercept heat rays; B, incan-
descent electric lamp; C, dark-chamber;
G, glass plate covered with wet filter
paper upon which the worms were made
to creep; Z, lantern; /, light-proof pas-*
sage used by Parker and Arkin, but
omitted in these experiments because
the apparatus was set up in a dark-room
with blackened wall.

reactions, the number of negative reactions due to the directive influ-
ence of the light was obtained; for, since the positive reactions are
not caused by the light, but by other influences, such as irregularities
of the surface over which the worm is creeping, etc., it is clear that
an equal number of the negative reactions are likely to be caused by
these other influences, and it is, therefore, necessary to subtract the
number of these from the total number of negative reactions, to get
the real number of such reactions due to light only.

Table I gives the result of the first set of experiments. The light
intensities in candle-metres are given in the first column on the left.
In the next two groups of columns are given, for the right and left
sides respectively, the number of times the worm moved indifferently
(0), negatively (—), and positively (+), and the excess of negative over
positive reactions. In the third group, the total reactions of the
worm for both sides are given.

From an inspection of the third group of the table, it appears that
the maximum number of reactions to be attributed to the directive

Negative and Positive Phototropism of the Earthworm. 29

influence of the light is found, not in the most intense light, but in
the third intensity used, 48 candle-metres. From this intensity there
is a rather constant decrease in the amount of this excess, corre-
sponding with a decrease of light intensity. The lowest intensity used
with this apparatus, 0.012 candle-metre, shows an excess of only 6
negative reactions out of a total of 200. There is also a less regular
increase in the number of positive reactions, as the light becomes
less intense.
SRABIER, 5
REACTION OF ALLOLOBOPHORA TO LIGHT OF DIFFERENT INTENSITIES.

Reactions with light on || Total reactions of

Light | Reactions with light on
left side of worm. worm.

intensity right side of worm.

in |
candle- |
metres. |

Excess of || | Excess of |
—over-+.|| — over -+. ||}

The record for negative light reactions (26 per cent) for the single
intensity of about 3.6 candle-metres obtained by Parker and Arkin
finds an appropriate place in this table, in that, falling between
5 candle-metres and 1 candle-metre, its negative light reaction (26
per cent) also falls between those of 5 (38.5 per cent) and of I
(24.5 per cent), though the position is not one that yields smooth
results. The fact that the number of negative light reactions varies
with the intensity makes it highly probable that the difference be-
tween the records of Parker and Arkin, and those of Smith, are to be
explained as due to difference of intensity. Judged from the series

30 George P. Adamis.

exhibited in Table I, the intensity of light used by Smith was not far
from 5 candle-metres, a condition easily obtainable from diffuse daylight.

The decrease in the excess of negative over positive reactions
corresponding with a decrease of light intensities suggested that if
the light were still further decreased, the negative light reactions
might entirely disappear, or even be replaced by positive ones. To
test this, the first piece of apparatus could not be used, because at
intensities of light lower than 0.01 candle-metre it was found impos-
sible to follow with the eye the movements of the worm. To obviate
this difficulty, the second piece of apparatus was devised. This
apparatus (Fig. 2), which may be called the low-intensity apparatus,
consisted of two dark chambers in contact with each other, shown in
plan in the figure. In the first chamber (4), a vertical white screen
(CD) was placed in a diagonal position,
so that it reflected light from an elec-
tric incandescent lamp (£) through a
small aperture (/) between the two
chambers. This reflected light entered
the second chamber (/), and there
passed over a horizontal glass plate
(G). It was found that when the rela-
tions of the light, screen, and plate were such as are given in the dia-
gram, with an incandescent lamp of 8.25 candle-power, and a circular
aperture of 1 mm. diameter, an illumination of 0.0011 candle-metre
was obtained at the centre of the plate. Wet filter paper was used
on the plate, and two thick, narrow strips of glass (H//), like those
employed by Smith (:02), were placed on the filter paper in parallel
positions, with only enough space between them to allow the worm
to creep. This narrow path between the two strips was in a direc-
tion at right angles to the central rays of light which came through
the aperture. After starting the worm in this path, the chambers
were closed so as to exclude all outside light. Time enough was
given to allow the worm to creep into the field beyond the two strips
of glass; chamber B was then opened, and the position of the worm
noted. If the worm was headed toward the light, it was called a
positive reaction; if away from the light, a negative one; and if the
worm had moved without changing its original direction, an indif-
ferent one.

Twenty fresh worms were used as before, and for each worm two
sets of readings were made. In one set, the worm was illuminated

FIGURE 2.

Negative and Positive Phototropism of the Earthworm.

31

by the beam of light of 0.0011 candle-metre intensity; in the other
set, the aperture was closed, that is, the worm was in total darkness.

TABLE II.
REACTIONS OF ALLOLOBOPHORA TO LIGHT OF 0.0011 CANDLE-METRE, AND IN
DARKNESS.
Rhee ie fone aaa Darkness.
of = :
| es aaa ee eee

| | | ;
1 Se ee a ee slic a eerepee fe

Pe yith°0 | 8 2 6 0 5 5 0
3 | ae (ele seal Sat 2 0 6 + 2
4 || o 6, |. 4 27 ico 7 3 44
Sve it 0 6 4 208 Nai 7 3 44

6 | 0 5 5 0 0 5 5 0
lene ‘ca ae: 3 4 a 5 3 2+
8 0 5 5 0 1 5 4 1+
Ooi 5 | l 0 4 6 2-
er 8 2 6 0 4 6 _
LM 8 2 6 || 0 6 4 2+
12 1 0 6 4 ae fea 7 3 44
13 0 8 | 2 6 1 4 5 —
14 || 0 8 2 6 0 4 6 ce
lig 71)" | 6} 4 2 1 5 4 i
16 | i ee | 1 7 0 6 | 4 2+
Ll aie ame ae 6 1 te i
18 | o | fem Fae 4 0 3 ji 4—
19 0 Oh ld 8 0 a ae vas

20 1 a a tes 0 hl eee 0

| | : |
| Totals | ey has 59 | 79+ 7 | 10 Sar inate
ll |

In each set, ten readings were made on each worm, five with the
worm’s right side toward the aperture, which in one set was open and

in the other closed, and five with the left side toward the aperture.

32 George P. Adams.

Table II gives the reactions of the twenty worms. The readings
of the right and left sides of the worm have been combined. In the
first column on the left is given the number of the worm. In the
left-hand group of columns are given the reactions of the worm in
light of 0.oo1r candle-metre; and in the right-hand group are given
the reactions in darkness. At the bottom, the total reactions for the
twenty worms are recorded (see Table II, page 31).

It appears, first, that there is a great reduction in the number of
indifferent movements over the number obtained with the first piece
of apparatus. This is explained by the fact that in the second appa-
ratus the slightest deviation of the worm from the indifferent path,
which was marked by a pencil line on the filter paper, could be
observed; and thus many reactions which would have been recorded
as indifferent in the first apparatus were here shown to be either
positive or negative. Of the reactions of the worms in total darkness,
in all 200, 7 were indifferent, 100 were toward the aperture (which
was now closed), and 93 were away from it. The 200 reactions of
the same worms in light of 0.0011 candle-metre at the plate, were as
follows: 3 were indifferent, 138 were toward the light, and 59 were
away from the light, there being an excess of 79 positive reactions.
Moreover, there was never found an excess of negative over positive
movements in the case of any worm. This clearly points to the
conclusion that light of low intensity induces positive phototropic
reactions.

To ascertain whether the results obtained by the high-intensity
and by the low-intensity apparatus were fairly comparable, a third set
of experiments was tried, in which worms were tested at the same
ligkt intensity (1 candle-metre) in both pieces of apparatus. Twenty
fresh worms were used, each one first in one apparatus, and then in
the other. In each apparatus, ten readings were made, five with the
light on the right side, and five with it on the left side of the worm.

Table III shows the result of this set of experiments. The figures
given are the totals for the twenty worms, right and left sides
combined (see Table III, page 33).

A much larger number of indifferent movements were obtained on
the high-intensity apparatus, but the total excess of negative over
positive movements is so nearly alike for each (25 per cent and
29 per cent), that one may assume that the records of the two sets of
experiments are fairly comparable, and therefore, from the observa-
tions recorded in Tables I and II, draw the conclusion that Allolo-

Negative and Positive Phototropism of the Earthworm. 33

bophora fcetida is negative to light intensities between 192 and 0.012
candle-metres, and positive at 0.0011 candle-metre.

The results of these experiments are in harmony with certain daily
habits of earthworms. During the daytime, while the light is of

TAB ELL

Excess of

Direction of reactions.
€ — over +.

High-intensity apparatus

Low-intensity apparatus

relatively high intensity, these animals retreat into their burrows
because of negative phototropism. At night-time, however, they
emerge from their burrows, not because of the absence of light, but
by reason of their positive phototropism to light of low intensity, for
even in the darkest night there is faint light.

CONCLUSIONS.

1. Allolobophora fcetida is xegatively phototropic toward the light
from electric incandescent lamps varying in intensity from 192 candle-
metres to 0.012 candle-metre; the percentage of negative head-move-
ments referable to lights of different intensities are as follows: 41.5
Mervecent C192 cm:.), 41.5 per cent (90 cm.); 59 per cent (48 cm:),
maper cent (31 cm-), 45.5 per cent (12 cm.), 38.5 per cent (5 em.),
24.5 per cent (I cm.), 14 per cent (0.128 cm.), 12 per cent (0.050 cm.),
5 per cent (0.020 cm.), and 3 per cent (0.012 cm.).

2. A. foetida is fosztively phototropic toward an electric incan-
descent lamp of 0.0011 candle-metre intensity.

3. Earthworms retreat into their burrows during daytime because
of their negative phototropism. They emerge at night not so much
because of darkness as because of their positive phototropism for
faint light.

BIBLIOGRAPHY.
FRANDSEN, P.
:ort. Studies on the reaction of Limax maximus to directive stimuli. Pro-
ceedings of the American Academy of Arts and Sciences, xxxvii,
pp. 185-227.

34 George P. Adams.

HouMESs, S. J.
:ol. Phototaxis in the Amphipods. American journal of physiology, v

pp. 211-234.

?

LOEB, J.
’93. Ueber kiinstliche Umwandlung positiv heliotropischer Thiere in negativ
heliotropische und umgekehrt. Archiv fiir die gesammte Physiologie,
liv, pp. 81-107.
PARKER, G. H.
:02. The reactions of Copepods to various stimuli, and the bearing of this
on daily depth migrations. Bulletin United States Fish Commission
for 1901, pp. 103-123.

PARKER, G. H., and ARKIN, L.
:o1. The directive influence of light on the earthworm Allolobophora
foetida (Sav.). American journal of physiology, iv, pp. 151-157.
SMITH, AMELIA C.
:02. The influence of temperature, odors, light, and contact on the movements
of the earthworm. American journal of physiology, vi, pp. 459-486.
WILSON, E. B.

b)

gt. The heliotropism of Hydra. American naturalist, xxv, pp. 413-433-

oN EXPERIMENTAL STUDY OF THE SUGAR CONTENT
PND EX) RAV ASCULAR COAGULATION © OF ©THE
BLOOD AFTER ADMINISTRATION OF ADRENALIN.

By CHARLES H. VOSBURGH anp A. N. RICHARDS.!

[Carried out under the auspices of the Rockefeller Institute for Medical Research at the
Laboratory of Physiological Chemistry, of Columbia University, at the
College of Physicians and Surgeons, New York.J

INTRODUCTION.

pe in 1902 the discovery was made by Herter and Richards?

that the injection of solutions of adrenalin chloride into the
peritoneal cavity of dogs was followed by an intense though transient
glycosuria. It was also found that the application of adrenalin solu-
tion directly to the surface of the pancreas produced a similar effect.
As a result of a number of experiments in this direction, the sugges-
tion was offered that this form of glycosuria was in reality of pan-
creatic origin.

In extending these observations, Herter and Wakeman? have
found that the power of adrenalin to produce glycosuria, when applied
to the pancreas, is not specific but is shared with various substances.
The number of such substances is comparatively large, and apparently
the only quality common to the series is a reducing activity. A
seeming exception to this rule was found in potassium cyanide.
When solutions of this substance were applied to the pancreas in
amounts far too small to produce general toxic symptoms, glycosuria
resulted. This substance, like hydrocyanic acid, while it has no
reducing power, exerts a specific action on the animal cells in pre-
venting them from absorbing oxygen.* It is natural to suppose that,
in the absence of the normal amount of oxygen in the cell, an excess

! Research scholar of the Rockefeller Institute.

2? HERTER and RICHARDS: The medical news, 1902, Ixxx, p. 20I.

3 HERTER and WAKEMAN: Virchow’s Archiv fiir pathologische Anatomie und
Physiologie und fiir klinische Medicin, 1902, clxix, p. 479; HERTER : The medical
news, 1902, ]xxx, p. 867.

4 GEPPERT: Zeitschrift fiir klinische Medicin, 1889, xv, p. 208; /zd., p. 307.

35

36 Charles H. Vosburgh and A. N. Richards.

of reducing substances may be formed. It is possible that these
substances may act in a manner comparable to those of the above-
mentioned series in bringing about the excretion of sugar. From
the facts brought out by their observations in this regard, Herter and
Wakeman are inclined to attribute the production of glycosuria upon
the application of adrenalin and other substances to the pancreas
to a toxic action on the cells of that gland which is closely connected
with the power of reduction.

If this view of the matter is correct, an important relationship sug-
gests itself between this form of experimental glycosuria and condi-
tions in the human organism which may give rise to an excretion of
sugar. The fact that many organs of the body may form reducing
substances capable of easy oxidation which may reach the pancreas
in the blood stream, carries with it the possibility that, if the normal
balance between the amount of these substances and the oxidizing
power of the pancreas be disturbed, the production of glycosuria may
occur.

Concerning the mechanism through which adrenalin brings about
the excretion of sugar no positive statements can as yet be made.
The work of Minkowski! and his followers, which has furnished the
basis of the belief in the existence of an internal secretion of the pan-
creas exercising a controlling influence on carbohydrate metabolism,
justifies an assumption that the sugar elimination is the result of an
alteration in the nature, activity, or amount of this secretion. The
glycosuria-producing effect of injury of certain parts of the central
nervous system,” and the increase in sugar formation in the liver
which follows stimulation of the coeliac plexus? or of the vagus nerve,‘
may lead to the supposition that adrenalin glycosuria results from
the action of a nervous mechanism. Finally, it is known that under
the abnormal conditions which follow the injection of phlorhizin® or
chromic acid,® glycosuria may occur, owing to an increase in the
permeability of the kidney cells. The possibility that adrenalin

' MINKOWSKI: Untersuchungen iiber den Diabetes Mellitus nach Extirpation

des Pancreas, Leipzig, 1893.
* CL. BERNARD: Lecons sur la physiologie et la pathologie du system nerveux,
Paris, 1858, i, p. 401.
* A. and E. CAVAZZANI : Centralblatt fiir Physiologie, 1894, viii, p. 33.
* LEVENE: Centralblatt fiir Physiologie, 1894, viii, p. 337.
’ v, MERING: Zeitschrift fiir klinische Medicin, 1889, xvi, p. 431.
» Kossa: Archiv fiir die gesammte Physiologie, 1901-1902, Ixxxviii, p. 627.

Sugar Content and Coagulation of the Blood. ag

glycosuria is the immediate result of changes in the kidney has not
yet been excluded.

Whatever may be the manner by which the effects of. adrenalin
are brought about, it is probable that the mechanism involved is one
which is active, though in a different degree, under normal condi-
tions. A determination of the identity of the mechanism is therefore
of importance, not only in explaining the phenomenon in question,
but also from the fact that it may throw light on some of the
processes connected with the normal metabolism of carbohydrate
within the organism.

Before such a determination can be made, however, a more accu-
rate knowledge of the internal conditions antecedent to the excretion
of sugar is necessary. With this end in view we have made a some-
what detailed study of the sugar in the blood, after intraperitoneal
injection of adrenalin, as well as after application of that substance
to the pancreas.

SuGAR CONTENT AND COAGULATION OF ARTERIAL BLOOD
AFTER TREATMENT WITH ADRENALIN.

It has long been known that the glycosuria produced by extirpa-
tion of the pancreas,! puncture of the floor of the fourth ventricle,”
and poisoning with certain substances, such as carbon monoxide,’
is the immediate result of an increased accumulation of sugar in the
blood. On the other hand, injections of phlorhizin* are followed by
the excretion of sugar due to the effect on the kidney. In the latter
case the percentage of sugar in the blood never rises above normal,
and may even fall below that amount. To determine in which
class adrenalin glycosuria belongs, we have made a number of deter-
minations of the sugar content of the blood of dogs which had been
subjected to treatment with adrenalin. In this series also we have
attempted to ascertain the rapidity with which this substance acts,
and the course and duration of its influence.

Method of collection and analysis of blood. — Healthy, well-nourished
dogs were anzesthetized with pure ether, a cannula introduced into a
femoral artery, and a portion of blood taken. The solution of adrena-

1 MINKOWSKI: Loc. cit.

2 CL. BERNARD: Loe. cit.

8 SENFF: Ueber den Diabetes nach der Kohlenoxydathmung, Inaugural dis-
sertion, Dorpat, 1860.

4 v. MERING: Loe. cit.

38 Charles F1. Vosburgh and A. N. Richards.

lin chloride! was then injected by a hypodermic syringe into the
peritoneal cavity or, after an incision through the abdominal wall, was
painted on the surface of the pancreas with a soft brush. Portions
of blood were then drawn from the femoral artery at various
intervals.

Having in mind the possible production of glycosuria by means of
anzesthetics,? as well as by asphyxia,* care was taken to keep the
anesthesia as light and as constant as possible. Moreover, we
believe that this factor may be left out of account in these experi-
ments, since the control portion of blood, taken before adrenalin
treatment, was collected under the same conditions of anzesthesia
as the subsequent portions which are compared with it.

The portions of blood were analyzed according to the following
procedure. The blood was drawn directly into a beaker containing
a solution of phosphotungstic acid in dilute hydrochloric acid. The
beaker was counterpoised on a balance and the blood weighed imme-
diately after its withdrawal from the artery. On boiling this mixture
the blood proteids are precipitated in a granular form leaving a
water-clear fluid free from proteid. The precipitate was washed
thoroughly with hot water, a process which is rendered easy by its
porosity and its friable character. The combined filtrate and wash-
ings were nearly neutralized with sodium hydroxide and evaporated
_to small bulk on the water bath. The evaporated residue was made
up to known volume (50-100 c.c.) with water, and was filtered. The
reducing power of this solution was determined by the Allihn method.
The results were calculated in terms of dextrose from the weight
of the metallic copper. The figures given represent the averages of
closely agreeing duplicates.

Method of determining coagulation.—JIn one of our early experi-
ments we noticed that a portion of blood drawn for the purpose of
rinsing the cannula clotted very rapidly. As a result of this observa-
tion, in a number of later experiments we have taken separate por-

* In all the experiments outlined in this paper, the adrenalin chloride solution
(;0'00) Prepared by Parke, Davis, & Co., by the method of Takamine, was used.

* CuSHNY: Pharmacology and Therapeutics or the action of drugs, 1899,
p. 160.

* DASTRE: Comptes rendus des séances de l’academie des sciences, 1879,
Ixxxix, p. 669.

* This solution contained 70 gms. of phosphotungstic acid and 20 c.c. of hydro-

chloric acid, sp. gr. 1.20, in a litre. About 5 c.c. are sufficient to completely pre-
cipitate the proteids in 1. gm. of blood.

Sugar Content and Coagulation of the Blood. 39

ao

tions of blood to be tested regarding this point. The amount drawn
in each case was 2 c.c., collected in a graduated cylinder of 5 c.c.
capacity. The time which elapsed between the collection of the
blood and the time at which the cylinder could be inverted without
loss of its contents, was noted as the time of the coagulation of the
blood.

The results of our determinations are given in Table I, pages 40, 41.

These experiments show unmistakably that the administration of
adrenalin chloride either by intraperitoneal injection or by painting
it upon the pancreas is followed by a marked increase of sugar in the
blood. The increase is very noticeable within the first five minutes
after the application and reaches its maximum within three hours.
A very gradual fall then ensues, which may continue until the per-
centage of sugar becomes subnormal (Exp. 2). In a dog recently
fed (Exp. 1), the blood sugar may be double the normal quantity
fourteen hours after the injection. A marked rise occurred in the
case of a dog (Exp. 7) which had been starved for six days. In Ex-
periment 8 a fatal dose of adrenalin was given. A slight increase in
the sugar content of the blood occurred shortly after. One minute
before death ensued, twenty-four hours after injection, the percentage
of sugar was approximately normal.

Simultaneously with the production of hyperglycemia, an effect on
the coagulability of the blood is observed. In every case, without
exception, the time of coagulation is lessened after adrenalin is given.
This diminution is equal in some cases to four-fifths of the coagula-
tion time of the control.

Arthus! has shown that the time of coagulation decreases if the
blood is allowed to come in contact with blood already clotted or with
an exposed tissue surface. Special care has been taken therefore in
these experiments to remove the clot from the cannula before each
collection. Furthermore the portion for the coagulation test was
collected just after that for sugar analysis, a circumstance which
insures the rinsing of the cannula.

The recent observation? by the same author that the mere with-
drawal of large amounts of blood from the body hastens the coagula-
tion of subsequent portions, raises the question whether the results
which we have observed may have been due to loss of blood alone.
To test this point, a control experiment was made in which the

? ARTHUS: Journal de physiologie et de la pathologie générale, 1902, iv, p. 283.
2 ArRTHUS: /éid., p. 273-

Charles 1. Vosburgh and A. N. Richards.

40

9870 ” » ‘SIU ZG 2°W'd CTOL £1
1620 Suguted 1ayye ay T+" 'd 716 SZ
'W ‘d Z['S Je svaroued LE70 ” ” » OF :'W'd CHS +06
Jo aovyins wo poyured (521) uonntos L120 ” mi» SL 2d £38 TZ
apuoyyo ulpeuerpe ‘o°o f fyuewTed gee Suyured aayye ‘ur ¢ :'Wd JTS Cue
*xo JO pus [[) SPL, Worl peznoyyy €Z10 feultou +"W i'd OCs 61% ‘SI Sv TL +
S9T0 ” ” » € 3? W'd 0O'9 80
+1¢0 » » ‘SIGS = Wid 90'S Fé
+0¢'0 » » IMT + Wd 804 0°02
‘Wid 80'E Je Ysniq & YIM Ssveto S8T0 ” we ase) We rN Se L6l
-ued 0} paidde (°%°') uonnjos apr 8/10 ” te Sill PSE orate SCC
-O[Yyo UljeuaIpe Jo ‘9°90 ¢ SyuaUITIadxa ZSTO Suyured 10yje ur Y :"W'd STE Sa 4
jnoysno.1y} panuyuos uoneziuayyy ZIlO jewiuiou : Wd OOS b'CS "sy Qf OZI le
00 ¢ 660°0 ” ” » AL? WV 8e'01 ELS
‘Bulusow yxou “W'VOEOT| SO € LOE0 » on eae ee Wiese OIL SOL
ye ose “Wd OF OT Ie pezizayja ulese | OT + | OL ” » “UT i'd Ly'9 STI
S'W'd QOL [1G panuyuos visayjsxur | QZ Z Ler'0 i Smet apple Sho A NORE ATO YAKS) LOI
“Wd Lg ye Aqreouoqadenurt payaf | 00 Z | O0FO ” we J LOT SW a LG'S SIL
-ut (©°°T) uorjnfos apuoyyo uljeuaipe |} CT Z 1S¢O uonoofur 19qye suru ¢ ) s"W'd ZS'S OST |
JO 9°9 § f'W'd OES YW paezuoayriy 00 + Ssc0 [eULIou Wd Pps 062 "SIU CE | é
“O98 “Ulur
‘OF TT 3v U1ow SurMoy[oF uo pazusyya 1eo0 x pe er WG Cree
urese {WV OFZ] [2 penuuoo visayy ZS°0 n Sy ps SAMAR OS PAT O'+?
-sw@ue {IW ‘d $¢°6 Ye Alpvouozodesjut 950 a SSL ge Na OS TL O'S?
poyafur (29°F) uonjos apoyo ure ZO uolqoalul aye ay T sWd SEOL oog
UdsIpe dg SW 'd $16 Ie paziieyay 9T'0 [vWAoU : "Wd OSG 0°62 sly ZT v8 I
toy [7899424 SURI | -nay ase] cr al ieee
*SyALUIDY ~e[ndvoo arene ‘uaye} UdYM SUIT, | Soe ke aus | ‘Sop Jo | -Ladxe
jo OWL, ou, | I4sleM| JOON
|

“NITIVNAUGY HLIM LNANLVAYL YALAV GOOTY IVIYALUY AO NOILVINOVOD AGNV LNADINOD AVINS

eS aL eVare

| JO JUnOWULY

Sugar Content and Coagulation of the Blood.

*payoo][O9 SVM Pood Fo uo]
-10d ysv[ 1aqye ‘ur T palp ‘saoay ul
possed snoonwu Apoosq o14stte}oV1eYyo
£7 ou ye udAIS Iayya ou payeajsoad
yontu A19A [eUULUR fSUTULOUSULMO][OF
OF OLIV paztayie S Wea CT TL UY 49419
Japun yday Nd OF'EESH Je AT[VIUo}
-ade.nurt pozoolur (28°41) -os aptopyo
Ul[vUAIPe FO ‘O79 ZI faayjo pue wu0jJ
-OLOTYD FO "PX YILM pazyaysxuy

‘Wd ¢f'¢ Ww svosdued 0} pardde

UOIIN[OS Ul;LUdaIpe Jo "9.9 ¢ fjuduT
ssodx9 ynoysnory} uaars IYI,

‘Wd QO'S 3@
svasourd 0} pardde (°°°1) uonnios
aplo[yo ulpruatpe jo ‘oo Z fjuUaUT

-tiadxo =ynoysnory} UALS JOY

“A ‘V
CSOT 3 svosoued 07 pardde (22°7)
apluolyd ulpeusipe jo ‘oO ¢ fjUuaW
-l1adxa ay} JNOYSsno1Yy; UdAIS 19YI9]

O¢ S cOv0 ” ” » oO + Ned S£'6
Or 2 sso'0 Uotoalur 19yzV ‘sty ET 2 WV TS OT
cL. -e 6610 uorqoalur ye une ay [2d €O'TT
Gr Zz $600 uooolur rye ‘ulut of :"W'd ST'OT
OZ) 1 +00 uoloalut ‘yye sas Qe ull TT ‘Wd $f6
og + Selo euliou : "I ‘d £76
Set 6910 7 Jo) Se eo ORAS)
CCeaT C+L0 Sunuied 1ayje say Z td S'S
Giaez 9020 Sunuied yye ur ¢T ay Ts "W'd Sg}
SO @ 16T0 » ” » 6¢ * Wid 22+
Ob @ eZT0 » ” » Sl + Wid [0+
AN HA Z6LO Suyured rayze curt Qs “d GE'S
OOMee FST'O [euttou sW'd $o°"
Ob I ILO ” ” “say 4 : IN ‘d OO'L
Ly 1 STe'0 ” » AGT +W'd 009
0% i! +970 ” ” ” oe cua O¢'s
Or I L1c0 ” 15 eee ele eg
Cores $0z'0 Surured s9yye ur ¢ :*W'd Q'S
os 9 Igloo [eULtoU iW id ZO'+
OS 0 S90 ” ” ” + WN 'd oS %
SI il 6LS'0 ” ” ” ¢ : "IN ‘d tS T
ce I LLY'0 ” ” “SI 6. Nd PS CL
Alt fer'0 » ie ae Ube aoa Vora
10 @ 88e 0 » # on OF SOW Soil
LS. tS€'0 » oan ail Ne OL TE
00 ¢ 1620 Suyuied saye cut ¢ :W'v OO'TT
0c § 6270 [etustou FWY $F OT

‘oes *UTUr

S97
O6L
FOL
{1
OST
Ol

SSI
LLY
S91
OLT
csi
LT
OLT

OLT
est
Oot
'0¢
OLT
6 LT

00
£61
ole
£06
Fol
8'S¢
8&2
661

sXep 9

"SIU $7

*sIU $2

Let

ee

é fl

I~

42 Charles 1. Vosburgh and A. N. Richards.

amount of blood taken was very small. The details of the experi-
ment are as follows:

A small dog, 5.3 kilos in weight, which had received no food for twenty-eight
hours previous to the experiment, was etherized and a cannula introduced
into the femoral artery. 2 c.c. of blood collected at 2.34 P. M., coagulated
in 5 min. 50sec. An incision was made through the abdominal wall and
2 c.c. of adrenalin solution applied to the pancreas with a soft brush

at 2.43 Pp. M. Subsequent samples of blood (2 c.c. each) coagulated as
follows:

5 min. after adrenalin application (2.48 P.M.) coagulated in ] min. 45 secs.

eh er ot ‘ é (259 P.M.) ‘ Ze
33 “ec “ “ “ee (3.16 P. M.) “ 2, “cc 32 “ec
1 hr. us ss = (3.45 P. M.) 3s SO Ree

Other control experiments in which large amounts of blood were

removed, and no treatment with adrenalin given, show no such marked
changes as are seen in the observations in Table I.

THe SOURCE OF THE EXCESS OF SUGAR IN THE BLOOD IN ADREN-
ALIN GLYCOSURIA, AS INDICATED BY COMPARATIVE ANALYSIS
or BrLoop COLLECTED SIMULTANEOUSLY FROM THE PORTAL
AND HEPATIC VEINS AND THE FEMORAL ARTERY.

The results just detailed show clearly that the phenomenon of
adrenalin glycosuria is due to an increase of sugar in the blood.
The source’ of this excess of sugar is of great importance in deter-
mining the mechanism by which this effect is brought about. We
have endeavored to trace the source of the sugar by means of analysis
of blood taken simultaneously from the portal and hepatic veins and
from the femoral artery. In these experiments it was necessary to
collect successive portions of blood at various intervals from the same
blood-vessel without interfering with the normal circulation in that
vessel. The methods which have hitherto been devised for the col-
lection of portal and hepatic blood appeared to be inadequate for our
purpose, as well as somewhat difficult and uncertain of application.!
We have therefore made use of an original method, the description of
which follows. A cannula of special design? (see accompanying dia-
gram) is the essential feature of the method.

1 For description of older methods, see SEEGEN: Die Zuckerbildung im Thier-
korper, 2 Auflage, Berlin, 1g00, p. 66.

* The special cannulas used in these experiments were made for us very skil-
fully by Mr. John T. Hoyt of the Department of Physiology in this institution.

Sugar Content and Coagulation of the Blood. 43

General method. — The method of fixing the cannula into a vein is as follows:
The vessel is carefully exposed, and the outer connective tissue sheath dis-
sected away. Loose ligatures are passed about the vessel at each end of
the cleared portion, which should be about 2 cm. in length. Before intro-
ducing the cannula, its parts should be so arranged that the flange of the
outer tube is separated by a space of about o.5 cm. from that of the inner
tube. The brass rod should be in position, totally filling up the bore of
the inner tube. The ligatures about the vessel are then tightened momen-
tarily by an assistant, and a longitudinal slit made in the wall of the vessel
between them. This slit should be a trifle shorter than the diameter of
the flange. The flange of the inner tube is then introduced into the
interior of the vessel through the slit, the outer tube pressed down till the

EXPLANATION oF DIAGRAM.— The cannula consists of three parts, viz.: an inner
tube (4), an outer tube (2), and a nut (C). The inner tube is about 6.5 cm.
in length. Its outer diameter is 4 mm., its inner diameter,2 mm. On it a
screw thread (A’) is cut, extending for 3.5 cm. from one end. The other end
is provided with a flange (D) 9 mm. in diameter. The end view of this flange
is shown at D’. The outer tube (4),3.5 cm. long, fitting closely over the inner
tube, terminates in a flange (Z) similar to the one on the latter. By means of
the nut the two tubes may be held in such a position that their flanges are in
close contact. All parts of the instrument are made of brass. The cannula
is also provided with a brass rod (not shown in the diagram) about 28 cm.
long, of such a thickness that it fits closely into the bore of the inner tube.
A small shoulder brazed on the rod at the proper point insures the complete
occlusion of the cannula when desired.

4

wall of the vessel is held tightly between the two flanges, and the nut
screwed down so that the hold is retained. The ligatures are then loos-
ened and the normal circulation is resumed. With a little practice the
operation can be accomplished with no loss of blood and an interruption
of the circulation of only thirty to forty seconds. On connecting a rubber
tube with the inner tube of the cannula and removing the brass rod, blood
can be withdrawn at pleasure.

In our experiments to determine the sugar content of the blood flowing to and
from the liver, we have introduced cannulas of this type into the portal
vein at its juncture with the pancreatico-duodenalis, and into one of the
larger hepatic veins at a point between the liver and the diaphragm.

In order to expose the vessels, a transverse cut was made through the abdomi-
nal wall, following the curvature of the free border of the ribs and extend-
ing for about three inches on each side of a point on the median line just
below the xyphoid cartilage. Bleeding was prevented by ligaturing the
vessels which it was necessary to cut. The abdominal organs were pro-
tected from exposure by cloths moistened with warm saline solution. The
time necessary for the operation and the introduction of the cannulas into
both veins usually amounted to about one hour.

44 Charles LH. Vosburgh and A. N. Richards.

In drawing blood from the hepatic vein it was necessary to use a suction pump
to overcome the negative pressure, which is very manifest in the venous
circulation at this point. For this purpose, the beaker containing the
phosphotungstic mixture, previously weighed, was placed under a small
bell-jar which was connected by its lower opening with a suction pump.
A small glass tube inserted through the rubber stopper which closed the
upper opening of the bell-jar, and terminating at a point directly over
the beaker, served to conduct the blood into the precipitating fluid. On
connecting the cannula in the hepatic vein with the glass tube, and apply-
ing suction, the estimated amount of blood is easily obtained.

For collecting arterial blood, a glass cannula of the ordinary type was intro-
duced into the femoral artery.

The order of procedure in obtaining the portions of blood simultaneously was
as follows: The beakers which were to hold the femoral and portal blood
were counterpoised on balances placed at the side of the operating table.
Small glass tubes were clamped in a position to lead the blood into them.
The beaker for collection of hepatic blood was placed under the bell-jar
arranged as described above. ‘The brass rods were removed from the
cannulas in the portal and hepatic veins, and connection with the proper
glass tubes made, passage of blood being prevented by clamping the
rubber tubes. The connection between the cannula in the femoral and
the third glass tube was made. -At a given signal the clamps were re-
moved and the blood collected. The brass rods were immediately replaced
in the cannulas, and the portions of blood weighed. The time necessary
for collection of all three portions of blood has seldom amounted to more
than ten or fifteen seconds.

The analysis of the blood was carried out in a manner exactly similar to that
previously described.

By means of these methods we’ have succeeded in obtaining blood
simultaneously from the three vessels mentioned, both before and
after the application of adrenalin chloride to the surface of the pan-
creas. The results of sugar determination are given in the table on
page 45.

The results of Experiments 1, 2, 3, and 6 present, we believe, a true
picture of the events in this connection succeeding the application of
adrenalin to the pancreas. The samples of blood taken before treat-
ment with adrenalin agree fairly closely in their sugar content. On
the ground of these figures it cannot be said that the amount of
sugar in the blood issuing from the liver is greater than that of the
femoral artery or portal vein, After treatment with adrenalin, how-
ever, the relations are changed. In Experiment 3, four minutes after

45

Sugar Content and Coagulation of the Blood.

“UIdA OFeday ayy OFUL V[NUULD AY} JAISUL 07 JapsO UL ysayD dy} Vado 0} Aressaoau punoy 31 sem Aquo Juawttodxa sty} uy 7

*puosas Jo awit} ye ayourdss 197° C77: q 5 5 9 ’ ( ~ ,
pue paysasuoo zeus pure pajsasuo9 Lec’ £co'0 8020 eae SO occ 10'9 Tél rel O'oT
spainueg ‘Suolde{[09 jsiy ye aouvived : 9€"S
-dv ul [euliou saansazur pu sva1ourg 8020 | ISTO | 10¢0 | 20jJeq “Ulm OT 9S SUE SSSI Lv1 SG, MN ras)
*puosas Jo aug . ‘T: . ate : : ; ;
qv orjouvdd pure palsasuod auijsaqut [[Rus OFE'0 8s1'0 cre'0 SHE OS ULG ee ore Oc! Lit vel
pur pajsasuod svasourg *Suoloa{[[O9 }S.1y Jo sce
DUI} JV [RUILOU SaUT]Sa}UT puke svatoUR OG OM ISLCG0 1220 a1OFaq “ULU COT Ste Sel SOT ‘CL 9Z | SE'S ¢
Wath? |I : 2 d
*at0uRAD . : eter : 7 ‘ Titra "7 ;
puv pajysaSuoo sea aunysaiut [peurs at) ‘puo SreO | FIO s8e"0 AS TEE PLUSH ae 60°9 sh'st [0c eT
-oas dY1 yy [RWOW Aram Sauysaiut pure ors é : =
svasourd at) ‘poojq jo Suimesp ysay IV OLEO |\<28c 0% TREO. | atoyeq “arUr, eT 6S |) S8ie SSL Ossi BO: (oo Sie | Gar
*paysaSuod yoNut a1am suvs.i0 6020 | 200 8020 ” 3 99); OFS T6l rail Fel
asap) ‘uarjer 910m saiies yang pue pay | ZSZ'O | LST | $120 eo ee A | 90$ | SI | SHT| el
ayi uayA, ‘souvivedde ut [euliou atom Via C JOT: qoue ‘ur | : ‘Cc Y aa
saulsaq [[ews puv seatoured oy} ‘poolq t0z'0 LILO A810 ‘ ay | j bby cel FLT Jel
Jo suonaayion pudoss pue way joPMN IY | oe | cong | esto | oxozoq “wim 6T | ’ lice | ot | eet| ezt | oz | se | ¢
: sil Eee! | ; [ L C ‘L
anjq A1aA oun Kae : | : f
-sazul [jeuts pure paisasuos yonuw AOA SUM oFe'0 S+Z'0 ZEEO i Tt | 02'S aT +8 oeT
svaioued ayy ‘poorq yo Suimesap paryy 7 ; : de | : aN os :
‘orjouvdd qeroaltine decteavar “aaieedlios b8¢ 0 L610 LYe0 Jae LOD 6b c OL Vol atall
yeyMatuos svaioued ‘uoysa]]09 puodsas JW | 6¢°>
‘aouvavedde ut [euliou seuyseayut pu svaso +9T'O SOTO O9T'O d1OFoq “Ulu OQ] CO + SOL 6 +1 Leer +7 76 Gs
-uvd ‘poojq Jo woryoa[foo Jsuy Jo oy TW ‘ :
spafordurs | $20 | 800] STO | axe ura QZ | v6 | OO€ | #IT| +41
uoneitdsar [eioynse pure ; pauado yaya eee
800 Oro OTO | ex0F9q “UlU g If6 cle VIE SLI cL | 06 l
WO Jag | yO 19g | "JUAaD 19g ‘Wid | ‘Wid | ‘sw | ‘suy |] ‘swy | ‘SIA | OMT
‘ulaA | ‘utaa | *A19q18 ‘uonvordde OYE | oy, oe uae | IRA *Ar9}.18 pay dx
-sy1ewiay onedayzy] Teqog | [e1oway | Fo aut] YM | -usaipe -99|[09 anedarzyz | jeq10d | [e1owla,T SU] = jo
pareduros uo | Fo u0y 3 Sire hme ee ones
-oa]joo jo awry, | -eordde F : sf N
WO1F Poor ur IesNs euTT | mosy poojq joqunompy (PULL) =

jo aut],

"NITIVNAYGY HLIM INANLVAY], YWALAV GNV ANOAAA AVALUV IVUOWNT YY

46 Charles H. Vosburgh and A. N. Richards.

the application of the substance to the gland the sugar content of the
arterial blood rises 0.028 per cent, that of the portal blood remains
practically the same, while the increase in reducing power of the
blood emerging from the liver amounts to 0.065 per cent. The same
relation, though on a higher plane, is apparent twenty-six minutes
later. Sixty-six minutes after, the sugar percentage from all the ves-
sels is approximately the same. Precisely similar results are obtained
in Experiments 1 and 2, and in a lesser degree in Experiment 6.
Judging from the results of these analyses then, a formation of sugar
in the liver must be the cause, in part at least, of the increase of
sugar in the blood.

Experiments 4 and 5 apparently form exceptions to this conclusion.
It will be noticed, however, that the percentage of sugar in the sam-
ples taken before adrenalin treatment are abnormally high, especially
in Experiment 4. It is possible that the mechanism which takes part
in the production of adrenalin glycaemia has already been affected by
the operative disturbance. It is not an unfair assumption that the
additional impulse given by the application of adrenalin is on that
account less effective and its result more transient. Consequently at
the time of the collection of the second portions of blood the secretion
of sugar is lessening. The same reasoning may hold good with
regard to Experiment 5.

Comparison of the sugar content of the portal blood with that from
the femoral artery and hepatic vein in Experiments 1, 2, 3, and 5 shows
that the increase of sugar following treatment of the pancreas with
adrenalin is least in the portal vein. While in the control samples
the sugar percentage of the portal is as high or higher than that of
the femoral or hepatic blood, after treatment with adrenalin, it is
lower in every case. In this connection we would call attention to
certain changes in the appearance of the organs of the abdomen. At
the time of collection of the first samples of blood the appearance of
the intestines and pancreas was normal. As the experiment pro-
ceeded, however, the pancreas became congested and the intestines
cyanotic. The latter symptom is due, probably, to a partial obstruc-
tion of the circulation by the formation of a clot at the flange of the
cannula. The effect of this partial obstruction is a partial asphyxia
of the tissues. That the ve/ative decrease in the sugar of the portal
blood is dependent upon increased utilization within the tissues
through which it passes there can be no doubt. Whether this con-
sists in a mere increased oxidation of sugar, owing to the increased

of the Blood. 47

On O

Sugar Content and Coagulat

SsT‘0 6bL0 SLEQ) | Wid Fee O41 Lol Vel
*‘jewi0Uu |
Seu Son Lap Ole Seattle cl 20 0 +0¢'0 90¢'0 "Wd 69°C 6 +1 T¥1 L +1 +e VIL S
‘Teulon svo49 0610 S60 a1 0) ‘Wid OF C1 L0¢ Ll 6 ET
~uRq ‘UODaT[O9 puodas je |
SUNIL GLENS EUOIS) SSM SEH ALE F6l'0 9/10 SsTO TRENCOS TY SIESIIL L eT Site 4G L's +
(EMIIOU SBSTIUEd UO oe] 0920 81¢0 4970 | Wea crcl alt Sc FL
-[09 poo[q puoras Fo ay ye |
SHEILA USHA oU eH 91¢0 | FLL stc0 ‘Wd OCI eal esi SSL St SAIL €
VPA) rer lX6) FOWO |) airs cel 6 ll SST
‘TeuIIoU svaIDUeg “pooyq jo |
SHULOK ON) HOS! [ORAR AE [OMEN JONGSOISISES 8sto0. | 6LT0 €810 | Wd OF L9T sitell +1
jO 9UIT} oY} ye onjq yey |
SONOS CMB e1G | Se SHAUL 9910 0910 FLT 0 Wid tOb Ser ctl OCL 67 LL c
4170 =| SssTt0 LT@O =| ‘W'd S0'9 Ler Sit tell
s0z0 | 86T'0 oro | waoss | LET LUI ora
‘jusutiadxe jnoysno1yy cre 0 £600 070 Wd $0'S OPI 6C1 671
jewiiou sauljsejyur pue seai9
“urd ay} jo aouviwaddy 01 0 £1c0 9€¢ 0 Wid Sth +91 Ver S+1 cé Cell I
“Usd teq || 3ued tag || juo0!1eq) ‘suivis) | ‘sulviIn, | ‘suivig |‘sinoyy{] ‘Olly
“UIOA “ULOA ‘Arayre | “ULOA “UTDA *A1OV Ie ‘Juaw
sed v onedafy [v10g qerouag | “OPT | syedazy jewog | [eioway | ‘pes -lladxa
Silanes cl Jo DUIS | TYUSIOA\ jo
SE Sul ON

KAALAV IVAOWAY

WoO poolq ul Iwsng

wo1f poolq fo JuUNOWY

‘GALLINO ONIAL NITVNANGY HIIM LNAWLVAYY,

AHL GNV SNIAA IVLYOd ANV OILVdAR FHL WOVA AISNOANVLIOWIS NAMVL GOOTG AO INALNOD UVONS

48 Charles 1, Vosburgh and A. N. Richards.

supply of that substance, or whether there is a decomposition of
another character in increased amount, owing to lack of oxygen in
the tissues, our experiments do not decide.

In referring to Experiments 4 and 5, the idea has been expressed
that the high sugar content found in the control samples was possibly
due to the effects of operative disturbance. The question might
naturally be raised as to whether the effects noted in our other experi-
ments might not be due to that cause rather than to the influence of
adrenalin. To settle this point we have made a series of five control
experiments in which the blood was collected in a manner similar
to that described, the treatment with adrenalin being omitted. The
results are given in Table III, page 47.

These figures show an essential difference from those given in Table
II. In only one case does the blood of the hepatic vein contain con-
siderably more sugar than that of the femoral artery. In only one
experiment (3) is there an essential rise in the sugar of the femoral
blood. The results indicate, therefore, that while in a small percent-
age of experiments carried out according to this method, the opera-
tion may give rise to a hyperglycaemia similar to that produced by
adrenalin, in the majority of cases we are justified in attributing the
results to the action of adrenalin.

SUGAR IN THE BLOOD OF THE PANCREATICO-DUODENAL VEIN AFTER
TREATMENT OF THE PANCREAS WITH ADRENALIN.

It has been shown in the experiments of Series II, page 45, that in
the hyperglycaemia which follows the application of adrenalin to the
pancreas, the increase of sugar is least in the blood of the portal vein.
We have attributed this circumstance to an increased decomposition
of sugar in the intestinal tissues, and have suggested that it may be
connected with a partial obstruction in the circulation of the blood
through those tissues. To ascertain whether the congestion of the
pancreas which is regularly observed after treatment of that gland
with adrenalin takes part in this phenomenon, we have tested the
biood from the pancreatico-duodenal vein. Though an increased de-
composition of sugar in the pancreas would be at variance with our
ideas regarding the events taking place there, its possibility has not
been positively excluded.

The method of collecting blood was as follows: A cannula of the
design previously described was introduced into the portal vein at a

49

d.

Sugar Content and Coagulation of the Bloo

*pajysasuos yonut Ao
|

‘paysasuos yonu Ara A

‘poysasuood yeyMowog

‘ayed Ala

‘[BWL0 NT

96 '0
Sic 0
é1c 0

| 1aqze ‘99S QT “ULL OZ

“LO}Je 99S Of UIUL JT

“LOY “IIS Op “UL QZ

‘Ta}Je ‘Oas GT UW OT |

“AO}JeOIS OZ ULUTS

"19}Fe ‘00S Sh

‘pajsasuoo yvyMawos

raped Ara 4

*[BULIO NT

*paisasuos Ayqusis Aso;
| [Yo] A

Wd

*[BWLION

C070

“‘svarourd

jo
souvivaddy

“doqye ‘oas og uULtU S

“ADJJFE “O9S CZ

‘a1ofoq “ULL 6

S0°90'+-0€ +0'+ |

"layeoas gull g

‘aye “urut ¢

c

‘a10foq “UTUr CT |

quad Jag | yuU99 19g
“ihe Ti Woemiae

oea19uNd | [Blow ay

WOJF poOolq UL AwsNS

‘uonrordde
jo ou} YIM
paiedwmos uoy
-J9[[OD JO SWI],

‘Wd
"Ul[eVuaIpe
jo
uonrvoydde
FO SUIT,

Of 90'r

‘Wd

“WOT]IAT[OD
jo

aun,

‘suviry | *suei4
UIDA *A1gq1e
d1yvaINUR | [RIOWI YT

WO. poolq fo "Wy

‘SINOy]

"pez
QouIs
oun,

‘quot
-1adxa
jo
‘ON

50 Charles H]. Vosburgh and A. N. Richards.

point just opposite the entrance of the pancreatico-duodenalis. Loose
ligatures were placed about the portal vein, one on either side of the
cannula. The cannula was opened at the same time that the ligatures
were tightened. Blood is thus obtained from the desired vein, free
from admixture with portal blood. The portal circulation is inter-
rupted for a few seconds, but the pancreatic not at all.

We have collected blood in this manner both before and imme-
diately after painting the pancreas with adrenalin, and analyzed it for
sugar. The results are given in Table IV.

The results of these experiments are very uniform. In Experi-
ment I, the percentage of sugar in the pancreatic blood rose 0.073
per cent in the first three minutes after adrenalin was applied. In
Experiment 2, an increase of 0.068 per cent occurs within twenty-five
seconds. In Experiment 3, the rise in the first forty-five seconds
amounts to 0.032 per cent. In the last experiment, we have con-
tinued the collection of blood when the gland was very much con-
gested, and have compared these samples with portions taken at the
same time from the femoral artery. The analyses show a continued
rise in the sugar percentage, and only a slight difference in the blood
from the two sources. We are forced to conclude, therefore, that
there is not an increase in the decomposition of sugar in the pan-
creas antecedent to the rise of sugar in the general circulation, and
that the difference observed in the second series, between the reduc-
ing power of the blood of the portal vein and that of the femoral
artery, is not dependent on processes of this nature in that gland.

SUMMARY OF CONCLUSIONS.

1. The intraperitoneal injection of adrenalin chloride, as well as
the application of that substance to the pancreas, gives rise to a
marked increase of sugar in the blood. This hyperglycaemia makes
its appearance immediately after the administration, reaches its maxi-
mum in from one to three hours, and may continue for over fourteen
hours.

2. Simultaneously with hyperglyczemia occurs a decided diminution
in the time of extravascular coagulation of the blood. This phe-
nomenon appears to be due also to the application of adrenalin to
the pancreas.

3. The cause of this form of hyperglyczemia, as indicated by com-
parative analysis of the blood flowing to and from the liver, is to be

Sugar Content and Coagulation of the Blood. 51

attributed, in great part at least, to an increased sugar formation in
that organ.

We are indebted to Dr. C. A. Herter for the suggestion of the
subject of this work, and for valuable counsel during its progress.
We also wish to express our obligation to Mr. William D. Cutter for
assistance in a number of the operations.

THE IMMEDIATE INFLUENCE OF: EXERCISE UPOGNei ae
IRRITABILITY OF HUMAN VOLUNTARY MUSCLE.

By THOMAS ANDREW STOREY.
[From the Physiological Laboratory of Stanford University.]

1? is commonly the practice of athletes to go through a few pre-
liminary “warming up” movements before attempting a com-
petitive test. Sprinters always take a few short dashes immediately
before their events. It is said that horses never make their best
records in the first heat of a race.

The writer has noted such facts as these in connection with his
work in the Encina Gymnasium of Stanford University, and the
following experimental results are presented here because they seem
to offer some explanation.

In general, the experimental procedure adopted may be outlined
as follows: A single small muscle was substituted for the many
groups that participate in normal exercise. The condition of the
muscle was made evident by the amount of its contraction in response
to the single break-induced current. The muscle was exercised,
‘“warmed up,” by voluntary work with an ergograph, and the influ-
ence of that work upon the condition of the muscle was then tested
by means of the break current.

APPARATUS.

The apparatus employed here has been described by the writer
elsewhere. The essential pieces were: an ergograph for the index
finger; ! an electric pendulum for regulating the opening and closing
of the primary current to the inductorium, and for cutting out the
make-induced current; 2 and an inductorium with 10340 turns on the
secondary coil.

In those experiments in which the muscle was excited electrically,
the primary current was furnished by three Edison-Lalande cells,
“type S,” and the secondary coil was placed at a distance of 10 cm.

' STOREY: This journal, 1903, vill, p. 355-
2 SToREY: This journal, 1903, vili, p. 435.
52

Influence of Exercise upon Human Muscle. Pe

from the primary coil. The rate of excitation, electrical or voluntary,
was once a second. The resistance to contraction in response to
electrical excitation was formed by a light, twisted, rubber band which
was attached to the recording device. With voluntary excitation the
resistance was furnished by spiral springs of different strengths, as
indicated with the different figures below.

The first dorsal interosseus muscle of the left hand was used in
every experiment.

The writer experimented upon himself in the cases herein reported.
Other individuals have been used, and the evidence which they have
furnished is the same as that shown below.

THE INFLUENCE OF A SERIES OF SINGLE ELECTRICAL EXCITATIONS
UPON THE IRRITABILITY OF HUMAN MUSCLE.

It is well known from studies on experimental animals that a
muscle stimulated by a succession of induced currents of equal
intensity responds by contractions that gradually increase in height;
in other words, the irritability of the muscle is increased. Harley!
has noted this “staircase” in a series of voluntary contractions of
human muscle, Waller? and the writer? have reported it with
electrical excitation of human muscle.

Summation caused by electrical stimulation of human muscle
has, however, not yet been noted so far as I am aware. I have
observed that a strength of the induced current which is at first
insufficient to produce a visible contraction of the human abductor
indicis will give rise to contraction after from five to fifteen stimuli
have been applied. In such a case, subsequent contractions will
rise in the staircase form.

THE INFLUENCE OF A SHORT SERIES OF VOLUNTARY CONTRACTIONS
UPON THE IRRITABILITY OF HuMAN MUSCLE.

In the two experiments recorded in Fig. 1, the procedure was as
follows: The muscle was made to contract a number of times against
the resistance of the rubber band mentioned above, in response to

' HARLEY: Journal of physiology, 1894, xvi, p. roo.

2 WALLER: Report to the Scientific Grants Committee, British Medical
Association, 1885, p. 68.

3 STOREY : This journal, 1903, viii, p. 373.

54 Thomas Andrew Storey.

the single break current. After this several seconds were spent in
increasing the resistance. In experiment A, a light spiral spring
was used, and in experiment B, a heavy spiral spring was added.
As soon as the resistance was arranged, the muscle was contracted
voluntarily a number of times; but the contractions were not con-
tinued long enough to cause fatigue in either experiment. Then
followed three or four seconds of rest, during which time the spring
resistance was removed. Finally, the muscle was again excited
electrically, precisely as in the first part of each experiment.

It is evident that the response of the muscle to electrical excitation
with the single induced shock is very much greater after than before

“a i

(1) (2) (3) (1) (2) (3)

FicukE 1.— The curves read from left to right. 4 (1), contractions upon electrical stim-
ulation, against very slight resistance; (2), voluntary contractions against consider-
ably increased resistance ; (3), conditions as in (1). 4. Conditions as in Series 4,
except that the resistance to voluntary contraction was very much greater.

a series of voluntary contractions against some resistance. The irri-
tability of the muscle then is considerably increased by a moderate
amount of ergographic work, that is to say, by a “warming up”
process.

THE INFLUENCE OF VOLUNTARY MUSCULAR FATIGUE UPON THE
IRRITABILITY OF THE WORKING MUSCLE.

In Fig. 2 the experiment is formed of three parts, as in Fig. 1.
There is, first, a series of electrical excited contractions, then a series
of voluntary, and finally a second series of electrically excited con-
tractions. The voluntary work was done with a moderately strong
spiral spring, and was continued until some fatigue was present, as is
evident in the curve inscribed.

Influence of Exercise upon Human Muscle. 55

There are no staircase contractions in the series of electrically
excited contractions beginning the experiment in Fig. 2. This is
due to the fact that some muscular work preceded the electrical
excitation, so that the period of the staircase contractions had been
passed.

On comparing the effect of electrical excitation before fatigue

iy Lh
| Tee
il SAHIN AVHEULHNOASREA !

(2) (3)
FIGURE 2, — i. Electrical stimulation, very slight resistance; (2) voluntary contraction
against considerably increased resistance ; (3) conditions as in (1).

1
_ shovhod bak \ anh

with the effect of electrical excitation after fatigue (Fig. 2), it is
evident that the irritability of the muscle has been immediately
very greatly reduced by the fatiguing work.!

CONCLUSION.

The experiments herein presented make it evident that human
voluntary muscle is made more irritable by successive excitations ;
that the irritability of human voluntary muscle is immediately very
greatly increased by a moderate amount of work; and that it is
very greatly decreased by a fatiguing amount of work.

It may be concluded that the ‘ warming up” habits referred to in
the opening of this paper are of value in athletic contests, because by
such means the irritability of the acting muscle groups is greatly
heightened, and they are then able to attain more quickly their
maximal strength and speed of contraction.

1 See also, Srorey: This journal, 1903, viii, p. 360.

Peoeu DY OF THE VASOMOTOR NERVES OF THE
RAUBIT'’S EAR CONTAINED, IN- THE THIRD
CERVICAL.-AND: IN. THE CERVICAL
SYMPATHETIC: NERVES.

BY S.-J. MELTZER Anp CLARA MELTZER!
[From the Rockefeller Institute for Medical Research. |

INTRODUCTION.

INCE the classical experiments of Claude Bernard and Brown-
Séquard on the effects which section or stimulation of the
cervical sympathetic nerve exerts upon the blood-vessels of the ear,
the sympathetic has been considered to be the carrier of the vaso-
motor nerves of the ear. In most of the leading text-books no men-
tion is made of the third cervical nerve in this connection. Yet the
fact that the auricularis magnus, a branch of the third cervical, carries
vasomotor fibres to the ear became known soon after the discovery of
vasomotor nerves. As early as 1854, Schiff? stated that section of
the cervical nerves in which run the fibres of the great auricular nerve
causes congestion of the ear, while stimulation of the peripheral end
causes constriction of the vessels. Schiff gives no particulars regard-
ing the extent or the intensity of the dilatation. Moreover, accord-
ing to Schiff, the congestion which follows section passes off in a few
hours.

Besides these statements of Schiff, there are in the literature a few
reports of other investigations bearing upon this subject. These re-
ports, however, are somewhat contradictory. Lovén? (1866) stated
that the auricularis magnus contains vasoconstrictors for the tip and
the sides of the ear. Pye-Smith * (1884) states that in his experiments
the great auricular nerve exerts no vasomotor influence upon the ves-
sels of the ear. Morat® (1891), who experimented on dogs, concluded
that the auricular nerve is a vasoconstrictor of the ear. In one

1 Research Scholar of the Rockefeller Institute.
2 SCHIFF: Beitrage zur Physiologie, 1894, i, p. 148.
8 LovEN: Arbeiten aus der physiologischen Anstalt zu Leipzig, 1867, p. 1.
* PyE-SMITH: Journal of Physiology, 1887, viii, p. 25.
5 Morat: Archives de physiologie, 1892, p. 92.
57

58 S. J. Meltzer and Clara Meltzer.

experiment in which the right sympathetic was cut eight days before,
section of the cervical nerve on both sides caused a congestion of both
ears, which was greater on the side on which the sympathetic was
intact. Fletcher! (1898) studied the effect of stimulation of the periph-
eral cut end of the third cervical nerve and found that it sometimes
causes a constriction even of the proximal third of the central artery.
According to Tigerstedt, Moreau? also claims that the auricular nerve
contains vasoconstrictors for the ear.

Our observations were made on rabbits. We have cut the third
cervical nerve on one side and have compared the effect upon the
blood-vessels of the various areas of the ear of the same side, with the
effect of cutting the cervical sympathetic nerve or of removal of the
superior cervical ganglion, either on the other side of the same animal
or on the same side of other animals. We have also compared the
effect of stimulation of the cut end of the third cervical nerve of one
side with that of the stimulation of the sympathetic of the same side,
as well as with the stimulation of the third cervical or of the sympa-
thetic nerve on the other side. . The effects of cutting the nerves as
well as of their stimulation were often studied on one and the same
animal. In some cases the effects of stimulation were studied im-
mediately upon cutting the nerves, in other cases the nerves were
stimulated many days after section, and in still others the freshly
cut nerves were studied on one side, while on the other side the
nerves had been cut a few days before.

The hair of the ears was removed by scissors. The use of depila-
tories, although giving satisfactory results otherwise, was not advis-
able, on account of their irritating effect upon the skin of the ears,

producing a degree of vasodilatation which usually lasted for some
time.

RESULTS OF THE CUTTING EXPERIMENTS.

The operation was performed while the animal was under ether.
When the third cervical nerve was handled, the animal had to be kept
under deep anzsthesia. We attached but little importance to the
changes which followed immediately upon cutting the nerves while
the animal was still under anzsthesia. Frequently both ears are well
congested, while the animal is still under the influence of ether, espe-

1 FLETCHER: Journal of physiology, 1897-98, xxii, p. 259.
* MoreEAu: Memoires de physiologie, Paris, 1872. Quoted from TIGERSTEDT:
Lehrbuch der Physiologie des Kreislaufes, Leipzig, 1893, p. 482.

A Study of the Vasomotor Nerves of the Rabbits Ear. 59

cially as long as it remains tied on the board. Indeed the ear on the
non-operated side is the one which at times appears the more con-
gested, apparently on account of some actively vasodilating influence
of central origin. Immediately upon removal from the board, how-
ever, there is already a marked difference in the congestion of the two
ears. Our results are derived from many comparisons made at fre-
quent intervals for many days following the operation.

We have made a special study of the subject in hand on twenty-two
animals. We have, however, also made a good many more observa-
tions on animals which were similarly operated, but chiefly for the
purpose of other studies. Our observations brought out the following
results :

1. In five out of the twenty-two animals, section of the third cervi-
cal nerve was followed by a dilatation of all the blood-vessels of the
entire ear, including the central artery. In four of these animals, the
sympathetic nerve was cut, or the superior cervical ganglion removed
on the other side. The contrast in the congestion of the two ears
was striking. The vasodilatation due to the removal of the influence
of the sympathetic was confined to the lower two-thirds or three-
fourths of the central artery and adjacent parts. The ear on the side
where the third cervical was cut was uniformly congested throughout
its entire extent and with greater intensity. The central artery was
dilated throughout its entire length, and was distinctly wider than on
the sympathetic side. In all parts of the ear, numerous fine vessels
made their appearance; while the entire ear had a pinkish hue. The
picture on page 60 is from a photograph! taken ten days after
the third cervical nerve was cut on the left, and the sympathetic on
the right side. The difference is well represented ; but, in reality, the
contrast was still more marked.

2. In four animals, section of the third cervical nerve caused only a
very moderate dilatation of some blood-vessels at the tip and upper
half of the sides of the ear. The subsequent cutting of the branches
connecting the third with the second and fourth cervical nerves
caused, in one animal, a uniform congestion of the entire ear, in an-
other, a dilatation of the blood-vessels of the entire ear, except the lower
two-thirds of the central artery ; in the other two animals, the cutting
of the connecting branches caused no change.

! We are indebted for this photograph to Dr. EpwARD LEAMING, Depart-
ment of Pathology, College of Physicians and Surgeons, Columbia University,
New York.

60 S.J. Meltzer and Clara Meltzer.

3. The remaining thirteen animals can be put in one group, in
which, after cutting the third cervical, all the animals had dilatation
of the blood-vessels of at least the sides and top of the ear, including
the bifurcation of the central artery, and an absence of dilatation of
at least the lower third of the central artery. Within these limits
there has been a considerable variation in the extent and in the in-
tensity of the congestion of the ears in different animals. In some,
the entire ear looked very congested with the exception of a pale
nucleus corresponding to the lower part of the central artery; in
others, the ear presented a
normal appearance with
the exception of a moderate
flushing at its periphery.
In nearly all cases, the cen-
tral artery has shown at
some point an abrupt tran-
sition between the dilated
and non-dilated parts. In
two cases, the subsequent
cutting of the sympathetic
filled out exactly the non-

dilated part without chang-
Rabbit’s ears, Experiment 47, ten days after section

i ea
of sympathetic on the right and cervical on the ae — Ms Bes or of the
left side. other parts of the ear. In

another case, the subse-
quent cutting of the branches connecting the third with the fourth
cervical nerve completed the dilatation, increasing however at the
same time the intensity of the congestion of the entire ear. In two
other cases, the cutting of the third cervical alone caused originally
a congestion of the entire ear; the next day, however, the conges-
tion was found to have become restricted more to the periphery,
leaving a pale area at the centre of the ear. For this reason, we put
these animals into the third instead of the first group. We have to
add that the subsequent cutting of the connecting branches of the
third cervical nerve we began only when we were already near the
end of our series of experiments. We are therefore unable to state
how many of the animals of the third group would have shown a con-
gestion of the entire ear had we subsequently cut these connecting
branches.

Judging from our experiments, it appears that section of the third

A Study of the Vasomotor Nerves of the Rabbit's Ear. 61

cervical nerve on the left side causes a more extensive congestion than
section of the same nerve on the right side. It also appeared that
the congestion was greater in gray or brown animals than in white
ones. However, the number of animals we have experimented upon
is not large enough to permit positive conclusions to be drawn in this
direction.

4. The dilatation which followed section of the sympathetic nerve
was in the majority of cases restricted to the central artery and the ©
adjacent region. In most of these cases the dilatation of the artery
extended only over its lower two-thirds, and in some not even so far.
In these cases the dilatation had mostly an abrupt termination. There
was one case in which section of the sympathetic caused a congestion
of the entire ear; the dilatation of the central artery, however, was not
of even width throughout its entire length, but tapered distinctly
toward the top. This was in marked contrast with the character of
the dilatation of the central artery when it occurred after cutting the
third cervical nerve. There was only one other experiment in which
the entire ear was congested after section of the sympathetic, but a
few hours after operation the congestion had already receded consid-
erably. The difference between the right and left sides which we
have noted above with regard to the effects of the section of the third
cervical, seemed also to hold good for the sympathetic; the congestion
was more marked after section of the left than after section of the
right sympathetic nerve. We may add here that in only one exper-
iment did the subsequent removal of the superior cervical ganglion
seem to improve somewhat the effect which followed simple section
of the sympathetic. In all other experiments the subsequent removal
of the ganglion brought no additional changes, not even when the
removal of the ganglion occurred a few days after the section of
the sympathetic, and the primary effect was already distinctly
diminished.

5. Our experiments show that, in general, the sympathetic and
cervical nerves carry vasomotor fibres for different areas of the ears, .
the former controlling the centre, and the latter the periphery of the
ear. The influence of the auricularis magnus, however, seems to be,
in many respects, greater than that of the sympathetic. (a) The con-
gestion of the parts, due to section of the cervical nerves, appeared
always more intense than that which followed section of the sympa-
thetic. (0) Often the centre was fully congested after section of the
cervical nerves, but the periphery was very rarely fully congested after

62 S.J. Meltzer and Clara Meltzer.

section of the sympathetic. (c) We noted in two experiments that
the central artery, which was dilated after section of the sympathetic,
became distinctly more dilated on subsequent cutting of the third
cervical nerve. But we find no note in our experiments indicating
that the subsequent cutting of the sympathetic improved the conges-
tion of the periphery or of the centre of the ear, which followed sec-
tion of the third cervical nerve. (a) The congestion which follows
section of the third cervical nerve (and its connections) persists dis-
tinctly longer and in greater intensity than that which follows section
of the sympathetic. When the cervical nerve is cut on one side, and
the sympathetic on the other, the difference is striking for ten or
fourteen days. While the blood-vessels on the sympathetic side be-
come constricted to nearly their original width, and show rhythmical
changes again, the ear on the cervical side still looks well congested,
with practically no rhythmical changes. In view of our experience
we can hardly understand the statement of Schiff that the congestion
which follows the cutting of the third cervical lasts only a few hours.
On the contrary, in some of the experiments at least, the congestion
did not develop fully until some time after the operation.

6. As is well known, and as was mentioned above, the congestion
following section of either nerve decreases more or less rapidly.
Many writers ascribe this decrease of the vasodilatation to the as-
sumption of the constricting tonus by some uncut vasoconstrictor
nerve fibres, which normally.do not participate in the maintenance of
the tonus, but which, in the absence of the chief constrictors, are
ready to substitute the latter, —a sort of collateral innervation. In
our experiments, we had a few animals in which on one side the su-
perior cervical ganglion was removed, and the third cervical nerve and
all its connecting branches were cut. We had reason to believe that
we had thus excluded all vasoconstrictors of the ear. Nevertheless,
a few weeks after the operation, there was little left of the original
marked congestion. In these cases, the recurrent constriction could
not be the work of substitution or collateral innervation. Neither
could it be the result of regeneration. Aside from the shortness of
time, the occasional awéopsies in vivo have shown that there was as
yet no regeneration. Apparently the walls of the blood-vessels them-
selves possess the capacity of resuming their tonicity without the aid
of extrinsic nerve influence. Furthermore, we had experiments which
seem to demonstrate that at least in these cases no substitution took
place. For instance, the ganglion was removed and a dilatation of the

A Study of the Vasomotor Nerves of the Rabbit's Lar. 63

central artery followed. When, later, after this dilatation had dis-
appeared, the third cervical and its connecting branches were then
cut, a congestion of the periphery appeared while the centre remained
pale. In this case the subsequent constriction of the central artery
was apparently the work of the blood-vessel itself, and was not due to
an assumption of the tonus by the other vasoconstrictor fibres of the
ear. It is true, we had a number of experiments in which, after the
congestion following the primary cutting of the sympathetic had
diminished, a subsequent cutting of the cervical nerves caused, in
addition to the congestion of the peripheral parts of the ear, also a
redilatation of the central artery. This, however, does not necessa-
rily mean that the cervical assumed the control over the central artery
only after the sympathetic had been cut, since we know, as we have
seen above, that the primary cutting of the cervical nerves alone often
causes a dilatation of the central artery.

Our experiments, therefore, furnish evidence that the constriction
of the blood-vessels can return apparently without the aid of extrinsic
nerves. There is, furthermore, satisfactory evidence that, at least in
some cases, when the cervical nerve does not participate originally in
the maintenance of the tonus of the central artery, this nerve does not
assume the tonus after the influence of the sympathetic is eliminated.
Finally our experiments afford no sufficient evidence that a collateral
innervation ever takes place in the tonus of the vasomotor nerves of
the ears; z.¢., we have no evidence that the cervical can take the
function of the sympathetic, or the sympathetic that of the cervical
nerve, if the substituting nerve had originally no active share in the
maintenance of the vascular tonus.

On the other hand, we had experiments in which the dilatation of
the central artery due to the cutting of the sympathetic was distinctly
increased throughout the entire length of the artery by cutting the
cervical nerve a few minutes later. This can only mean that the
vasomotor tonus of the entire central artery was simultaneously
maintained by the sympathetic as well as by the cervical nerves.
This condition was observed to exist more frequently in the tonus of
the upper part of the artery. Our observations in this respect, how-
ever, are too few in number to permit a detailed discussion of these
complicated conditions.

64 S.J. Meltzer and Clara Meltzer.

RESULTS OF STIMULATION EXPERIMENTS.

As far as we know, Fletcher is the only investigator who has given
a detailed account of the effect of stimulation of the third cervical
nerve. According to his experience, electrical stimulation causes the
greatest constriction in the distal third of the central artery, but the
proximal third also shows a marked constriction. To obtain an effect
from this nerve, Fletcher states that the stimulating current must be
stronger than that which elicits an effect from the sympathetic.
Again, the latent period in the stimulations of the third cervical
nerve is much longer than that in stimulations of the sympathetic.
Finally, according to Fletcher, there is a difference in the order of
the return of the flush after discontinuation of stimulation. After
stimulation of the sympathetic, the return flush begins distally and
travels towards the base of the ear; after stimulation of the third
cervical nerve, it begins proximally and affects the terminal bifurcation
of the artery last.

Our stimulation experiments were made on sixteen animals. In
many of them the cervical nerve as well as the sympathetic was
stimulated on both sides. In a few cases, the nerves were stimulated
five days after they had been cut. We shall give only a brief sum-
mary of our results.

7. For the majority of animals, it can be stated in a general way
that stimulation of the third cervical nerves caused chiefly a constric-
tion of the vessels of the sides and top of the ear, including also the
upper end of the central artery; stimulation of the sympathetic
caused chiefly a constriction of the lower three-fourths of the cen-
tral artery and adjacent parts of the ear. In a smaller number of
cases, stimulation of the third cervical nerve caused also a complete
constriction of the entire central artery, and in very few instances
did stimulation of the sympathetic cause pallor also in the sides
and top of the ear.

8. In nearly all cases, the pallor following stimulation of the
third cervical nerve spread from the top downward, and the con-
striction of the artery caused by stimulation of the sympathetic
spread from the base of the ear upward. Regarding the order of
the refilling of the vessels after discontinuation of the stimula-
tion, we have too few notes tu permit a general conclusion to be
drawn.

A Study of the Vasomotor Nerves of the Rabbit's Ear. 65

9. Regarding the intensity of the constriction which is caused by
the stimulation of either nerve, we found in our experiments no differ-
ence in favor of the sympathetic. On the contrary, in our experi-
ments the general pallor, as well as the degree of constriction of the
individual vessels, was often greater after stimulation of the cervical
nerve than after that of the sympathetic. The same must be said of
the strength of the stimulus which is required to bring out an effect
from each nerve, and of the length of the latent period in each case.
Fletcher as well as Morat state that the interrupted current which is
sufficient to cause constriction upon stimulation of the cervical nerve,
must be considerably stronger than ordinarily employed in nerve
stimulation. Fletcher ascribes this to the thickness of the sheath of
the cervical nerves. In our experiments, we had a relatively large
number of instances in which a weaker current brought out a strong
effect from the cervical nerve, while a stronger current brought out a
comparatively slight effect from the sympathetic of the same side.
For instance, in two experiments, stimulation of the right cervical
nerve with the interrupted current, while the secondary coil was at a
distance of 130 mm., caused great pallor, and a constriction of the
blood-vessels of the entire ear, except the lower two-thirds of the
central artery; while stimulation of the right sympathetic with
the secondary coil at a distance of only 90 mm., caused only a very
moderate constriction of the lower two-thirds of this artery. Our
experience with regard to the strength of the stimulus is as follows:
There have been cases in which equally strong currents brought out
equally strong effects from both nerves. In other cases, as stated
above, a weaker current brought out a strong effect from the cervical
nerve, and a stronger current caused a mild effect from the sympa-
thetic. In still other cases, only a strong current elicited some effect
from the cervical nerve, while a weaker current elicited from the
sympathetic a constriction leading to the entire disappearance of the
central artery. Only this latter class of cases, but few in number,
corresponds with those seen by Fletcher.

Similar variations were observed with regard to the latent period.
But here we can say, in general, that in all cases in which the stimula-
tion caused a strong effect, the latent period was short. When the
effect was moderate, and the current had to be strong, the effect set
in later. It would seem to us that the thickness of the sheaths of the
nerves has little to do with these variations ; they seem to be due
rather to individual variation, which can, of course, only be recognized

66 S. J. Meltzer and Clara Meltzer.

when the experiments are extended over a large number of animals.
Fletcher seems to have experimented upon only four animals.

We have also notes on the length of the after-effect in some experi-
ments. As far as the few data permit any conclusion, it would seem
that the long after-effect followed generally moderate primary effects
brought on by strong currents. When the primary effect was a
strong one, the return flush usually set in very soon after the stimula-
tion was discontinued. However this may be, we can positively state
that the long and short after-effects were equally divided between the
cervical and the sympathetic nerves.

We have thought it necessary to dwell especially upon these points,
because the strength of stimulus, latent period, and after-effect are
criteria sometimes employed to distinguish between different kinds of
nerves, and might have been looked upon also in our case as physio-
logical criteria distinguishing the sympathetic from the cervical
nerves. Furthermore, such a physiological distinction would seem to
run parallel with a certain anatomical distinction which appears to
exist between the sympathetic and the cervical nerve fibres. Accord-
ing to the investigations of Fletcher, the fibres which pass to the ear
by the route of the cervical nerves arise from the cells of the ganglion
stellatum. These vasomotor fibres are therefore post-ganglionic,
while the vasomotor fibres within the cervical sympathetic are pre-
ganglionic. We could then be misled into the belief that the longer
latency of period, and the requirement of greater intensity of stimulus,
etc., might be qualities peculiar to post-ganglionic nerve fibres. We
therefore took occasion to state especially that in our experiments
these qualities were not peculiar to either set of nerve fibres, but seem
to depend rather upon the individual variations of the animals. The
basis of these variations, whether they are simply due to variations
in the distribution of the number of vasomotor fibres of the same
character between the cervical and sympathetic nerves, or whether
there is a variation in the distribution of fibres of different character,
we do not wish to discuss for the present.

10. In nearly all cases in which both sides were compared in the
same animal, stimulation of the left sympathetic gave a distinctly
better effect than stimulation of the right, with regard to the degree
as well as the extent of the constriction. This holds good in a gen-
eral way also for stimulation of the cervical nerves; here, however,
we had one animal in which stimulation of the right cervical nerve

gave the better effect, and two in which the effect was about equal on
both sides.

A Study of the Vasomotor Nerves of the Rabbit's Ear. 67

This effect of stimulation coincides with the above-mentioned effect
of section of the nerves in which the congestion was found to be often
more intense on the left than on the right side. While the number
of our experiments is, perhaps, as yet too small to justify the general
conclusion that in all rabbits the left nerves contain a larger number
or more efficient vasomotor fibres than those on the right side, it is at
least well to keep in mind that this was the case in a number of con-
secutive experiments. The importance of this will become evident
when we consider that among the methods employed to study the
question, whether there is any difference in the effect between simple
section of the cervical sympathetic, and the removal of the superior
cervical ganglion, there was one which consists in the comparison of
the effects following section of the sympathetic of one side and re-
moval of the ganglion on the other side. In view of our experience,
it is obvious that this method is incapable of leading to decisive re-
sults. This method presupposes that the effect of section of the
sympathetic, or the removal of the ganglion, is in all animals the same
on both sides. Now assuming even that the greater effect obtained
on the left side in our experiments was a matter of accident, these
experiments demonstrate unmistakably that presupposition of simi-
larity on both sides is not permissible, and therefore any inference
drawn from a method which contains this supposition as a premise
cannot be conclusive.

11. In our experiments, we met with cases in which the stimulation
of the nerves had but little effect, while their section brought out a
considerable flushing of the ear. This shows that there is a differ-
ence between the normal stimulation which maintains the arterial
tonus, and the artificial stimulation of the end of the cut nerve. This
difference might be due to the fact that the normal tonus is main-
tained by stimuli which are certainly more adequate for nerve excita-
tion than the electrical current; but it might also be explained by the
assumption that the normal stimulus affects solely or preferably one
set of nerve fibres, while the artificial stimulus affects simultaneously
and indiscriminately two antagonistic sets of fibres. ‘This recalls the
relations between the inhibitory fibres of the vagus and the accelerat-
ing nerve fibres. While a simultaneous artificial stimulation of both
kinds of nerve fibres always favors inhibition, nevertheless reflex ac-
celeration is possible, and it seems that a normal tonus of the accel-
erating nerve fibres is being continually maintained.

68 S. J. Meltzer and Clara Meltzer.

SUMMARY.

The more essential points of our investigation are:

1. In the majority of cases, the third cervical nerve carries the vaso-
motor fibres for the blood-vessels of the rabbit’s entire ear, except for
a comparatively small area around the lower two-thirds of the central
artery, for which the sympathetic is'the carrier of the vasomotor fibres.
In a good many cases, the cervical nerve innervates also the vessels
of the centre, and in very few exceptional cases, the sympathetic
carries fibres also for the periphery. There is probably in all cases
a zone in which the vascular tonus is maintained by the nerve fibres
of both nerves simultaneously.

2. The congestion following the section of the cervical nerve lasted
in all cases longer than that following the section of the sympathetic.

3. Section as well as stimulation of both nerves caused a distinctly
better effect on the left side than on the right.

4. There was no constant proportion between the effect of section
and that of stimulation.

5. Even after section of all the vasomotor carrying nerves, the
blood-vessels, sooner or later, become constricted again, probably
through some intrinsic activity of the blood-vessels themselves.

Mee or BCirIC “ROTATION OF ) THE NUCLEIC ACID
OF THE WHEAT EMBRYO.

bY THOMAS 5B. OSBORNE.

[From the Laboratory of the Connecticut Agricultural Experimental Station. |

AMGEE and Jones! have recently described nucleoproteids
from several sources, all of which showed the unexpected
property of right polarization. As it is possible that the nucleic acid
component may, in whole or in part, be the cause of this dextro-
rotation, I have, following the suggestion made by Dr. Gamgee, ex-
amined the rotation of the nucleic acid which I have obtained from
the wheat embryo.”

The specific rotation was determined by suspending the dry acid
in water and gradually adding decinormal potassium hydroxide solu-
tion until all was dissolved. In this way a perfectly clear solution
was obtained which reacted strongly acid with litmus and contained
the nucleic acid as acid potassium nucleate. The rotation of the solu-
tions was observed at 20°, with the following results : —

I. Observed angle +3.16°. Amount of dissolved acid =.0236 gr.
perc.c. Length of the tube 2 dm: —

(a) = +66.95°.
2. Observed angle +5.84°. Amount of dissolved acid =.0400 gr.
perc.c. Length of tube 2 dm: —
(a) = +73°.

3. Observed angle 2.89°._ Amount of dissolved acid =.0393 gr.
per c.c. Length of tube 1 dm: —

(oj = Piaga:

1 GAMGEE and JOoNEs: This journal, 1903, vill, p. 447.
2 OsBORNE and Harris: Zeitschrift fiir physiologische Chemie, 1902, xxxvi,
p. 85 ; also Report Connecticut Agricultural Experimental Station for 1gol.
69

70 Thomas B. Osborne.

These solutions showed no change in rotation after standing twenty-
four hours.

From these results it is evident that this nucleic acid is strongly
dextrorotatory and that the degree of rotation is considerably influ-
enced by the concentration of the solution.

In order to determine the rotation of a mixture of protein substance
with nucleic acid, a quantity of triticonucleic acid was dissolved in
water with addition of an amount of potassium hydroxide which just
sufficed for solution, and then one half as much pure ovalbumin was
added, and the solution examined in a 200 mm. tube, with the follow-
ing result : —

4. Observed angle 2.50°. Amount of dissolved substance .0404
gr. per c.c. Length of tube 2 dm:—

(a)? = +30.94°.

This ,is approximately the rotation calculated for a mixture of one
part of ovalbumin (a)? = —30° and two parts of wheat nucleic acid
(a)? = +67°, the mean rotation of which would be +35°.

It is thus evident that a combination of protein substance with
nucleic acid may show strong right polarization, and that this dextro-
rotation may be wholly due to the nucleic acid component.

Gamgee and jones give no data from which the proportion of pro-
tein and nucleic acid in their nucleoproteids can be inferred, except in
the case of ‘‘ Hammarsten’s preparation,” which designation, although
they do not say so, presumably refers to the nucleoproteid obtained
by Hammarsten’s method from the pancreas.!_ This nucleoproteid
according to Hammarsten, contains 4.5 per cent of phosphorus, from
which we may assume that it contains about 50 per cent of nucleic
acid, since Levene? found 8.65-9 per cent of phosphorus in his prep-
arations of the nucleic acid of the pancreas. If the dextrorotation
of the nucleoproteid of the pancreas is wholly due to the nucleic acid,
the specific rotation of the acid in ‘‘ Hammarsten’s preparation ” must
be very high in the light of Gamgee’s and Jones’s figure. This, how-
ever, was obtained in such an extremely dilute solution that it is quite
possible that it is too high.

The figures given for the other nucleoproteids are such as might
be caused by the dextrorotation of the nucleic acid. That this is the

' HAMMARSTEN: Zeitschrift fiir physiologische Chemie, 1894, xix, p. 19.
* LEVENE: Jdid., 1901, xxxii, p. 548.

The Specific Rotation of Nucletc Acid. VB

case is indicated by the fact that, as Gamgee and Jones say, ‘The
specific rotation of the nucleoproteid is +38°, that of the nuclein
+65°, while it can- be indirectly shown that a substance is contained
in the residual material whose specific rotation is about +81°.”
We thus see that with a probable increase in the proportion of

nucleic acid in the compounds examined there was an increase in the
dextrorotation.

THE INFLUENCE OF COLD ON THE ACTION OF
SOME HAMOLYTIC AGENTS}

By G. N. STEWART.
[From the Hull Physiological Laboratory of the University of Chicago. |

HAVE already shown? that sapotoxin causes an increase in the
electrical conductivity and in the permeability to electrolytes of
formaldehyde-hardened red corpuscles, and have, without being able
to directly prove this, brought forward evidence that it exerts a
similar action on the unfixed corpuscles. The difficulty in making
observations on the latter consists in the rapidity with which sapo-
toxin causes laking. This increases the conductivity by the liberation
of electrolytes on the one hand and diminishes it by the liberation of
hemoglobin on the other, and so obscures the phenomenon which
comes so clearly to light in the case of formaldehyde corpuscles.
The fact that at o° C. certain of the biological laking agents are said
not to act, suggested that by cooling the blood to 0°, and using small
doses of sapotoxin, an action on the conductivity might be revealed
before any laking occurred. Dr. Peskind showed that, as a matter
of fact, under the conditions mentioned, laking is delayed for a con-
siderable period, a period which, as I have found, may extend to
many hours.

The procedure was as follows. The dose of sapotoxin necessary to cause
laking of blood at air temperature was first determined. Then a meas-
ured volume of defibrinated blood was cooled in ice to o°, and a pre-
determined quantity of ice-cold sapotoxin solution (of course in NaCl
solution) added to it. The mixture was rapidly shaken up and returned
to the ice. A control specimen of defibrinated blood containing as much
of the 0.9 per cent NaCl solution used in making the sapotoxin solution,

* Some of the experiments included in this paper were made in conjunction
with Dr. S. PESKIND, in the Physiological Laboratory, Western Reserve Uni-
versity, during his tenure of the H. M. Hanna Fellowship.

* Journal of physiology, 1899, xxiv, p. 211; I90I, xxvi, p. 470; Journal of
experimental medicine, 1902, vi, p. 257.

72

Influence of Cold on the Action of Hemolytic Agents. 73

_ as was added to the blood of the sapotoxin solution, was also kept in the
ice. Another specimen of defibrinated blood, containing as much of the
sapotoxin solution as the first, was kept at air temperature. From time
to time resistance measurements of these mixtures were made, a measur-
ing tube being immersed in crushed ice, into which dipped a thermometer.
The readings of the thermometer never varied more than one or two
tenths of a degree from o°. The tube containing the electrodes was
always immersed for some time in the ice before the blood mixtures were
poured into it. In the tables the conductivities (at o°C.) are expressed
in reciprocal ohms X 10°. The commencement and progress of laking
in the mixtures was controlled by putting from time to time into a cooled
graduated centrifuge tube 1 c.c. of the mixture, filling up the tube to
15 c.c. with the ice-cold salt solution, and centrifugalizing in a cold room
with a centrifuge of such speed that complete sedimentation occurred in
four and one-half minutes. The amount of hemoglobin, if any, in the
supernatant liquid was then determined colorimetrically.

In this way it has been shown that sapotoxin, when added to
defibrinated blood at 0°, produces for some time no effect on the
conductivity and no laking. For example, in Experiment I, a dose of
2 per cent sapotoxin solution, corresponding to 6 c.c. to 100 c.c. of
blood, caused for at least two and one-half hours no change in the
conductivity and no laking. Then an increase in the conductivity
developed, which reached its maximum after about twenty-one hours,
practically no laking having taken place up to this time. The dose
of sapotoxin was insufficient to cause any increase in the conduc-
tivity of the specimen left at ordinary temperature, except for the
relatively short period before Jaking had begun, or while the amount
of hemoglobin liberated was still insignificant.

In Experiment II, with a different specimen of blood, but the
same dose of sapotoxin, the period of delay before a change of
conductivity occurred was much less. In one hour, an increase in
conductivity was marked, and indeed had reached its maximum;
whereas in thirteen minutes it had not begun. In one hour and
thirteen minutes, the blood had not begun to lake. This dose was
sufficient to cause a slight increase in the conductivity of the blood
laked at ordinary temperature.

In Experiment III (1), in which the blood was practically fresh,
the same dose of sapotoxin caused but little increase of conductivity
for an hour. After two hours, the conductivity was distinctly, and
after eighteen and one-half hours, markedly increased, while no laking

74 G. NV. Stewart.

took place before the twenty-third hour. After forty-eight and one-
half hours only 15 per cent of the hamoglobin was in solution. The
dose of sapotoxin was sufficient to cause an increase of conductivity
in the partially laked blood left at room temperature, but when laking
was almost complete this increase was completely masked by the
liberated haemoglobin. When, to the same defibrinated blood after
standing in the cold for twenty-four hours, a dose of sapotoxin
two-thirds greater was added, a distinct increase in the conduc-
tivity had already taken place in thirteen minutes. The maximum
conductivity (before laking) occurred in thirty minutes, and obvious
laking had appeared within one and one-half hours of mixture. Not-
withstanding the depressing influence of the liberated haemoglobin,
the conductivity was increased by laking (both in the specimen
kept at o° and in that kept at air temperature), as always happens
on the addition of a dose of sapotoxin more than sufficient to
just cause laking, owing to the liberation of electrolytes from the
corpuscles. ‘

In Experiment III (2), the effect of an increased dose of sapotoxin
(equal to 10 c.c. of the 2 per cent solution to 100 c.c. of blood) was
tried on a specimen of the same blood as was used in Experiment
III (1). While there is some increase of conductivity before laking,
the typical effect of the large dose is seen in the marked increase
of conductivity after laking, particularly in the mixture kept at air
temperature.

In Experiment IV, with perfectly fresh blood, and a dose of sapo-
toxin corresponding to 8 c.c. of the solution to 100 c.c. of blood,
the maximum change in conductivity was produced in thirty-two to
thirty-eight minutes, before which laking had not begun.

In Experiment V, with the same blood as in Experiment IV, but
after keeping it forty-eight hours in ice, a dose of 6 c.c. of sapo-
toxin solution to 100 c.c. of blood, produced a slight increase of
conductivity, even in five minutes after mixture, and the change
went on increasing till a maximum (without laking) was reached
in about one and one-fourth hours. For one hour and eight minutes
longer, no laking occurred, the conductivity maintaining itself at
the Jevel previously reached. The dose of sapotoxin was sufficient
to cause a distinct increase in the conductivity of the blood allowed to
lake at air temperature, in other words, it was sufficient to cause a
liberation of electrolytes which, in spite of the discharged haemoglobin,
increased the conductivity.

Influence of Cold on the Action of Hemolytic Agents. 75

The simplest explanation of these phenomena is that the sapotoxin,
when it acts at 0°, produces first such an effect on the superficial layer
(envelope) of the corpuscles, that its permeability to electrolytes is
increased. This change does not necessarily, or at any rate does not
immediately, lead to a discharge of haemoglobin. The second stage
of the action of the sapotoxin consists in the liberation of the hamo-
globin, presumably by breaking up the compound which it forms with
the stroma. This obviously will be a gradual process, since the pene-
tration of the sapotoxin will take time, especially when it is delayed
by a low temperature. If the dose of sapotoxin is minimal, it does
not appear that the process goes beyond this second stage, either at
O° or at air temperature. But if the dose is greater than that just
sufficient to cause the change in the envelope and the discharge of
the hamoglobin (with, it may be, such portion of electrolytes as ex-
ists in ordinary solution in the corpuscles), the electrolytes which are
bound to the stroma, chemically or physically, appear to be liberated,
and the conductivity of the laked blood is increased more or less
markedly in proportion to the dose. The decided increase which is
produced in the potential difference between the belly and the ten-
dinous extremity of the frog’s gastrocnemius, when sapotoxin solution
(in 0.9 per cent NaCl) is applied to the muscle next the tendon, may
be due either to a change in the permeability of the sarcolemma, or to
the liberation of electrolytes in the fibre contents, or to both. A
similar, though less marked, increase is produced by sapotoxin in the
current of rest of nerve.

It may be asked whether there is any proof that the sapotoxin is
taken up and fixed by the corpuscles during the stage when the con-
ductivity is increasing, without any laking having been produced.
Clear evidence has been obtained that this is the case.

For example (as in Experiment I), measured quantities of the mixture of
blood and sapotoxin kept at o° were removed after the increase of
conductivity had clearly developed itself, but before any laking had
occurred. They were rapidly centrifugalized in a cold room after being
shaken up with a known volume of ice-cold salt solution (usually 14 times
as much as was taken of the blood). The supernatant liquid, seen to be
free from haemoglobin, was decanted off, and the washing with ice-cold
salt solution and centrifugalizing repeated once or twice more. In this
way, all the sapotoxin not in the corpuscles was removed. The sediment
was then distributed in salt solution, and left at air temperature or placed
in a bath at 40°C., and it was seen that laking occurred. The amount of

76 G. N. Stewart.

laking was determined by centrifugalizing and estimating colorimetrically
the quantity of hemoglobin in solution.

It was found that the ultimate amount of Jaking was not markedly
less in the specimens of sapotoxin blood kept at 0° and washed free
‘from extraneous sapotoxin than in similar specimens of sapotoxin
blood kept at air temperature, and then washed free from extraneous
sapotoxin. Not only then do the corpuscles fix sapotoxin at 0°, but,
apparently, at the stage when the conductivity is increased while no
actual laking has yet taken place, they have fixed as much, or nearly
as much, as they would have done had laking been allowed to proceed
at air temperature. Formaldehyde-hardened corpuscles also fix sapo-
toxin, as can be shown by adding to a suspension of the washed cor-
puscles in salt solution a small amount of sapotoxin, allowing the
mixture to stand and then centrifugalizing. The supernatant liquid
will be found to have no hzemolytic power.

Since, as is well known, the cholesterin of the serum neutralizes the
action of a certain amount of sapotoxin and to this extent prevents it
from laking the corpuscles,! it was thought that still more striking
effects on the conductivity of the corpuscles would be obtained if a
suspension of corpuscles washed free from serum constituents were
employed. Experiments VI and VII were performed in order to test
this, but to my surprise, I could obtain no distinct evidence that the
phenomenon in question is to be observed with washed corpuscles at
all. The observations are rendered more difficult by the great sensi-
tiveness of washed corpuscles to sapotoxin. Thus in Experiment VI
a suspension containing about 75 per cent of corpuscles (by volume)
was laked almost immediately at 0° by addition of sapotoxin in the
proportion of 2 c.c. of the 2 per cent solution to roo c.c. of the sus-
pension, and so violent was the action, that no intact red corpuscles
could be discovered by the microscope, although swollen leucocytes
were plentiful. In accordance with the size of the dose the conduc-
tivity of the laked blood was greatly increased. Partial laking oc-
curred very rapidly when sapotoxin was added in the proportion of
1.4 c.c. of the 2 per cent solution to 100 c.c. of the suspension.
Even the addition of sapotoxin in an amount corresponding to 0.6
c.c. of the 2 per cent solution to 100 c.c. of the suspension caused

* Cholesterin suspended in saline solution removes sapotoxin completely from
solution, and prevents the laking of blood added to the mixture. After filtering
off the cholesterin, the filtrate has no laking action.

Influence of Cold on the Action of Hemolytic Agents. 77

such a rapid laking that the decline of conductivity associated with
the escape of the hemoglobin was proceeding three to five minutes
after mixture. That the dose was by no means a large one (as tested
by its effect in liberating electrolytes from the corpuscles) was shown
by the fact that, instead of any increase in the conductivity being
produced as laking went on, a marked and progressive diminution
occurred. It was shown by the method already described that the
corpuscles had fixed some sapotoxin, which produced a further laking
after all extraneous sapotoxin was removed.

In Experiment VII, in which a minimal dose was much more care-
fully sought for than in Experiment VI, the attempt to discover an
increase of conductivity preceding the liberation of haemoglobin from
washed corpuscles was unsuccessful, although the corpuscles had
fixed enough sapotoxin to cause a considerable amount of further
laking. It could not be determined whether any sapotoxin was fixed
before laking began, since the interval between addition of the sapo-
toxin and the commencement of laking was so short even at 0°. On
the other hand, it was quite clearly shown that in the case of the
washed corpuscles, as well as of the unwashed, the laking process is
much retarded by the low temperature.

If in reality the preliminary stage of increased conductivity without
liberation of the hemoglobin is not passed through in the laking of
washed corpuscles, two explanations suggest themselves: (1) that
the preliminary increase of permeability of the corpuscles in entire
blood is produced not by the sapotoxin itself but by a compound of
the sapotoxin with some constituent of the serum, the cholesterin,
é. g., which compound, less violent in its action than the sapotoxin,
may affect the envelope of the corpuscles without liberating the
hemoglobin; (2) that the ions to which the corpuscles become more
permeable do not include Na and Cl. The former would appear the
more probable hypothesis and I hope that further experiments will
settle the point. A possible diminution in the internal friction
produced by some action of the sapotoxin on the serum, which, of
course, would cause an increase in the velocity of the ions and there-
fore an increase in the conductivity, is excluded by the fact previously
shown ! that the addition of sapotoxin to serum does not increase its
conductivity. Further, the gradual increase in conductivity of the
sapotoxin blood at 0° is against this idea.

1 STEWART: Journal of physiology, 1901, xxvi, p. 484.

78

G. N. Stewart.

This experiment shows well, as several of the others do, that with minimal

The

doses of sapotoxin there is for a long time, even when laking is allowed
to go on at ordinary temperature, little or no liberation of electrolytes
from the corpuscles. Instead of an increase of conductivity there is
a diminution due to the depressing influence of the discharged hamo-
globin, and approximately proportional to the amount of blood-pigment
set free. Thus in A! (Experiment VII) when 57 per cent of the total
hemoglobin was in solution A was 40.50, while for the “‘ control” A was
47.92. If we take the proportion of hamoglobin in the corpuscles as
4o per cent, and the proportion of corpuscles in the suspension (calculated
from the conductivity), as 36 per cent, the hemoglobin in A’, if it were
all liberated, would make up a solution containing 13.8 per cent of blood
pigment. With 57 per cent of the hemoglobin discharged into a volume
of “serum” constituting 64 per cent of the suspension and about 65.4
per cent of A’, the serum would contain about 12.2 per cent of hamo-
globin. But the serum is, of course, increased by the escape of water
from the laked corpuscles. On the assumption that 57 per cent of the
corpuscles has been completely laked (an assumption not perfectly
correct, since partial laking of some corpuscles must have taken place),
the amount of serum would be about 77 per cent, and the proportion of
hemoglobin dissolved in the serum about 10.3 per cent. I have shown ?
that when one gram of oxyhemoglobin is dissolved in 100 c.c. of serum,
the conductivity is depressed by 1.82 per cent. If the conductivity of
the control (in Experiment VII) be diminished by 1.8 X 10.3, that is,
18.5 per cent, we get 39.1, which is not very different from the actually
observed X of A! (40.5). The liberation of a small amount of electrolytes
from the laked corpuscles and the increased permeability of the unlaked
corpuscles would easily account for the slight difference.

addition of a further amount of sapotoxin to A’ caused a very great
increase in A. In previous researches I did not study the effect of really
minimal doses of sapotoxin. My former conclusion, that sapotoxin causes
the liberation of electrolytes in considerable amount from the corpuscles

during laking, must therefore be supplemented by the statement that this

is the case only when doses greater than those just sufficient to produce
laking are employed. After a minimal dose of sapotoxin has caused the
hberation of the hamoglobin from the corpuscles, unaccompanied by any
large proportion of the electrolytes of the latter, the addition of a further
quantity of the poison brings about a marked discharge of electrolytes
from the stromata, just as the addition of water or sapotoxin to the ghosts
of heat-laked blood causes the liberation of electrolytes.

* STEWART: Journal of physiology, 1899, xxiv, p. 216.

L[nfiuence of Cold on the Action of Hemolytic Agents. 79

Since, as previously shown, the conductivity of formaldehyde-fixed
corpuscles is increased by sapotoxin at air temperature, it became
of interest to determine whether cooling to 0° abolishes this effect.
Experiment VIII shows that the increase of conductivity is practi-
cally as great as at air temperature. Apparently a low temperature
does not much, if at all, retard the action, at least for the dose
employed.

This result is suggestive as regards the portion of the formaldehyde-
fixed corpuscle affected by the sapotoxin. If the sapotoxin had to
penetrate to the interior of the corpuscle in order to produce the
change of conductivity, the low temperature might have been expected
to markedly retard the effect. The fact that the retardation is not
striking corroborates the view taken in previous papers that it is the
surface layer (envelope) of the formaldehyde corpuscle which is
attacked by it. I take the opportunity of mentioning here certain
results obtained by Dr. C. C. Guthrie in my laboratory in an investi-
gation, which is still proceeding, on the influence of formaldehyde in
various doses on blood, and especially on the action of laking agents.
. He finds that if blood (dog’s or rabbit’s) be drawn into an equal vol-
ume of 1 per cent. formaldehyde solution in 0.9 per cent NaCl solu-
tion, the corpuscles can be laked (by water or sapotoxin) after many
hours. Even after forty-eight hours, a mixture of one part of blood
with two parts of 1 per cent formaldehyde solution (which, like
mixtures containing still larger amounts of formaldehyde, remained
unclotted) was laked by water and more slowly and partially by
sapotoxin. In most specimens of dog’s blood, clotting is completely
prevented when the formaldehyde present amounts to one volume of
the I per cent solution to 0.6 volume of blood. In some specimens
somewhat more is needed, in‘others less. Much smaller quantities of
formaldehyde (ce. g., two to twenty parts of blood to one part of I per
cent formaldehyde) retard coagulation in proportion to the quantity of
formaldehyde present, and although they do not prevent its ultimate
appearance, they exert a marked influence on the process, as is shown
by the slow and imperfect shrinking of the clot as compared with that
in control specimens of normal blood. One of the ways in which
formaldehyde prevents or retards coagulation seems to be by hinder-
ing the development of the fibrin ferment (perhaps by a rapid partial
fixing of the leucocytes or blood-plates), and this appears to be a
more powerful action than its inhibition of fibrin ferment already
formed. The haemolytic action of dog’s serum on rabbit’s corpuscles

80 G. N. Stewart.

is hindered by formaldehyde in proportion to its concentration and
the time it has acted. This is true whether the formaldehyde is
added to the dog’s serum or to the rabbit’s blood, and indeed the
addition of a given amount to the dog’s serum causes a greater
diminution in the hamolytic activity than the addition of the same
amount to the rabbit’s blood, when the formaldehyde is allowed to
act for the same period before mixing the blood and foreign serum.
The laking of the blood by such haemolytic agents as have been
investigated is not essentially altered, even after a considerable.
period, by such quantities of formaldehyde as are necessary to prevent
putrefaction.

It might be expected that other hemolytic agents than sapotoxin
would produce a similar effect on the permeability of corpuscles as a
preliminary to laking. I investigated bile salts (Na taurocholate) as
a representative of the other chemical lakers, and foreign serum as a
representative of the biological lakers. As is shown in Experiment
IX, while there appeared to be a slight increase of conductivity pro-
duced by the bile salt in the case of unwashed corpuscles, it is not so
pronounced as when sapotoxin is added. In this connection it may
be recalled that bile salts, while they increase the conductivity of for-
maldehyde-fixed corpuscles, do not cause so great an increase as sapo-
toxin! A further fact, presumably related to this difference, is that
in sapotoxin laking the corpuscles always swell before discharge of the
haemoglobin, becoming smooth in outline if they have previously been
crenated, and therefore take up water, while in sodium taurocholate
laking they need not do so. In Experiment X, an attempt was made
to determine whether the conductivity in a suspension of washed
corpuscles was increased by sodium taurocholate before laking, but
with the same negative result as in the case of sapotoxin. The
sensitiveness of the corpuscles to the bile salt is increased by the
removal of the serum constituents just as in the case of sapotoxin.

The fact that some of the biological haemolytic agents, e. g., snake
venom, according to Flexner and Noguchi? are unable to cause laking
in washed corpuscles, while in the absence of serum constituents
sapotoxin and bile salts are apparently unable to produce the prelimi-
nary increase of permeability of the corpuscles, might seem to sug-
gest that such biological lakers may act primarily by augmenting
the permeability of the envelope; and that just as in the absence

‘ STEWART: Journal of medical research, viii, p. 268.
* FLEXNER and NoGucui: Journal of experimental medicine, 1902, vi, p. 286.

Lnfluence of Cold on the Action of Hemolytic Agents, 81

of the complement of the serum, the venom is unable to complete
the reaction on which this increase of permeability depends, so, in
the case of these chemical lakers, something which can act as a
complement or as an intermediary body, to use Ehrlich’s termin-
ology, is necessary to produce the preliminary increase of conduc-
tivity, although ultimately laking can take place without it. In
Experiment XI, however, no evidence was obtained that dog’s serum
can produce any change in the conductivity of rabbit’s blood before
liberation of the haemoglobin. This is in agreement with the fact
that the permeability of the formaldehyde-fixed corpuscles of the
rabbit is not affected by dog’s serum. It was observed that with
foreign serum complete laking will take place at o°, although the
statement has been made that it does not occur below about 5° C.

82 G. NV. Stewart.

EXPERI-

February 8.— At 12.20 p.m. added to 25 c.c. of dog’s defibrinated blood obtained
nineteen hours before and kept in cold 1.5 c.c. of 2 per cent sapotoxin solution in 0.9 per
cent NaCl solution, both ice-cold. Call the mixture A. Kept in ice. At 11.55 a.m.
added to 25 c.c. of the defibrinated blood 1.5 c.c. of 0.9 per cent NaCl solution. Call the

P.M rt
12.38 Control cooled in ice before putting in U tube . ; 34.73
12.403 DR SUR ne aaa ea ene ae ‘ ; 34.63
12:25 BN ; 34.96
ZY, ; : Sante
2129 , 34.73
12.53 ¥ 5 34.26
12.56 SORE hia acer PT ge : 34.26
lls Absolutely no laking 2 : (342
34.77
134.77
2.45 : 34.73
. 34.73
4.48 Absolutely no laking.? (Sediment 3) . . 35.36
4.50 Pere AL Ue ergs rs, Mey ea er 5 35.26
Feb. 9
A.M.
9.383 37.06
oil RAR aO UME SS Rar Rue THATS A cen, ot Recent a ute At Io cae shee 36.90
9143 Very little laking. Not more than 1.8 per cent of the hamo- :
globin in solution.2 (Sediment 4) ohvd, eth amen { 36.95
10.01 Control a BF Sollee ve, To ett Sa ee ance Sheen
10.05 No laking whatever "(Sediment 5) } 34.49
P.M.
3.17 Serum from A (after centrifugalizing for three hours contains
some hemoglobin) . . 83.66
Serum from control (after centr ifugalizing for three hours ¢ con-
tains a little haemoglobin, but less than in serum from A) 82.57
3.46 Sediment fromvA. yun. a) oe ste Geeks ee ee 6.98
SAE IS GORA: SLE nd ON) | Conn tee ci ee ar 3
6.91
4.07 Sediment fromlcontroly 2-9 9es epee te ere 6.81
4.09 5 HD. SO bese a= eaters aeeOnahR 1 a 6.69
4.103 6.55

! As in all the tables,A is an abbreviation for X (0°) 108, the conductivity (at
° C.) being expressed in reciprocal ohms.
2 By centrifugalization test, 1 c.c. of the blood being mixed with ee c. of ice-cold
NaC] solution, and centrifugalized. “Sediment 3,” etc., means that the sediment
in the centrifuge tube was so numbered and set aside.

L[nfiuence of Cold on the Action of Hemolytic Agents. 83

MENT I.

mixture “control.” At 3.12 p.m. added to 25 c.c. of defibrinated blood 1.5 c.c. of 2 per
cent sapotoxin solution, both ice-cold. Call the mixture A’. Kept at room temperature.
For the sapotoxin solution A = 90.22; for the NaCl solution A = 88.95; for the defibri-
nated blood A = 32.09; for the serum from the clot, \ = 73.05 (Serum contains much fat).

3.20 1A eae hah <8 Pe RMN T Fae Cx oo de oneal an an a tae ee 34.96
Soh To eye no laking whatever . AE te Sg ee PS. Sa 34.77
3.59 Pong ety ker 36.74
4.02 36.74
4.03 Ee a arc ce ar eas |i Sb as eee Sas 36.63
4.06 Practically no haemoglobin in solution, certainly less than
ippericenteof the wholes (Sediments). 2 9s. = 2. 36.63
4.30 ERIS Eerie ta Lik men Glp time hae arse gt hats ee Ge te Wt 36.79
4.31 3.5 per cent of hamoglobin is now in solution.! (Sediment 2) | 36.79
Feb. 9
A. M.
10.305 ee RSP Sa eS CAR gh a See ae oe gi een ae ae ee an ear eS ee ( 33.89
10.32 ee ROR ROA A oe HE PCON NS dP cies ae aE acre eae 33.80
10.345 54 per cent of the hemoglobin in solution. (Sediment 6) . (33.71
Feb. 10 5
P.M.
+590 | } 34.54
4255 ME RCE M a reine g She Crsieee | Suara ome tg ae toe ee ee eee 34.35
Even after twenty-four hours longer A’ is not completely laked.!

February 10.

. 2P.M. Sediments 1 and 2 shaken up in NaCl solution and centrifugalized.
Considerable laking. In 1, 41 per cent of hemoglobin has gone into solution since
sediment first obtained, and in 2 about 50 per cent. Some laking in 3, thongh not
so much as in 2 (23 per cent of the hemoglobin is in solution). In 4 practically
no laking. (All kept at air temperature since first sedimentation.)

February 11.

2P.M. In ] and 2 laking has made much further progress, and only a small
residue of unlaked corpuscles is now left, about the same in] asin 2. The same
is true of 3, in which, however, the residue of unlaked corpuscles is a little greater
than in | and 2.

The residue in 3 is not quite 0.1 c.c. In 4 laking has also advanced markedly,
and residue is little greater than in 3. In 5 laking has advanced very little, and
the residue of unlaked corpuscles amounts to 0.5 c.c.

In 6 laking has hardly advanced any further than yesterday. The amount of
sediment in 6 is fully as great as in 4, being a little over 0.1 c.c. All these were
kept at air temperature (19°).

A’ is not yet completely laked, as shown by centrifugalizing.

Control kept at air temperature is red, and shows no laking to eye.

1 By centrifugalization test, 1 c.c. of the blood being mixed with 14 c.c. of ice-
cold NaCl solution, and centrifugalized.

84 G. N. Stewart.

EXPERIMENT II.

October 21, 1902. — At 4.12 p.m. added to 50 c.c. dog’s defibrinated blood drawn
twenty-five hours ago and kept on ice, 3 c.c. of 2 per cent sapotoxin solution (in NaCl
solution), the blood and solution having been previously cooled to 1.5° C. Call the
mixture A. Kept in ice.

Another sample of A .

(A ‘specimen of A centrifugalized after the addition of FeCl!
shewed no laking)

Another sample of A .

A specimen of A centrifugalized after the addition of FeCls
showed distinct laking, about 40 per cent of the haemoglobin
having been liberated from the corpuscles, as shown colori-
metrically. Twenty-two hours later A, which was kept in
ice all through, was seen to be completely laked.

Added to 5 c.c. of blood 0.3 c.c. of 0.9 Bers cent NaCl.
This mixture .

Added to 5 c.c. of blood 0.3 c.c. of sapotoxin solution and left
at room temperature.

This mixture is much laked, though not completely.

RM eA On cr Ma ete ye een Lee TE Pa 32.92

32.892

1 FeCl; was added in small amount to a measured quantity of the blood, accord-
ing to the method of precipitating red corpuscles described by Dr. S. PESKIND.

Then the mixture was rapidly shaken up with excess of ice-cold 0.9 per cent salt
solution, and centrifugalized.

Influence of Cold on the Action of Hemolytic Agents. 85

EXPERIMENT III (1).

November 22, 1902.— At 4p. mM. added to 25 c.c. of dog’s defibrinated blood drawn
34 hours ago, and kept in ice, 1.5 c.c. of 2 per cent sapotoxin solution in 0.9 per cent
NaCl solution, also ice-cold. Call mixture A. At 3.55 p.m. added to 10 c.c. of blood
0.6 c.c. of the sapotoxin solution. Call mixture A’. A was kept in ice, A’ at air tem-
perature. At 4 p.m added to 10 cc. of blood 0.6 cc. of 0.9 per cent NaCl solution.
Call the mixture “control.” Kept in ice.

For defibrinated’ blood’ ><", =~ s+ - AN=2840.
For sapotoxin solution . . Stee ete N= 85:98:
For the sodium chloride eattiack; a ion) Sate Oa 4k

A’ (30 per cent laked) 1

. 50 per cent laked 1
‘Absolutely. no Taking
Control. 4 :

WD WWW HWW Wo
Or AS Seco

Is not yet completely

laked. Apparently

| about the same

aN ps2 OA Se Re Nae Oo ! amount of laking as
INovaking ssc 2 =: aly yesterday.

|| Twenty-four hrs. later

= siete Daa te a ‘ (greater part of ha-

Woplaking+ 5 chee o>. —- moglobin is now in
solution)

About 15 per cent of
the hzmoglobin is
now in soiution.}

A was now placed at
room temperature
for twenty-four hrs.,
at end of which time
A= 29 61, the greater
part of the haemo-
globin being now in
solution.

i

1 Determined by addition of FeCl3 and cold NaCl solution and centrifugalizing.

86 G. NV. Stewart.

EXPERIMENT III (2).

November 23.— At 11.35 a.m. added to 25 c.c. of same defibrinated blood 2.5 c.c. of
sapotoxin solution, both ice-cold. Call mixture A. At 12.25 added to 10 c.c. of blood
1 c.c. of sapotoxin solution. Call mixture A’. Kept A in ice, A’ at air temperature.
At 12.30 added to 10 c.c. of blood 1 c.c. of 0.9 per cent NaC] solution. Call the mixture
“control.” Kept in ice.

11.48 Pe eet Rn Oe A! completely laked.
PSS sle 4 2.5%) 8) an eee nea
TESS
11.573
P.M.
TRIOS tN eset Mas Rn ee eae ae ee
12.08 | Nolaking! —..:

1.05 Now there is obvious
laking.

1.08 Sih ce ©) Aaa aiee ee
1.10 | Most of the corpuscles
are laked“:— 2 =. = 38.53
12.50 “Controle sete et 33.22

1 Determined by addition of FeCls and cold NaCl solution and centrifugalizing.

Influence of Cold on the Action of Hemolytic Agents. 87

EXPERIMENT IV.

October 16, 1902. — At 4.25 p.m. added to 5 c.c. of dog’s defibrinated blood drawn an
hour ago 0.4 c.c. of a 2 per cent sapotoxin solution in 0.9 per cent NaCl solution, both
cooled to 0° C. Kept the mixture in ice and made conductivity measurements at 0°.

For the sapotoxin solution . . ... A=88.95.
For the 0.9 per cent NaCl solution . A= 88.64.

Miaxtureron blood: andssapotoxin Pos.) wy) itntspetecisist iss tates ok { 28.81
Seeley othe a i sie ok CON RE Baar aie Ye ee: hg Soe end em nee | 29.82
+ 29.96
| 29.68
{ 29.58
Wee

Shook up U tube

33.66

5 c.c. blood + 0.4 c.c. of 0.9 per cent NaCl pines ns -four 29.34 -
hours ago and kept on ice). é 29.27

The original defibrinated blood, kept ir Tt TEI Cs omen mem ie are 24.61

24.37

5 c.c. defib. blood + 0.4 c.c. sapotoxin, kept at 17° C., laked completely in 5 min.
5 c.c. defib. blood + 0.3 c.c. sapotoxin, kept at 17° C., laked in 20 min.

5 c.c. defib. blood + 0.2 c.c. sapotoxin, kept at 17° C., laked in 23 hrs.

5 c.c. defib. blood + 0.3 c.c. sapotoxin, at 0° C., almost completely laked in 21 hrs.
5 c.c. defib. blood + 0.2 c.c. sapotoxin, at 0° C., only slightly laked in 2] hrs.

88 G. NV. Stewart.

EXPERIMENT V.

October 18, 1902. — At 3.33 p.m. added to 15 c.c. dog’s defibrinated blood (obtained
forty-eight hours ago and kept in ice-chest) 0.9 c.c. sapotoxin. Kept mixture in ice.
Call mixture A. For the sapotoxin solution A = 88.95; for the 0.9 per cent NaCl
A = 88.64.

tn tn Go
nADdure ©

Not laked .
Now took out U tube and left in air ten minutes, ‘then put back
in ice.

The blood is now v pretty well laked . er
Refilled U tube from stock of A kept in ice all the time since
mixture Rae tear eae eee

Not laked. . .

Took out U tube from ice, ‘heated in hand, shook up ‘and put
back in ice. Blood is now somewhat darker, though not
by any means completely laked.

Now fairly well laked

Added to 15 c.c. defibrinated bioad 091 c.C. " sapotoxin, antl left
at air temperature (18°).

Completely laked.

roo
ne AR

Wun

Wn
NWO Oo

Added to 15 c.c. defibrinated blood 0.9 c.c. of the 0.9 per cent
NaCl solution.

monn
¢

Refilled U tube from stock of A kept in ice (does not seem
laked) sn Syl Be Se eee ae

WY)

L[nfiuence of Cold on the Action of Hemolytic Agents. 89

EXPERIMENT VI.

January 11.— To 1 c.c. of dog’s corpuscles, washed thoroughly and suspended in 0.9
per cent NaCl, added 0.04 c.c. of 2 per cent sapotoxin solution in 0.9 per cent NaCl.
Immediate laking. For the suspension, A = 13.78; for the NaCl solution, A = 89.26;
percentage of corpuscles in suspension, calculated by the electrical method, 75 per cent.
At 11.13 A.M. added to 10 c.c. of the suspension 0.2 c.c. of the sapotoxin solution, both
ice-cold, and kept the mixture in ice. It begins to lake almost immediately.

MER CUTGR erat. ebm a bel LOSS
Lig eek aa eae IPO eh Gea lel oF or)
After 20 hrs. at 8-10°. . . | 31.89 | Microscopically: No intact corpuscles,
but numerous swollen leucocytes and
hemoglobin crystals.
At 11.45 added to 5 c.c. of the suspension 0.07 c.c. of the sapotoxin
solution, both ice-cold, and kept the mixture in ice.

The mixture. .
58 per cent of the hemoglo-

bin is in solution!. . . | 10.31 | Microscopically examined at once; many
hzmoglobin-containing corpuscles,
mostly round, but some crenated.
Some ghosts. On standing until
next morning, few or no unaltered
corpuscles.

At 12.40 added to 5 c.c. of the suspension 0.3 c.c. of a 0.2 per cent sapotoxin solu-

tion in 0.9 per cent NaCl solution, both ice-cold, and kept the mixture in ice.
PEt om MOVIIEUING: Ata. ce) ire tose 2. || 15/65
2S oa ee ee otra eR Omer amen (ek 0]
WZeS Ot ieks oe 12.67
12.56 | 24 per cent of haemoglobin
isin solution! . . 11.63 | Microscopically: Numerous corpuscles

containing hemoglobin, mostly round,
but some crenated, and some ghosts.
On standing until next morning, few
or no unaltered corpuscles.

1 Sediment was repeatedly washed in cold NaCl solution to remove all sapotoxin
in the solution. Evidently the corpuscles had fixed some sapotoxin, since they
laked partially on standing till next morning at air temperature, while a control
specimen of blood containing no sapotoxin did not lake. The washed sapotoxin
corpuscles did not lake so completely as corpuscles from the same mixture left still
in contact with the salt solution containing sapotoxin.

go G. NV. Stewart.

EXPERIMENT VII.

January 20. — Dog’s defibrinated blood obtained January 18 at noon. Washed twice
in 0.9 per cent NaCl solution in the cold.

1 c.c. of the suspension plus 0.1 c.c. of 0.2 per cent sapotoxin solution in 0.9 per cent
NaCl solution lakes at once at air temperature.

1 c.c. of the suspension plus 0.05 c.c. of 0.2 per cent sapotoxin solution lakes partially
in one minute at air temperature. In one hour not completely laked. Completely laked
next morning at 10 A.M.

1 c.c. of the suspension plus 0.03 c.c. of 0.2 per cent sapotoxin solution. No laking in
five minutes. In twelve minutes is a little darkened. In one hour distinctly darkened,
but not by any means completely laked. Completely laked next morning at 10 A.M.

For the suspension A = 46.86 on January 20 and 46.43 on January 21.

At 3.03 p.m. added to 20 c.c. of the suspension 0.8 c.c. of 0.2 per cent sapotoxin.
Kept at air temperature (18° to 20° C.).. Call the mixture A’. At 3.16 added to 20 c.c.
of the suspension 0.8 c.c. of 0.2 per cent sapotoxin, both ice-cold, and kept in ice. Call
the mixture A. At 4.10 added to 20 c.c. of the suspension 0.8 c.c. of 0.9 per cent NaCl,
and kept in ice. Call the mixture “control.”

For the 0.9 per cent NaCl solution A = 89.26 For the defibrinated blood A= 29.41
For the 0.2 per cent sapotoxin . . = 89.26 Serum fromiclot7)) 9. s-.0A— "Ou

L[nfluence of Cold on the Action of Hemolytic Agents. 9I

EXPERIMENT VII (continued).

P.M. d P.M. r
Bee a AS. { 48.94 3.09 | A’ is somewhat dark-
3.25 ‘ | 48.29 ened.
ESR MM MOO et Mh 'o) fie? Seencalturict oa ial a IPRA 3.30 | Very distinctly dark-
3.33 Slight laking, 8 per cent 4 ened.
of the hemoglobin in | Eighteen and one-half
solution.t (Sed.1). | | 47.21 hrs. later completely
yee an aces eee, ac ky WOOO laked.
PSO Re encore Oe ian cone ere (e+ oro BAA: NERS { 40.56
3.58 | 16 per cent of the he- | { CE tae sophie Salers @Oen ken eae 4 40.50
moglobin ‘in solu- od 57percentofhzmoglo- | |
tion! (Sediment 2) 45.51 bin in sol! (Sed. 5) | | 40.50
5.02 radar se: ats Se ¢5| (PAS 2gel anal |
5.05 Recs Noe pepatenae de ae oe ile one 3.20 | 64percent of hemoglo- ( 40.50
5.08 | 20 per cent of the he- | { bin insol (Sed.8) | -
moglobin in solu- | | 24 hours later : 1 45.19
tion.! (Sediment 4) | (45.03 || Jan. 22
Jan. 21 3.15 To 5 cc. of A’ added
DAO) Bele a a (phatase kyon oto ye 45.51 0.2 c.c. of 2 per cent
2.35 | 43 per cent of the he- sapotoxin. Lakes
moglobin in solu: completely. In sedi-
tion! (Sediment 6) | | 45.43 ment many leuco-
Jan. 20 cytes and granules,
4.40 | Control 48.01 but no ghosts.
aut oath ak CM DEE Ses ar ATOZ \ 3.332 | >This mixtures)... >. 66.01
4.46 No laking.t (Sed. 3) 47 83 3153 A’ plus as much of 0.9
Jan. 21 per cent NaCl as was
2.59 Slight laking.! (Sed.7) 46.43 added of the 2 per
cent sapotoxin sol. . 47.83

Microscopically, A’; numerous ghosts perfectly round. Some corpuscles still con-
tain some hemoglobin, although all are altered and swollen. None are crenated.

1 Tested by centrifugalization method. “Sediment 1,” etc., means that the sed-
iment in the centrifuge tube was so numbered and set aside. At 5.30 P.M., on
January 20, sediments 1, 2, 3, 4, and 5 were washed twice in ice-cold salt solution
(each time in 16 c.c.) and then left at room temperature. Considerable laking took
place in all except 3.

January 21 at 4.20 p.M. heated sediments 1, 2, 3, 4, 5, 6, 7, and 8 to 45° C. (after
washing twice in NaCl solution), then made up to 15 c.c. with NaCl solution and
centrifugalized.

In 3 and 7 some laking has taken place, but much less than in the others. In 1,
sediment is somewhat greater than in 2; in 2, two or three times as great as in 4;
in 4 and 5, about the same, the supernatant liquid in 5 being less deeply tinged than
in 4; in 6, sediment is not half as great as in 5, and about equal to that in 8, the
supernatant liquid in 8 being less deeply tinted than in 6,

92 G. N. Stewart.

EXPERIMENT VIII.

October 27, 1902. — Made a suspension of washed formaldehyde-fixed dog’s corpus-
cles in 0.9 per cent NaCl. Call it A.

For the 0.9 per cent NaCl solution . . . A=87.11.
For the sapotoxin solution . . . . . . AX=85.64.

A.

Added to 10 c.c. of A 0.6 c.c. of sapotoxin solution. Kept at air
temperature.
This mixture .

Refilled U tube .
Added to 10 c.c. of A cooled to 0° C., 0.6 c.c. of ice-cold sapo-

toxin solution. Kept the mixture in ice.
This mixture ; ee:

Refilled U tube .

Added to 5 c.c. of A 0.3 c.c. of 09 per cent NaCl. Kept mix-
ture at air temperature.

Lnfluence of Cold on the Action of Hemolytic Agents. 93

EXPERIMENT Ix.

December 16, 1902. At 2.52 p.m. added to 10 c.c. of dog’s defibrinated blood ob-
tained 24 hours ago and kept in ice, 2 c.c. of 2 per cent Na taurocholate in 0.9 per cent
NaCl, both ice-cold. Call the mixture A. Ais kept inice. At 3.40 added to 10 c.c. of
the defibrinated blood, 2 c.c. of the taurocholate solution and kept at room temperature.
Call the mixture A’. At 2.58 added to 10 c.c. of the defibrinated blood, 2 c.c. of 0.9 per
cent NaCl. Call the mixture “control.” Kept in ice. To 10c.c. of the blood added 2
c.c. of a NaCl solution whose A was 97.12. Call the mixture “second control.”

Bor detibtinated blood ss c28 6 2 eee = 295.
Horthe serum trom tnerclot aa. #1. ok ck. os ==. 7 Ob.
For the taurocholate solution. . . . . A=96.75.
For the 0.9 per cent NaCl solution . . . A=91.20.

DTS Ae a Co aan : A’ (partially laked).
te tee Se 28 per cent of he-
Bg a Sores PP Te moglobin being in
A is now slightly solution.}
laked, 7.7 per cent
of the hamoglo-
bin being in solu- A’ is now com-
tion.t pletely laked.
A’ (shows no sedi-
Serum of A (sepa- ment) .
rated by centrifug. : :
and _ containing
more hamoglo-
bin than last night)
Sediment of A :
INS BS ee a

Control . sue
Serum from control
(contains trace of
hemoglobin in so-
TOETO Mi a) oie cote = Wiad
Second control . . 39.42

1 Estimated by diluting with ice-cold NaCl solution and centrifugalizing.

94 G. N. Stewart.

EXPERIMENT X.

December 20. 9.35 A.M. To 1 c.c. of a sediment of dog’s corpuscles washed free
from serum by NaCl solution, added 0.1 c.c. of 2 per cent Na taurocholate solution in 0.9
percent NaCl, and kept at air temperature. In one hour partial laking has occurred. At
2 P.M. it is not yet completely laked. To 1 c.c. of the sediment added 0.2 c.c. of tauro-
cholate solution. It begins to darken in a few minutes. In an hour it is much darkened.
At 2 p.M. well laked. At 10.40 a.m. added to 10 c.c. of the sediment 1 c.c. of the tauro-
cholate solution, both ice-cold, and kept in ice. Call the mixture A. At 1] a.m. added
to 4 c.c. of the sediment 0.4 c.c. of a NaCl solution, whose A = 96.39. Kept in ice. Call
this mixture “control.”

For the’sediment, “4 Gus. ce ee = LO
For the taurocholate solution . . A=97.12

AC orp te ed tetae.: : Control

Partially laked; 40%
of the hzmoglo-
bin is in solution

Took A out of ice
and left at room
temperature.

Much darkened,
though not com-
pletely laked.

70% of the haemo-
globin is now in
solution .

Lnfluence of Cold on the Action of FHlemolytic Agents. 95

EXPERIMENT XI.

November 24. — Added to rabbit’s blood, obtained twenty-four hours previously and
kept in ice, an equal volume of dog’s serum obtained from clot drawn sixty-nine hours
before, and kept in ice. Complete laking at room temperature in three and one-half
hours. (No doubt it was laked before this.) A control with ten parts of dog’s serum
(heated to 56°C. for forty-five minutes, and to 60° C. for five minutes) added to 1 part of
rabbit’s blood, showed no laking in three and one-half hours.

For rabbits blood: aay. <cfs ©. fa sons). (NES 44.08
Hor unheated dog/s'serum 7.2 5. <3 N= 7659

At 3.50 p. M. added to 5 c.c. rabbit’s blood 5 c.c. of heated dog’s serum. Call mixture
A’. Keptinice. At 3:55 added to 15 c.c. rabbit’s blood 15 c.c. of unheated dog’s serum
(both previously cooled to 0°). Call mixture A.

Some laking . . . 58.: - No laking.

96 G. XN. Stewart.

SUMMARY.

1. At o° C the laking action of sapotoxin is much retarded, and it
can be shown that before any hemoglobin has been liberated the con-
ductivity of the blood is increased, presumably owing to an increase
in the permeability of the envelopes of the corpuscles to electrolytes.
At this stage sapotoxin has been fixed by the corpuscles. Bile salts
produce the phenomenon less distinctly, foreign serum not at all.

2. Both at o° and at ordinary temperature a dose of sapotoxin just
sufficient to cause liberation of the haemoglobin causes the discharge
of only a small proportion of the electrolytes of the corpuscles. When
the dose is increased the electrolytes are discharged.

3. It seems permissible to divide the action of sapotoxin on the
corpuscles into three stages: (1) an action on the envelope, which
does not necessarily nor immediately lead to the liberation of the
haemoglobin; (2) an action on the hamoglobin or the stroma which
causes the discharge of the pigment; (3) an action on the stroma
leading to the setting free of electrolytes.

QUICK METHODS FOR CRYSTALLIZING OXYHA:-
MOGLOBIN: INHIBITORY AND ACCELERATOR
PHENOMENA, ETC.: CHANGES IN THE
FORM OF CRYSTALLIZATION.

By EDWARD T. REICHERT.

I. Quick METHODS FOR CRYSTALLIZING OXYHAZMOGLOBIN.

HE most expeditious known method for preparing crystals of
the oxyhzemoglobin of the blood of the dog is to lake the blood
with ether, and subject it to cold. Crystals appear within a variable
period, ranging from a few minutes to many hours. I have found
that if to the blood there be added, before or after laking, from 1 to
5 per cent of ammonium oxalate, crystallization invariably begins
immediately, and that any quantity of crystals can be obtained within
a few hours at ordinary room temperature. If a drop of this blood
be placed under the microscope, crystals will be seen to form at once
near the margin of the drop, and to be deposited so rapidly that a
solid mass is formed in a few minutes.

The blood of the horse does not yield so readily to this treatment.
If a drop of blood so prepared be examined under the microscope, it
will be found that crystallization will not begin until after from fifteen
to forty minutes or more, and that it will proceed slowly. Better
results can be obtained if the blood be oxalated and set aside for the
corpuscles to subside, The supernatant liquid is then poured off,
and the remainder laked with ether.

Defibrinated blood of the rat, laked with water on a slide, and
covered with a cover-glass after the margin of the drop has become
dried, usually crystallizes very readily, as is well-known. Better
results can be obtained if the blood be oxalated before or after laking ;
and even more rapid crystallization occurs if the blood be laked with
ether instead of water, Crystals form so rapidly in the oxalate-ether
blood that a magma is formed in the test-tube within a few minutes.

The oxalate-ether process applied to the blood of the guinea-pig
gives most satisfactory results. Crystallization does not proceed

97

98 Edward T. Reichert.

quite so rapidly as in rat’s blood, yet within a minute or two innu-
merable tetrahedra appear, and any quantity can be obtained within
a couple of hours.

The blood of the necturus crystallizes readily when so treated,
which is a matter of especial interest because of the difficulty hereto-
fore experienced in obtaining crystals of oxyhamoglobin from the
blood of cold-blooded animals. The crystals resemble those of the
triple phosphates.

The rapidity with which crystallization begins and proceeds is in-
fluenced decidedly both by the method of laking and the percentage
of oxalate. Ethyl ether is a much better laking agent than water,
and acetic ether is stronger than ethyl ether. The presence of any
quantity of oxalate up to saturation increases crystallizability, but
I have thus far found from 1 to 5 per cent to be the best; the larger
the quantity the more is crystallization hastened. When more than
5 per cent is used, the oxalate also tends to crystallize upon the slide.
If the blood be prevented from drying, as in the test-tube, the oxalate
remains in solution.

Asphyxial blood yields crystals more readily than normal blood.

Il. INHIBITORY AND ACCELERATOR PHENOMENA, ETC.

If to the blood of one species, the blood, plasma, or serum of another
species be added, the laking of the blood may be retarded, accelerated,
or unaffected, according to the character of the mixture. The period
required for laking may be prolonged for five minutes or more.

The crystallization of the oxyhazmoglobin may be hindered or pre-
vented in such mixtures. Thus, by varying the proportions of a
mixture of the bloods of the dog and guinea-pig, crystals from one or
both may appear, but the process is invariably retarded, and some-
times to a marked degree. If crystals of both kinds of oxyhamo-
globin are deposited, those of one usually begin forming some time
before those of the other, and the crystallization of one may be
complete before crystals of the other are seen.

III. CHANGES IN THE ForRM OF CRYSTALLIZATION.

The typical forms of the crystals of certain kinds of oxyhemo-
globin may be modified or completely changed when the bloods of two
species are mixed. Thus, if to the blood of the rat there be added a
definite percentage of the blood of the guinea-pig, crystals of the rat’s

Quick Methods for Crystallizing Oxyhemoglobin. 99

oxyhzemoglobin may appear in unaltered form, but most, if not all, of
those from the guinea-pig’s blood will be changed; in fact, if any
perfect tetrahedra are found, they will have been formed at the very
end of the crystallization. If the proportions of the mixture be
properly modified, not a single crystal of what can be identified as
rat’s oxyhzmoglobin will appear, and all of the crystals will be modi-
fied tetrahedra and sfzuzd/es, and transitional forms between these.
The spindles resemble Charcot’s crystals in form but not in color ;
they vary in size, some of them being very large, and some may have
small spindles attached to them; and they can be obtained having
sharp angles, if crystallization has not been too rapid.

This complete change in the form of the crystals of oxyhzemoglobin,
when the bloods of two species are mixed, and the spindle-shaped
form of the crystals, are, I believe, unique facts in the crystallography
of this most important substance.

ARTIFICIAL. PARTHENOGENESIS IN NEKEES

By MARGIN: GE) Bills CEE Re
[From the Physiological Laboratory of the University of California. |

| eae year Dr. Loeb and I! published a note in this journal in

which we stated that it is possible to cause the unfertilized
eggs of Nereis limbata to develop into swimming larve by keeping
them for some time in sea-water, the concentration of which has been
raised by the addition of potassium chloride, and then returning them
to ordinary sea-water. We expressed the opinion that we were prob-
ably dealing with osmotic effects, and not with the specific effects of
K-ions, since we obtained swimming larve only from those mixtures
of sea-water and potassium chloride in which the osmotic pressure
had been considerably raised. We based our conclusions upon a
series of experiments on the eggs of a single female, but did not
publish them, as we were not able to obtain any more mature females
upon which to corroborate our findings. This year we were more
fortunate, however, and succeeded both in repeating our experiments
of last year, and of obtaining further proof of the correctness of our
conclusions.

All the precautions necessary to prevent the infection of the eggs
with sperm, which have been so often described by Loeb,? were fol-
lowed in these experiments. Each female Nereis, as soon as caught,
was kept in a separate dish of sea-water sterilized by heating to
80°C. After swimming about in this for some time, the worms were
washed in several changes of sea-water, then in fresh water, and
finally were opened in sterile sea-water with sterilized instruments.

I shall give first of all a detailed description of the series of experi-
ments which formed the subject of our first communication.

’ FiscHER, M. H.: This journal, 1902, vii, p. 313; LOEB, FISCHER, and
NeEILson: Archiv fiir die gesammte Physiologie, got, Ixxxvii, p. 594.
* Lor, J.: This journal, 1go1, iv, p. 424.
100

Artificial Parthenogenests in Nerevs. 101

Experiment I, July 20, 1901.— 12.30 a.M. The eggs of two females were
gently shaken together and distributed into the following mixtures of
potassium. chloride and sterile sea-water : —

95 c.c. sea-water+ 5 cc. KCl] 24m.
921 c.c. sea-water + 74 .c. KCl 25m.
90 c.c. sea-water + 10 c.c. KCI 25 m.
873 c.c. sea-water + 12} c.c. KCl] 23 m.
85 c.c. sea-water +15 c.c. KCl] 2h m.
823 c.c. sea-water + 174 c.c. KCl] 23 m.
80 c.c. sea-water + 20 c.c. KCl] 2im.
774 c.c. sea-water + 223 c.c. KCl] 24m.
75 c.c. sea-water + 25 c.c. KCl 23m.
10. 722 c.c. sea-water + 274 c.c. KCl] 23 m.
1l. 70 c.c. sea-water + 30 c.c. KCl] 23m.
12. Sterile sea-water (control).

CONN Wa

°

The eggs were taken out of the solutions and transferred to sterile sea-
water at intervals of thirty, forty-five, and sixty minutes. These will be
designated as Lots I, II, and III.

The first evidences of segmentation were found in the eggs exposed for
thirty minutes to the mixture of 80 c.c. sea-water + 20 C.c. KCl 23m.
One hour and twenty minutes after returning the eggs to sea-water several
were found in the two-cell stage. Ten minutes later an egg was found in
the four-cell stage. After two hours, with rare exceptions, all the dishes
contained segmenting eggs. In some of the dishes the eggs did not pass
through the two and four-celled stages, but broke up immediately into
eight and twelve cells. Up to five hours after the beginning of the
experiment, it was difficult to decide which salt solution would give the
most favorable results. After ten and a half hours (10.55 A. M.), however,
it was clearly noticeable that the optimum sea-water-salt solution mixture
lay between 774 c.c. sea-water + 223 c.c. KC] 24m and 85 c.c. sea-water
+15 c.c. KCl 24m. All these solutions contained eggs in the thirty-
two to sixty-four cell stages.

The first swimming larvee were found in the afternoon, fourteen hours
after the return of the eggs to sea-water, in Lot III of the mixture of 823
c.c. sea-water + 174.c.c. KCl2$m. Lots I] and I showed the same con-
dition of affairs. In one field of the latter every egg was in the blastula
stage, and many were beginning to swim.

At four o’clock, Lot I of the mixture of 80 c.c. sea-water + 20 C.c.
KCl 23m showed even better results than the foregoing, the dish teeming
with swimming larve. In Lots II and III of the same mixture many less
were found. Lots II and III of 773 c.c. sea-water + 22} c.c. KCl 23m
showed some swimming larve. The solutions containing a greater
amount of KC] than this showed many segmented eggs but no swimming

102 Martin H. Fischer.

larvee. In solutions less concentrated than 85 c.c. sea-water + 15 C.C.
KCl 23m, up to this time, no swimming. larve had developed. A few
hours later occasional swimming larvee were found in these solutions also.

At 1.00 A. M., twenty-four hours after the return of the eggs to sea-water,
with an occasional exception, every egg from Lot I of the mixture of 80
c.c. sea-water + 20 c.c. KCl 24m was swimming. In Lot I of 824 c.c.
sea-water + 17} c.c. KCl 25 m many were swimming. In the rest of the
dishes all were dead.

The control eggs left in the sterile sea-water contrasted sharply with
those which had been in the sea-water-salt solution mixtures. Not a
single egg segmented throughout the night. Eighteen hours after the
beginning of the experiment a few were found in the two to four cell
stage. Six hours later, when the other dishes were teeming with swim-
ming larvee, the control eggs were granular and surrounded by a granular
débris. The dead eggs floated in the water, and were held together by a
gelatinous material. Not a single swimming larva developed.

Many of the swimming larva lived through the next day. One was
removed to fresh sea-water in a separate dish. ‘This embryo lived fifty-
seven hours.

Experiment IT, July 6, 1902.— 11.05 P.M. It was necessary, first of all, to
repeat the experiment of last year. I distributed the eggs of a freshly
caught Nereis into the following solutions :

1. 25 cc. KCl 25+ 75 c.c. sea-water.
2. 225 c.c. KCI 23 m + 77} c.c. sea-water.
3. 20 c.c. KC] 24m -+ 80 c.c. sea-water.
4. 174.c.c. KC] 2} m + 821 c.c. sea-water.
5. 15 cc. KCl 24+ 85 c.c. sea-water.
123 c.c. KC] 25a + 87% c.c. sea-water.
7. 10 c.c. KC] 242+ 90 c.c. sea-water.
3. 74c.c. KC] 24m + 924 c.c. sea-water.
9. 5 cc. KC] 2im+95 c.c. sea-water.
x c.c. KCl 24m + 974 c.c. sea-water.

1. Sterile sea-water (control).

6
8.
10.
1

I removed one lot of eggs at 12.00 midnight, a second at 12.25 A.M.
These will be designated as Lots I and II.

At 3.00 A. M. July 7, every egg in the control was intact. A few eggs
were removed from Lot I of Solution 4 for microscopic study. These
were somewhat shrunken and opaque, had lost their nuclei, and no longer
showed the peripheral arrangement of the oil globules which is so charac-
istic of the unfertilized egg. At 4.30 a. M. I found the same condition of
affairs in Lot II of Solution 5. The eggs were fissured and broken up
into irregular masses (cells?). Only occasionally were these arranged
regularly as in the normal cleavage of the fertilized egg.

Artificial Parthenogenests in Nerets. 103

At 9.00 A.M. conditions were much the same, except that cleavage
had gone further. At 10.30 I noticed that many of the lines of cleavage
had become obscure or had disappeared entirely, so that the eggs were
again spherical. ‘The control eggs were still intact. At 11.30 A. M.,
however, a few were found in which the peripheral arrangement of the
oil globules was lost. The eggs from Solutions 7 and 8 were nearly all
intact. A few were somewhat shrunken, irregular in outline, and had
lost the peripheral arrangement of their oil globules.

At 9.30 P. M. conditions were as follows: Not a single swimming larva
was found in the control. Some of the eggs had undergone a granular
degeneration and were going to pieces. A few had lost the peripheral
arrangement of the oil globules; the remainder were intact, and looked
as though they had just been removed from the ovaries. Both lots of
Solutions 1 and 2 yielded no swimming larve. ‘lhe eggs were irregular
and contracted, with thick, clear membranes. Many were dying, and no
traces of segmentation could be found in Lot II of Solution 1. Several
swimming larve were obtained in both lots removed from Solution 3.
The optimum salt-solution sea-water mixture was Solution 4. This
yielded many swimming larvee, as did also Solution 5. A few swimming
larvee were found in Lot II of Solutions 6 and 7. One swimming larva
was found in Lot II of Solution 8, and one in Lot II of Solution 9. An
occasional egg, segmented into two or four cells, was found in Solution ro.

At noon, July 9, 1902, larve were still swimming in the dishes contain-
ing the eggs from Solutions 3, 4, and 5. ‘The control eggs still showed
the peripheral arrangement of the oil globules, or else were broken up into
a granular débris. In the other dishes, all were dead. I kept the control
eggs for another day, but no swimming larve developed.

The experiment seemed to leave no doubt as to the correctness of
the findings of last year. The unfertilized eggs of Nereis, if left un-
disturbed in sea-water, will not develop into swimming larve. After
several hours, however, they may show the changes which precede
cleavage, or may actually divide into two or four cells. If the eggs
are, however, kept for an hour in a mixture of 17}. c.c. KCl] 24m +
824 c.c. sea-water, and are then removed to ordinary sea-water, they
will develop into swimming larve.

Experiment III, July 8, 1902.— 2.10 a.m. I wished once more to assure
myself of the correctness of the results obtained in the last experiment ;
I wished also to discover whether the KCI acted only through its osmotic
effects. The following solutions were therefore prepared :

104 Martin FH. Fischer.

25 c.c. KC] 2im-+ 75 c.c. sea-water.
223 c.c. KCl] 23 m + 77% c.c. sea-water.
20 c.c. KCl 24 m+ 80 c.c. sea-water.
174 c.c. KC] 24m + 823 c.c. sea-water.
15 c.c. KCl 24m+485 cic. sea-water.
123 c.c. KC] 24m + 823 c.c. sea-water.
10 c.c. KCl] 23m+90 c.c. sea-water. —
74 c.c. KC] 21m + 921 c.c. sea-water.
5 cc. KC] 2im+95 c.c. sea-water.
10. 24 c.c. KCl 24m + 974 c.c. sea-water.
ll. 174c.c. NaCl 24m + 823 c.c. sea-water.
12. Sterile sea-water (control).

SOR CO SIT ON et Cont as

Lot I was returned to sea-water after 45 minutes; Lot II, after 75, and
Lot III, after 105 minutes.

To avoid repetition, I will only state in which solutions swimming larvee
were obtained, leaving a description of the histological findings until later.
At 4.00 P.M. the first ciliated larvae were discovered in Lot II of Solu-
tion 4. At 11.30 P.M., when it was certain that all the larvee which would
develop were swimming, conditions were as follows :

The majority of the eggs in the control were somewhat shrunken and
opaque, and many had lost the peripheral arrangement of the oil globules,
and were segmented into two and four cells. None were swimming.

All three lots of the NaCl-sea-water mixture were full of swimming
larve. Many swimming larve developed in Lots I and II of Solutions 4
and 5. Several ciliated larva were also found in Lot II of Solutions 3,
6, and 7. The eggs from Solutions 1 and 2 were shrunken and granular.
The capsules were swollen and the eggs were going to pieces. In solu-
tions 8, 9, and to, the majority of eggs had segmented into two and four
cells, but none were swimming.

At noon on the following day, no swimming larvae had developed in the
control; a large number of the eggs were dying. Lots II and III of
Solutions 4, 5, and 6 still contained many swimming larve. The eggs
which had been in the sodium-chloride solution were still swimming

beautifully. An occasional ciliated larva was found in Solution 7. In the
other dishes everything was dead.

‘The experiment again showed that when the unfertilized eggs of Nereis
are left for a certain time in sea-water, the concentration of which has been
increased to a definite degree by the addition of either KCl or NaCl, they
will develop into swimming larvee when returned to ordinary sea-water.

Experiment IV, July 10, 1902. — 2.15 A.M. In the experiments described
thus far, only electrolytes have been used to increase the concentration of
the sea-water. If parthenogenesis in Nereis is due simply to the ab-

Artificial Parthenogenests <n Nerets. 105

straction of water from the egg, we should be able to bring about the
development of the unfertilized egg, not only by mixtures of sea-water and
electrolytes, but also by mixtures of sea-water and non-electrolytes. ‘To
test this point, the following solutions were prepared :

50 c.c. cane-sugar 2 + 50 c.c. sea-water.
40 c.c. cane-sugar 2 7 + 60 c.c. sea-water.
30 c.c. cane-sugar 2 + 70 c.c. sea-water.
10 c.c. cane-sugar 2m + 90 c.c. sea-water.
Sterile sea-water (control).

KR st

The first lot of eggs was removed from the solutions after go minutes, the
second, after 110 minutes.

At 12.30 P.M., all the eggs which had been in Solutions 1 and 2 were
dead. Both lots of eggs from Solution 3 were full of swimming larve.
In the remaining dishes, every egg was intact.

A mixture of 30 c.c. cane-sugar 2 m + 70 c.c. sea-water has about the
same osmotic pressure as a mixture of 15 c.c. KCl 23m + 85 c.c. sea-
water. It seems, therefore, as though the essential factor in bringing about
artificial parthenogenesis in Nereis is an abstraction of water from the egg,
and that it does not matter decidedly whether the concentration of the
sea-water used for this purpose is raised by the addition of electrolytes or
by the addition of non-electrolytes.

The above experiments were repeated several times, but always
with the same results. The eggs of Nereis are not naturally parthe-
nogenetic, but cleave and develop to the swimming stage if immersed
for from one-half to one and one-half hours in sea-water, the concen-
tration of which has been raised a definite amount, and then returned
to ordinary sea-water. It does not matter whether electrolytes or
non-electrolytes are used for this purpose, — sodium chloride, potas-
sium chloride, cane-sugar may be used,—though sodium chloride
usually yields the largest number of swimming larvz, while sugar
yields the least.

A series of experiments had been begun in which I attempted to
obtain swimming larvz by altering the ion concentration of the sea-
water without altering its osmotic pressure, when the material gave
out, and put an end to further work. Thus far the experiments in
this direction have yielded only negative results; but they will be
continued next year.

106 Martin H. Fischer.

THE DIFFERENCES BETWEEN THE NORMALLY FERTILIZED, PARTHE-
NOGENETIC AND UNFERTILIZED Eocs oF NEREIS LIMBATA.

At ordinary summer temperature, the fertilized egg of Nereis
limbata throws out its polar bodies about one hour after fertilization,
and cleaves for the first time some fifteen or twenty minutes later.
The second cleavage occurs about one hour and forty-five minutes
after fertilization. Cleavage then proceeds in a regular manner, and
the blastula stage is reached in seven hours; the eggs swim in about
eleven hours, varying somewhat with the temperature.

When the unfertilized eggs are brought to the swimming stage
through a temporary residence in sea-water, the osmotic concentra-
tion of which has been increased a definite amount, no special changes,
save a slight shrinkage of the protoplasm, a disappearance of the
nuclear membrane, and a slight increase in opacity, are noted while
the eggs remain in the concentrated sea-water. An hour after being
returned to ordinary sea-water, the egg protoplasm becomes more

Cc

FicuRE 1.— A. Unsegmented egg of Nereis limbata immediately after its return to
ordinary sea-water after immersion for seventy-five minutes in a mixture of 17} c.c.
3m NaCl + 82} c.c. sea-water. The eggs were returned to sea-water at 3.25 A.M.
Experiment 3. 4. Same egg, one hour later. The increase in the opacity of the
protoplasm is not indicated in the drawing. C. Same egg, 4.55 A.M. J. Same egg,
5.10 A. M.

opaque and its outline somewhat irregular. The characteristic
peripheral arrangement of the oil globules also disappears, and they
become clumped irregularly in the central portions of the egg (Fig. 1,
B). During the second hour after their return to ordinary sea-water,
many of the eggs cleave into two cells, after which cleavage into four,
eight, and sixteen cells may occur in a more or less regular manner
(Fig. 1,C,D). Much oftener, however, no change whatever takes
place in the two hours following the return of the eggs into normal
sea-water, and then the eggs suddenly fissure irregularly, and break

Artificial Parthenogenests in lNerers. 107

up at once into a number of spherules (cells?) (Fig. 2, B,C). Cleav-
age may then continue until such pictures as are shown in Figures 3,
A, B, C, D, E, are obtained which were made when the first evidences

FIGURE 2.— A. Control egg from Experiment 3. 4, C. Eggs which had been in a solu-
tion of 90 c.c. sea-water + 10 c.c. KCl] 23m for seventy-five minutes, and were then
returned to sea-water. The eggs were removed from the solution at 3.25 A.M. The
drawings were made at 6.00 A.M.

of ciliary movement were noticeable among the eggs. This stage is
reached in about fourteen hours, though the majority of the eggs do
not swim until six or ten hours later. The swimming stage is there-
fore reached at a much later time than when the eggs are fertilized

FicureE 3.— Parthenogenetic larva from Experiment 2. The drawings were made four-
teen hours after the return of the eggs from Solution 5 to ordinary sea-water. Some
of the eggs were just beginning to swim.

by a spermatozoon. At other times, cleavage occurs in a much more
regular manner (Fig. 4, A, C), and the swimming larvae have an
appearance which cannot be distinguished from that of trochophores

108 Martin H. Fischer.

produced through fertilization with sperm. It often occurs that lines
of cleavage disappear, and cells coalesce, as Loeb found to be the
case in Cheetopterus.!

The unfertilized eggs of Nereis, when left in ordinary sea-water,

FIGURE 4.— Parthenogenetic larve from Solution 11 of Experiment 3. The eggs were
swimming, but the cilia are not indicated in the drawings.

F G

FiGuRE 5. — Unfertilized control eggs of Nereis, showing the various changes they suffer
when left in ordinary sea-water. 4, B. Control eggs twenty-eight hours after their
removal from the ovaries. The eggs cleaved, and then underwent a granular degener-
ation. C. Control egg from the same dish, practically intact. 2, £, #. Control eggs
in which cleavage occurred. G. A control egg in a state of granular disintegration.

' Lors, J.: This journal, 1go1, iv, p. 423.

Artificial Parthenogenesis in Nereis. 109

usually show no change during the first eight or twelve hours after
their removal from the ovaries. Often they are absolutely unaltered
after remaining in sea-water for thirty-six hours (Fig. 5, C), when
they become opaque and granular and suddenly break up into a
granular débris (Fig. 5, G). At other times the unfertilized eggs
become slightly opaque after lying in sea-water for several hours, lose
the peripheral arrangement of the oil globules, and go through one or
two cleavages (Fig. 5, D, E). They then become granular and go
to pieces (Fig. 5, A, B). Some of the eggs may even divide into
a dozen cells (Fig. 5, F). Never, however, do the unfertilized eggs
develop into swimming larve. It can often be seen that while one
dish of control eggs will show no evidences of cleavage whatsoever,
another dish of the same eggs will show a large number in the two or
four cell stage. Mere mechanical agitation is certainly not respon-
sible for this change, for I have often found that while none of the
eggs which had been transferred from one dish to another showed
any signs of development, those which had been kept absolutely un-
disturbed showed a large number of eggs in the two or four cell stage.
In looking over the notes of my experiments, I find that such evi-
dences of development occurred most often in dishes containing large
numbers of eggs. It is possible, therefore, that lack of oxygen or
the formation of carbon dioxide lie at the basis of the process, —a
question which will be studied next summer.

4.0 RF

hy .
.
eo"
'
.

cs

bo = 1 Lae Ge;
AST ESF Oy AE ey Gy be a ¥
; F i :
ite : _ nn Tes
é = 4 sors vey a oe vets
‘ i ‘
ake Oh > Lt oe coma Sh)
4 . A
.
aes t )
‘ A
‘ ft
ei “
* , * < 4
;
; n
7 i .
“
‘ a
,' j
‘ ’ ss
Fe
4
\
i .
4
ie sy
sh
‘
i
t .
¥ wal ' ‘
J i 4 a
y Tork.
{
‘
1
4
Pie tS
4
-

oo

ha)

aN

4

CARRE R INGO Pat hid

, i

.

~lTHE IMMUNITY OF FUNDULUS EGGS AND EMBRYOS
FO ELECIEICAL..STIMULATION.

BY ORVILLE EH. BROWN.
[From the Hull Physiological Laboratories of the University of Chicago.

T has been shown by Loeb! that the eggs and the embryos of the
small fish Fundulus heteroclytus are peculiarly resistant to
sudden changes of osmotic pressure of the media in which they live.
The eggs when fertilized will develop and live in distilled water, in
sea-water, or even in concentrated sea-water. The most probable
explanation of this curious fact seemed to be that the tissues were as
easily permeable to salts in the water as to the water itself. If this
is true, polarization will not easily occur in such eggs when subjected
to an electrical current; and if the stimulating action of an electrical
current is due primarily to polarization, the egg and the young fish
should be almost immune to such a current. Dr. A. P. Mathews
suggested that I compare the effect of the current on the egg, the
embryo, and the adult fish. The results confirmed our expectations.
It was found that the eggs are polarized with great difficulty; the
embryos show almost a complete immunity to electrical stimulation,
while the adult fish reacts to even a weak current in a positive
manner.

METHODS.

It was necessary to observe the eggs under a microscope. To
make this possible, a small paraffin cell was moulded upon an ordinary
glass slip. Into each end of the cell was inserted the tip of one of
Porter’s non-polarizable boot electrodes. The boot electrodes were
attached to Mesco dry batteries, and also to Daniell elements, which
were used in connection with a voltmeter. The Fundulus eggs were
surrounded in different experiments with water the density of which
varied from sea-water to distilled water. In some cases most of the
water was removed and the eggs were placed in contact from one
electrode to the other. Control experiments were performed upon

1 Logs, J.: Archiv fiir die gesammte Physiologie, 1894, lv, p. 390.
g y s 9 Pp. 39
Ii!

112 Orville H. Brown

Asterias and Arbacia eggs. As these Echinoderm eggs go to pieces
in distilled water, they were placed in a normal solution of urea. It
was necessary to place the eggs in solutions which were very poor
conductors, so as to cause the passage of the maximum amount of
current through the eggs. Exact comparisons of the effect of the
current on the Fundulus, and Asterias and Arbacia eggs cannot be
made, since the retaining solutions did not have the same power to
conduct the current. The free swimming embryos and the adult
fishes were worked upon in a rectangular dish, the dimensions of
which were two and a half inches by five inches. The dish contained
only sufficient distilled or other water to allow free swimming move-
ments. The current was conducted into the water in these cases by
means of large electrodes at the ends of the dish.

A. Effects of the current on the eggs. Hxperiment 7. — It had been ascer-
tained by previous experiment that currents of ordinary strength had no
apparent effect upon either the Fundulus eggs or the embryos. ‘The eggs
were placed in distilled water. The strength of the current was increased
to fifteen cells with but little effect; eighteen cells produced definite
results. To make sure that the current was good in every case, the cells
and the circuit were tested by the voltmeter. The protoplasm retracted
from the side near the negative electrode, and travelled toward the oppo-
site side, where liquefaction of the membrane occurred. Some of the
yolk of the egg and many of the oil globules escaped and travelled with
the cathode stream, confirming Lillie’s' results. This required a constant
passage of the current for fifteen to twenty minutes. With this strength
of current, however, many of the eggs did not burst and allow an escape
of the egg substance. ‘There was only a shrinking of the inner membrane
and the yolk of the egg.

Experiment I[f.— This was performed upon Asterias and Arbacia eggs in
normal urea solution, as a control experiment. Ten dry cells were used.
These were sufficient to cause a great many of the eggs to liquefy and
burst upon the side near the positive electrode, and allow an escape
of the egg substance (vzde Lillie).

Experiment I[1/.— The object of this experiment was to see whether the
developing Fundulus eggs are affected by the passage of a current of
medium strength. The eggs were fertilized, and when in the two-cell
stage, they were subjected in distilled water to the current of five cells,
which was allowed to flow for an hour, at the end of which time the eggs
had developed to the sixty-four-cell stage, apparently with no irregularities.

1 LILLIE, R. S.: This journal, 1903, viii, p. 273.

Lmmunity of Fundulus Eggs to Electrical Stimulation. 113

B. Effect of the current on the embryos. FAxferiment /.— Four day old

embryos, the hearts of which were just beginning to beat, were placed in
the paraffin cell containing distilled water through which the current from
twenty-one cells was passing. ‘There was no twitch at either make or
break. The heart maintained apparently the same rhythm as before the
passage of the current.

Experiment IJ. — Embryos twenty-one days old were placed in the rectangu-

lar dish containing a shallow layer of distilled water. Nothing less than
eighteen cells would cause a twitching of the small fish. By the use of
twenty-one cells, they responded somewhat as the adult responds to a
very much weaker current, as will be shown. When the fish were in sea-
water, or fresh water, the current had no effect.

C. Effect of the current on the adult. Hxferiment 7.— The normal adult

2.

Fundulus was placed in the same dish as was used in the previous experi-
ment. There was just sufficient fresh water in the dish to cover the fish
completely. By the use of one ceil the fish responded to the make
shock with a jump. It turned toward the anode and swam to it. By
the use of three cells, the following results were obtained :

If the fish was turned with its head toward the anode when the current
was closed, it was paralyzed and laid over on its side absolutely relaxed.
Breaking the current caused Ritter’s tetanus.

If the fish was turned with its head toward the cathode when the circuit
was completed, it immediately turned around until its head was toward the
anode, and turned over onto one side, paralyzed.

If the fish was held in the stream with its head toward the anode, it re-
mained quiet.

If held with its head toward the cathode, it endeavored to turn. ‘The
fins were in rapid vibration, and the tips of the pectoral fins pointed
toward the cathode.

When the fish was held by its head crosswise of the current, its tail
turned toward the anode. The fin on the anode side was held at an
angle of about forty-five degrees from the part of the body posterior to
the fin, while the fin on the cathode side was held nearly at right angles
to the same part of the body.

When a fish was held by the middle of the body, crosswise of the cur-
rent, its head was turned toward the anode.

Experiment IZ. — This experiment was to show the effect of the current on the

I.

fish after the brain has been destroyed. The method was the same as in
the experiment just preceding.

If its head was toward the anode, using four cells when the circuit was
closed, the animal made a feeble jump toward the anode and became
quiet.

114 Orville H. Brown.

2. If its head was toward the cathode at the make of the current, the fish
made efforts to turn away from the cathode, but it never succeeded.

3. The same response to the current that Loeb and Garrey* observed in
the larvae of Amblystoma, was also detected in the adult Fundulus with
the brain destroyed. ‘That is, when the current flows from the head to the
tail, the back becomes arched, while if the current passes from the tail to
the head, the back becomes concave, both the head and the tail being
raised.

4. When the animal is crosswise of the current, the fins move at make or
break of the current, but neither the head nor the tail turn toward the
anode as they did in the normal animal.

Experiment [[1.— 'Vhe object of this experiment was to show the effect of the
current upon the fish when both the brain and the spinal cord are de-
stroyed. The method was the same as in the two preceding experiments.
The same strength of current was employed. All the responses were
very slight.

1. When the fish was held on its ventral side, with its head toward the
anode, no responses other than a slight movement of the fins at make or
break were made.

2. When its head was toward the cathode, the pectoral fins were held
nearly at right angles from the body, with their tips slightly curved toward
the cathode.

3. When the animal was turned at right angles to the current, a twitch
of the fins was noticed, which was about as great on one side as the
other.

4. ‘The same phenomena as was observed in the animal with only the brain
destroyed, were also seen here, z.¢., the descending current caused the
body to be arched, while the ascending current caused the head and tail
to be raised.

SUMMARY AND CONCLUSIONS.

1. The fact that the eggs of Fundulus are immune to electrical
currents, as they are to osmotic changes of the medium in which they
live, is an interesting physiological confirmation of the theory of the
osmotic nature of electrolysis. The most probable explanation
appears to be that the membranes of the egg are so freely permeable
to ions and possibly to neutral particles that no polarization can occur.

2. This conclusion is supported by the behavior of Arbacia and

Asterias eggs, which show the contrary relation, being readily
susceptible both to electric currents and to osmotic changes.

* LoEB and GARREY: Archiv fiir die gesammte Physiologie, 1896, Ixv, p. 41.

Lumunity of Fundulus Eggs to Electrical Stimulation. 115

3. The fact that these eggs and embryos are not affected by a
current, except when very strong, supports the hypothesis that
electrical stimulation depends primarily upon polarization where the
current enters and leaves the cell.

4. There is a gradual increase in susceptibility to osmotic changes
and to the electric current as the embryo develops, the adult fish
being readily stimulated by the current from a single cell, which is
quite without action in the embryo.

5. It is suggested that the resistance of the electric and other
fishes to electric stimulation, noted by Du Bois-Reymond and others,
may not improbably be similarly explained by the permeability of the
walls of their tissue cells to ions.

6. The liquefaction of the eggs on the anode side, and the quieting
effect of the anode, supports Mathews’ hypothesis of the dissolving
action of the cathions and their inhibitory action.

7. The galvanotropic reactions are shown to depend primarily upon
the nervous system, as already made out by Loeb in other forms.

I take this opportunity to thank Dr. A. P. Mathews and Dr. R. S.
Lillie for suggestions and aid.

ON THE INFLUENCE OF VARYING INTENSITIES Sa
QUALITIES OF VISUAL STIMULATION? G¥er
THE RAPIDITY OF REACTIONS ae
AUDIVORY STIMUET:

By ROBERT MacDOUGALL.

HE interaction of hetero-sensorial stimulations and the influence

of sensory stimuli upon simultaneously existing motor processes
have been reported for a variety of activities. Such, for example, is
the influence of musical tones in sharpening the perception of color
when the stimulus is below the threshold of unassisted vision ; such
is also the reinforcement or reduction of force which a maximum
motor reaction receives from incident visual stimulations of specific
qualities. It is but a step to extend the application of the concept to
other sensori-motor relations. In the present paper is reported a
preliminary investigation concerning one of these aspects.

Reaction times—the period required to make a movement in
response to a preconcerted signal — have been found to vary in de-
pendence upon at least the following factors: a, the sensory type of
signal employed, 4, its intensity, c, its location, d, the presence or
absence of a preceding warning, e, the period intervening between
warning and stimulus, f, the nature of the reaction which is made,
and g,its energy. The question here raised is whether, in addition
to these special factors of variation, the intensity and quality of the
general stimulation falling upon the organism affect the rapidity of
its responses in a way analogous to the dynamogenic influence upon
reactions of these stimuli.

Two factors of possible variation were studied: first, changes in the
intensity of the stimulation falling upon the retina, reactions being
made: a, in darkness, 4, in weak light, and c, in strong light; and
second, changes in the quality of the visual stimulation, reactions being
made under illumination by the following homogeneous lights: a, red,
b, orange, c, green, and d, blue.

The subject sat before a table with his right forefinger on the re-

116

Rapidity of Reactions to Auditory Stimut. 117

action key. His head was enclosed in a blackened box having a front
of milk-glass, by which the distracting influence of a diversified field
of view was eliminated. The variations in light intensity were con-
trolled by the use of sixteen and two candle-power electric bulbs, and
by shutting off the illumination completely. The reaction stimulus
consisted of the click of a telephone receiver attached to the head
over the ear. A preparatory signal —also auditory —.was given
two seconds before the application of the stimulus. The reactor
directed his attention to the response to be made, as this attitude
appears most conducive to regularity, as well as rapidity in the re-
actions. Records were made by a Hipp chronoscope, corrected daily
by means of a gravity chronoscope. Preliminary training was given
in a series of experiments to establish the normal reaction times of
the subjects in ordinary daylight. Two persons took part in the
experiment.! In view of this fact and the circumstance that only five
series of ten reactions each were ordinarily used in making up the
individual averages, the quantitative results presented must be held
subject to revision.

The reaction times obtained by varying the intensity of the inci-
dent light are exhibited in the following table:

ARATE EHR le

Subject. Darkness.

140.6 o
139.7 «

Average 140.1 o

No dynamic relation appears between intensity in the illumination
and rapidity in the response. The reaction time is shorter in dark-
ness than in light in the ratio—dincluding both intensities — of
140.1 to 144.9. The reaction time is also shorter under the weaker
than under the stronger illumination, the difference being more
marked between weak and strong light than between illumination
and darkness.

1 Messrs. W. E. HockinG and W. P. Burris, to whom I am indebted for
carrying out the whole series of reactions.

118 Robert MacDougal.

The mean variations within the reaction series are given in the

following table:
TABLE IL.

Subject. Darkness.

6.4%,
10.7%

Average. 5%

The regularity of the responses under these conditions is not notice-
ably affected by the change from light to darkness, the mean varia-
tions, being, respectively, 8.1 per cent and 8.5 per cent. Both
reactors show an increase in inconstancy in passing from strong to
weak light, — 6.9 per cent to 9.3 per cent, —the latter showing also
more disturbance than darkness. The responses are thus swifter and
more irregular in faint light and in darkness than under conditions
of illumination more nearly approximating ordinary daylight, results
which suggest the influence of changing mental attitudes due to
factors of novelty and attention, rather than any direct dynamogenic
relation between sensory and motor processes.

The reaction times obtained under homogeneous light are given in

the following table:
TABLE III.

Subject. . Orange. Green.

129.2 «

132.2 ¢

Average. 129.6 « . 130.7 o

The reactions under colored light are made more rapidly than
under neutral light, in the ratio of 130.7 to 148.4. This accel-
eration is presented by both reactors under every quality of
homogeneous light. The reactions made under colored light are
shorter also than those made in darkness, in the ratio of 130.7 to
140.0. No constant relation appears between the various color quali-
ties and the time of reaction. Extreme rates, in the case of both

Rapidity of Reactions to Auditory Stimuli. 119

subjects, occur under red and orange stimulation, intermediate rates
with blue and green. The reaction times of the two subjects follow
the same curves in blue and green light, but converse curves in red
and orange, and the differences are in all cases slight.

The regularity of the responses is greater under neutral than
colored light, in the ratio of 6.9 per cent to 7.6 per cent. The
values of the mean variation under the latter conditions are given in

the following table :
TABLE Iv.

Subject. : Orange. Green.

6.7%

Average.

The responses are thus swifter and more irregular under colored
than neutral illumination, but, as in the latter case, no direct
dynamic relation appears between sensory and motor processes.

The differences which do appear point rather to a waxing and wan-
ing in the interest with which the reactor attends to a stimulus under
conditions of relative novelty and familiarity respectively, and to a
facilitation of the nervous discharges in general. When the reaction
series were arranged in pairs, and a change in the quality of the light
was made in passing from the first to the second set, the reaction
time in the later series was distinctly reduced. With each reactor
this change appeared in five out of the six color successions intro-
duced, as shown in the following table:

TABLE V.-

Subject A. Subject B.

Succession of colors.
Series 1. Series 2. Series 1. | Series 2.

Red followed by blue . . . 160 o
Blue followed by red . . . 5 7 168 «

Blue followed by orange. . 5 154 o

Orange followed by blue . . § 150 o

Green followed by blue . . 165 o

120 Robert MacDougal.

Such a shortening of the reaction time takes place as the result of
simple habituation, and the effects of this process might be expected
to appear within the period under discussion. That the phenomenon
is not a practice effect is shown by the results of introducing paired
series in which the reactions in each of the successive sets were
made under the same conditions of stimulation. Reaction times are
presented in regard to three colors only, two of which are partial,
since the repetition of color series was not made for the purpose of
determining the present point, and its occurrence is, in this connec-
tion, accidental.

TABLE VI.

Subject A. Subject B.

Colors.

Series 1. | Series 2. Series l. Series 2.

Riedie oan 162 o 167 o Ilo 15900;

Orange. . ae Ears l4lo 1526

Blue .

The second series, in each of these four cases, exhibits an increase in
the reaction time, instead of the decrease which is to be looked for
as the result of practice. The only discernible difference in the two
cases lies in the relation of the reaction to its sensory environment,
which in the one instance presented the maintenance of the preceding
conditions of general stimulation, and in the other involved a passage
to circumstances possessing elements of relative novelty. It is pos-
sible that colored light exerts a dynamogenic influence upon reaction,
and that its effects die away as the reactor becomes habituated to the
stimulation. It seems more reasonable, however, to attribute the
phenomenon to fluctuations in the general alertness of the reactor,
which take place in dependence upon changes in his surroundings.
The attitude of attention is characterized by nervous facilitation, and
an increase in the general bodily innervations, tending to produce
both an acceleration in the responses made to stimulation, and an
intensification of their force. This appears in the relation which has
been found to hold between the reaction time and the force with
which the movement is executed. It makes it probable also, if means
were devised for measuring the intensity of each reaction as it took

Rapidity of Reactions to Auditory Stimulz. 121

place, that the correlation between force and swiftness in the response
might account for the differences which exist between so-called sen-
sory and motor reaction times. It is in this relation, further, that we
should look for a partial explanation of the acceleration which follows
the use of a signal preparatory to the application of the stimulus, for I
am convinced that if the force of the reactions were recorded in the two
cases, it would be found that a higher general average would be pre-
sented when a warning is employed, than when the stimulus is applied
without a preparatory signal. In short, the factor determining the rate
of response to homogeneous sensory stimuli appears not to be mate-
rial differences in their quality or intensity, but the degree to which
they succeed in stimulating the attention of the reactor, and thereby
increasing the general nervous excitability of the moment.

ON THE RELATION OF EYE, MOVEMENTS TO LIMIT-
ING VISUAE STIMUET:

By ROBERT MacDOUGALL.

HE general phenomena of geometrical optical illusions point to
a constant relation between the movements of the eye and the
varying intensities and qualities of luminous objects within the field
of vision. Reflex movements are incessantly initiated by the appear-
ance of strong local stimuli, and voluntary movements, with scarcely
less frequency, are affected by the presence of unequal intensities and
qualities of illumination in different parts of the field of view, as the
point of regard moves over it. Theoretically, it may be said that
every bright point or object one sees, either causes a movement of the
eye, or inhibits it if it is in progress, — including within the term any
retardation as well as the complete cessation of motion. These un-
regarded factors, causing movement where none is intended or recog-
nized, and transforming it in greater or less degree when it is in
process, give rise to a variety of errors in our judgments of distance,
direction, size, form, and the like.

The investigation here reported, which concerns one of these types
of distorted judgments, took shape from the results of certain obser-
vations made in connection with an experimental study of the location
of the subjective horizon.!_ Judgments were made in a darkened room,
by means of an illuminated disc, running upon a vertical carriage,
placed at some two metres distance from the observer, who, by means
of a system of cords, adjusted the disc upon the imaginary horizon.
At the top and bottom of this carriage, electric bulbs were placed, by
which, at the pressure of a button, a small rectangle could be illumi-
nated with diffused neutral light. The purpose of these arrangements
was not made known to the observer, who was directed to adjust the
disc without reference to them. The results of one hundred and

* Published in Harvard Psychological Studies, Vol. I. (Monograph Supple-
ment 17, of the Psychological Review).
122

Relation of Eye Movements to Limiting Visual Simul. 123

ninety observations, distributed among six subjects, showed a differ-
ence in the location of the imaginary horizon under the two conditions
of 32.71’ of arc. In the article referred to, the following comments
are made: ‘The eye is uniformly attracted toward the light, and the
location of the disc correspondingly elevated or depressed. The
amount of displacement which appears is relatively large. It will be
found to vary with the intensity, extent, and distance of the illumi-
nated surfaces introduced. There can be little doubt that the practical
judgments of life are likewise affected by the distribution of light in-
tensities, and possibly also of significant objects, above and below the
horizon belt. Every brilliant object attracts the eye toward itself;
and the horizon beneath a low sun or moon will be found to be located
higher than in a clouded sky.”

In order to examine certain of these factors in more detail, a series
of experiments was arranged in which the observer was called upon
to locate the median point between two limits disposed in a horizontal
line, one of which remained constant, while the other was varied in
size, form, brightness, and color, successively.

A vertical black screen, eighty centimetres square, was set up at a
distance of two metres from the observer. Two rectangles of white
cardboard, five by twenty millimetres in size, and fifty centimetres
apart on a horizontal line, were attached to this screen with their
longer axes vertical. These formed the limits during the preliminary
series of determinations in which the normal variability was established
for each observer. A third card, similar to these limits, was attached
to a fine black cord passing over pulleys at the sides of the screen,
and thence to the observer, whose judgments were directly recorded
by adjusting this travelling index at what appeared to be the middle
point between the two extremes, the reckoning being always from the
inner boundaries of the limiting areas, whatever the character of these
might be. The observer then sat with closed eyes while his location
of the median point was recorded, and the index displaced, — now to the
right, then to the left, — in preparation for the following experiment.
Records were made in terms of displacement from the objective centre,
those toward the variable limits being marked (+), those toward the
constant limit (—). Twenty judgments were taken with the variable
limit on the right, and a similar number with it on the left, in each
case ten judgments being made after displacement of the index to the
one hand, and ten after displacement to the other. Forty judgments
thus formed the basis of each individual average, while three observers

124 Robert MacDougall.

took part in the experiments, giving a total of one hundred and
twenty reactions for each set of figures presented in the results.!

Six series of changes were made in the variable limit, as follows:
The constant limit was replaced, 1, by vertical strips of white card-
board two centimetres in width, and successively two, four, eight,
twelve, sixteen, and twenty centimetres in length ; 2,-by similar hori-
zontal strips; 3, by squares of two, four, and six centimetre sides ;
4, by two strips inclined to form an angle, which were, a, three
centimetres in length and enclosed angles of forty, ninety, and one
hundred and forty degrees, respectively, and, 6, six centimetres in
length, enclosing the same angles ; 5, by areas equal to the constant
limit, but of light gray, dark gray, and black cardboard respectively ;
and, 6, by areas equal to the constant limit, but of the following
colors: red, orange, yellow, green, blue, and violet.

The quantitative results of locations made under the first two series
of changes are given in the following tables:

TABLE 1.
VARIABLE LIMIT HORIZONTALLY EXTENDED.

Length.

ACD:
CE:

M.V.

GAB IVE all.
VARIABLE LIMIT VERTICALLY EXTENDED.

Length. - . 12 cm.

A.D.
(E5105
M.Y.

3.26

=0'25

2.98

Throughout the tables A.D. stands for the average deviation of the
locations from the objective centre ; C.E. for the constant error in-
volved in the series of judgments ; and M.V. for the mean variation of
the same series. The measurements are recorded in millimetres.

1 I am indebted to Messrs. BREWER, CAREY, and MEAKIN for the record of
udgments upon which this paper is based.

Relation of Eye Movements to Limiting Visual Stimuh. 125

Vertical elongation of the variable limit is followed by no noticeable
variation in the judgments; neither constant error nor mean variation
shows concomitant changes. The average deviation alone increases
progressively. Horizontal extension of the variable limit is followed
by a deflection of judgment toward it only at the beginning of its en-
largement, and in more than half of the series the error which appears
is negative in sign. In the series of horizontal changes the mean
variation increases with the elongation of the strip, except in the case
of the last increment, where an inversion of the curve takes place.
In the series of vertical changes no such increase is manifested. The
eye movements which the process of judgment involves are made in a
horizontal line, along which axis, in the latter case, the limiting stim-
uli do not vary; in the horizontal series, on the contrary, the limit
changes continually in the line of the eye movements, and in such a
way as to offer an enlarging range of possible variations in the eye
movements made, and therefore a ground for the progressive increase
in the irregularity which is manifested in the mean variation.

The results of substituting a series of increasing squares for one of
the original limits are given in the following tables:

TAB Ei

VARIABLE LIMIT A SERIFS OF SQUARES.

Length.

A.D.
CE:

M.V.

The influence of this series of changes also is relatively slight. The
location of the median point is deflected toward the variable limit, the
amount of error at first increasing with the enlargement of the square
but falling again when the latter is much larger than the constant
limit. Average deviation and mean variation follow similar curves,
appearing as constant functions of the deflection of judgment. The
form of these curves appears to be due to a change in the mode in
which the variable limit enters into the process of estimation. It is
probable that in the case of the smaller squares the whole area is
regarded as a unit, and the extent of the eye movements, together
with their variability, increases with each remove of the centre of the

126 Robert MacDougall.

figure from the opposite limit; but that when a certain magnitude has
been reached the square is no longer regarded as a whole, but only
the nearer border of it taken into account, in consequence of which
there results a decrease in the total excursions of the eye, and an in-
crease in the regularity of the judgments consequent upon the greater
definition of the termini between which the movements take place.

More constant and marked effects follow the introduction of the
fourth series of changes. Here the variable limit consisted of a piece
of white cardboard, cut so as to form an angle having its apex
toward the opposite limit, the sides composing the figure being five
millimetres in width and of either of two lengths, thirty millimetres
in the first sub-group and sixty in the second. In each set the en-
closed angle was made successively of 140, 90, and 40 degrees magni-
tude. The results are given in the following tables :

DABILE LV:

SIDES 30 MM. LONG.

TABLE VY.

SIDES 60 MM. LONG.

The deflection in the location of the median point presented in
these tables is positive in direction, large in amount, and varies con-
comitantly with the factors of change introduced. The constant error
in both sub-groups varies inversely with the magnitude of the enclosed
angle, the average deviation following a similar curve. The amount
of deflection also varies directly with the length of the enclosing sides.

Relation of Eye Movements to Limiting Visual Stimult. 127

The influence of the length of sides is least in the case of the smallest
angles, and greatest in those of largest magnitude. The whole exper-
imental series composes a quantitative estimation of the illusion in
one form of the Miiller-Lyer figure. The constant factor of change in
both the above sets of conditions is the spatial disposition of the total
figure which the varying angles and lines compose. The centre of
this system of lines is removed farther from the opposite limit by each
increase in length of the sides enclosing a constant angle, and by each
reduction in the magnitude of an angle enclosed by lines of a constant
length. With the variations in the position of this point, the amount
of the positive errors in judgment rises and falls.

At the same time the reactions toward this changing figure are
fairly definite and constant. The mean variation does not materially
change with the alterations made in the character of the limit. The
eye, in other words, apprehends the figure in each case as a structu-
rally constant unit, and returns to a specific point of rest determined
by the relations of the lines composing it.

The results of the last two series of changes, which may be dis-
cussed together, are presented in the following tables:

TABLE VI.

VARIABLE LIMIT: BRIGHTNESSES.

Intensity. Dark gray.

A.D.
CE:
M.V.

TABLE VII

VARIABLE LIMIT: COLORS.

Yellow. . : Violet.

2.63 F 3-57
—0.66 2.42

(ok ire nie 2.07

128 Robert MacDougall.

In the case of intensive variations the amount of deflection in-
creases with each increment of difference between the constant and
variable limits; and in each case the deflection is toward the stimulus
which both is absolutely the weaker of the two and presents less con-
trast with the general background, from which relation also a stimulus
to reflex movements of the eye might be conceived to arise. The
source of this constant error must lie in some factor other than the
sensational intensity of the limit, and to a common cause are ap-
parently due the distortions which arise from both intensive and
qualitative variations. In the latter case the following facts appear.
If the spectrum be divided into two parts, red, orange, and yellow, or
bright colors; and green, blue, and violet, or dark colors; the former
group, approaching the white limit more nearly in brightness, have
less disturbing effect upon the judgment than any of the latter group,
and the deflections which occur are not toward the more intense
stimulus, but toward the darker green, blue, and violet areas. The
series of judgments is also more erratic when colors of the dark group
are used than with those of the bright group. In general, then, the
results parallel those which appear in connection with the series of
brightnesses.

I am of opinion that these phenomena in general, namely, the factors
of variability in the visual perception of magnitude and distance, are
always to be explained in connection with the specific dynamogenic
values of the whole system of elements which the field of view at any
time contains. If illusions of direction appear, or deflections in the
judgment of distances, they occur because the visual objects on the
one side, or toward the one term, of the line of eye movement, exert
a greater reflex attraction upon the eye, and make it more easy to turn
it toward that side, and more difficult to turn it away than is the case
with movements in the opposite direction. The primary dynamo-
genic quality of a luminous surface is its intensity; the brighter or
more vivid an object, the more extended and uniform the reactions
of rotation which it is capable of calling forth. This relation is ex-
emplified in the distortion of judgment caused by the introduction of
a bright light to the one side or other of a point to belocated. Itisa
factor of recognized value in artistic composition, in which gold and
color and brilliancy are applied to objects of secondary importance,
—fabrics, decoration, foliage, skies, etc., — with the result that while
attention is not specifically attracted toward them as individuals, these
objects, nevertheless, decoy the eye to the centre of the system of

Relation of Eye Movements to Limiting Visual Stimuli. 129

motives incorporated in the picture, and thus preserve the artistic
balance of the composition.

But here the natural series of values is inverted ; the location of the
half-way point is deflected toward the darkest gray and dullest colors,
not toward the more stimulating white strip of paper. Other factors
than that of intensity must be present. We should undoubtedly find
that the whole series of elements which are taken into account in
pictorial composition, have their place in the system of factors which
determines such simple eye movements as those in question. Two of
these factors appear to be chiefly at work in producing the variations
presented by this set of experiments. ‘The first of these, which char-
acterizes the four geometrical series, is the habit of the eye, when ex-
ploring a system of visual objects in a limited field, to come to rest in
a position at which the stimulations to reflex movements of rotation
in opposite directions are in equilibrium. Such is the centre of a
_ circle, or the intersection of the system of diameters of any symmetrical
figure. This principle is most clearly exhibited in the phenomena pre-
sented by the introduction of the various angular forms, in which, as
has been said, the changes which appear in the location of the median
point will be found to parallel the shiftings of the centre of the figure -
defined by the apex of the angle and the terminations of its enclosing
sides. It is, perhaps, too much to say that this centre is represented
as the termination of the line whose centre is to be located; never-
theless the system of lines which compose this figure facilitate move-
ments toward its centre, and retard those directed away from it more
strongly than does the less striking patch of light which forms the
opposite boundary of the distance to be bisected, and thereby intro-
duces a characteristic distortion into the judgment.

The other factor to which reference has been made, is that of
novelty, —an element less easily demonstrable than the preceding.
The biological importance of habitual, prompt reaction upon novel
stimuli is as evident as are the power and constancy of the appeal
which such stimuli actually make. To this ultilitarian value an ele-
ment of intrinsic worth in the novel stimulus must be added; it is both
more important that the novel object should be attended to, and more
pleasant to receive the impression which it affords. These aspects
are both independent of the primitive sensational intensity of the
stimulus. In the series of brightnesses, and in that of colors as well,
a disturbance appears which does not reflect any consistent qualita-
tive or intensive change, but parallels the increments of difference

130 Robert MacDougal.

between the variable form and the unchanging, habitual limit. Since
these two sets of experiments took place at the close of the whole
piece of work, we may fairly attribute to both colors and neutral
shades an element of novelty, as compared with the constant white
strip which limited the prescribed eye movements on the opposite
side, and to this factor I ascribe the positive errors which are pre-
sented in the results. |

ON THE IRRITABILITY OF THE BRAIN DURING
ANAEMIA.

By WILLIAM Ji :Gl1ES:

[From the Physiological Institute of Bern University.]

I. INTRODUCTION.

Cae the summer of 1899 I had the pleasure of assisting
Professor Kronecker in a study of the irritability of the brain
during anzmia.’ Our research could not be concluded during my

stay in Bern that summer, but we both looked forward to completing
it together in the following year. Unfortunately for me, return to
the Physiological Institute has been impossible thus far, and the
work which has been delayed on that account has lately been resumed
by Professor Kronecker and Dr. Stumme. At the suggestion of
Professor Kronecker, the results of our investigation are presented
here in some detail though briefly.

In the preparation of these notes I have received numerous sug-
gestions from Professor Kronecker, who has also revised the state-
ments relating directly to our experiments. Throughout practically
all of our research, Professor Kronecker not only directed the work,
but did a very large share of it. His well-known generosity to his
pupils is again shown by his desire that this investigation, which was
chiefly his, shall seem to be wholly mine.

II. DESCRIPTION OF EXPERIMENTS.

In this research we sought especially to determine the order of
cessation, as well as the period of continuance, of certain reflexes dur-
ing anzemia of the brain.

Acute anzemia was brought about by perfusion with the solutions
indicated on the next page.

The animals employed were toads, frogs, rabbits, and dogs.

1 Gries: Report of the British Association for the Advancement of Science,
1899 (Dover), p. 897.

131

ee William J. Gres.

The solutions used were various strengths of pure sodium chloride,
Ringer’s solution, and modifications of it, Schiicking’s solution (both
of sodium and calcium saccharates), rabbit and horse serum, and
0.7 per cent sodium-chloride solution containing paraxanthin or
chloralbacid.

Experiments on toads and frogs. — Perfusion in the cold-blooded
animals was conducted with the least possible pressure through the
abdominal vein. In this series of experiments we used all of the
various solutions already enumerated, except serum.

Seventeen experiments were made, seven with toads and ten with
frogs, each of which was continued for a period of from one to nine
hours. The total amount of perfused fluid varied from 25 c.c. to
1590 c.c. In most cases perfusion was continued until the heart
ceased to beat.

The table on page 133 gives a summary of the more important
results obtained in this connection. The terms “skin,” “lid,” and
“nose,” in the table, refer to the reflex movements caused by pres-
sure on those parts.

During the period of perfusion, the following functions gradually
weakened, and then usually disappeared in this order: (1) respiration,
(2) skin reflex, (3) lid reflex, (4) nose reflex, (5) heart beat.

The relative time of cessation of these reflexes varied considerably,
not only with the character of the solutions, but also with the rapidity
of their perfusion and the amounts used.

Convulsive extension of the limbs occurred in all the experiments
in the earlier stages, but toward the close of each experiment and
before the reflexive movements of the eyelids ceased, no such mani-
festations were observed, nor could they be induced by mechanical
stimulation.

Perfusion of physiological saline solution containing 0.03 per cent
of paraxanthin induced hypereesthesia at first, but the reflex responses
quickly came to an end, as the perfusion continued. Cumulative
muscular rigor was the most pronounced feature of the experiment.
At the end of the experiment the body was perfectly stiff. With a
solution containing 0.015 per cent paraxanthin, moderate hyper-
zesthesia was observed at first, as in the case of the 0.03 per cent
solution, but the rigor of the former experiment was absent in this.

During perfusions with physiological salt solution containing
1 per cent chloralbacid, repeated spasmodic extension of the extremi-
ties was the main feature. With the solution containing 0.33 per cent

On the Irritability of the Brain during Anemia. 133

TABLE I.
Red cor-
: puscles at
ny Cessation of reflexes. 5. | the end of
: © .| Time after beginning the 53 | perfusion
o| 3 a6 perfusion. 4 5 | in fluid
2 : Bon OH
e| & Solution used. cap 4 2 from
5 S eu =
a e* 28] 3
Resp.| Ski : Heart}“ | 2] 5 | 3
Xesp.| Skin.| Lid. |Nose.|), 214 E13 |'s
O} a ja
ay a s h m. | h. m h. m ' h. m. | h. nm. h.m.| c.c
1|Toad| NaCl—06%..... GUS WP S/S 25). 901|'6" 0016 NS (A715. ooel se lise
2 || Ge g Loa ly ee Se LOGS Wie Ol iron lesan on OO} iSO) ae fen ter.
Salle a SOIR eho mel ss SHG) ZZ) i) IO) | ar Zen ed eye) OI Se Peal se
Pe) Ringers ae. ss 8 30|5 25|7 25|7 45|8 05/8 30) 740|—|+ | +
Sate ss Seo gkMtae atte) Hh, tf 9 10/6 00/6 15/6 15|6 30)9 10) 1590) —| + | 4
|
rt IU NEE Se Se cog 6,0 etsy) 2 00) Oman ioe oO) e-ronl meron mO2O0 | ctan Ne tea ete
Blas eRe aes eas SUG US OA ais | Asis) Nhe) US) | Sse Se
8 |Frog Ban Bin ees PO ous SH00)} TOON 230) 220) ZeZon SRO GOON Tat lene
Oy NaCl OG iretnene cs tO rat Bo Ze ROE AS 0.8 Leal ee sO ey Sl eee raed eee
TORN os a Leh ks anSO | 24OON 2: 202 50H 22 SOS 30"), (275!) she |baee|pace
iy ob So on a ae 120)" S30)\1, CONGR TS Sy 120 ISO! ae 38) ates
eee 0.7 % !
12. Calcium pee OIE NO Ae) Ve SOAS coca] COB) |S hae Se
l saccharate — 0.03 %
US eae ss * oN OOS) ANS: sas 25 | coos] VeWI = Se ae
« |§NaCl — 0.7% eae
ile 1Paraxanthin — 0.03% §| Le sy | Sas) Jal (LO) || ES CO) PEO daa shy) Sy ees ee ee
« |(NaCl—0.7% Be 3
15 Ee eeae ones + ADEE) 3X0) IN SO IU Sey | 8) |) 2 By) yl Sea) eal oe
« |§NaCl— 0.7% : z peal
16 | Choralband—1% . .{|9 50/0 24/0 35/0 35]0 33]0 50] | 25)+] +/+
« |{NaCl—0.7% t
17 \Chloralbacd’ 0.33% {| 1.05|0 29] 0 36]0 50]0 52/1 05] 120) +}+| +

1 White’s modification: 0.6% NaCl, 0.01% NaHCOs, 0.01% CaCl. 0.0075% KCI.
2 Howell’s modification : 0.7% NaCl, 0.026% CaCly, 0.03% KCI.

3 Not ascertained.

+ Heart continued to beat long after the conclusion of the experiment.

134 Witham J. Gres.

of chloralbacid, spasmodic twitching in the limbs was the most notice-
able incident.

At the end of the experiments with the solutions containing para-
xanthin and chloralbacid, after the heart had ceased to beat, solution
of calcium saccharate was perfused. In each case this solution
caused the heart to begin beating, and rapidly induced the normal
stroke and rhythm. .

Before passing to the next series, it should be stated that in each
of the preceding experiments the animal became cedematous. Even
those animals in which perfusion took place at the lowest possible
pressures, and for the shortest intervals, showed unmistakable signs
of oedema.

It was impossible to remove entirely the blood-corpuscles from the
capillaries in the heart and brain, even when the perfusion was con-
tinued uninterruptedly for eight hours, and as much as_1590 c.c. of
fluid had slowly passed through the body. In all cases the fluid
expressed from the heart and brain contained an appreciable number
of red and white corpuscles.

In most of the experiments, when the heart had come to a stand-
still after continuous irrigation with physiological saline solution, also
Ringer's solution, rhythmical contractions could be promptly induced
by perfusing Schiicking’s solution. This result was obtained even
when mechanical and electrical stimulation had failed to restore the
normal beating.

Experiments on rabbits. — We report the results of thirteen exper-
iments on rabbits. In this series we used all of the so-called “ in-
different ” solutions already mentioned.

Considerable difficulty was experienced in our efforts to devise a
method which would prevent almost instant death of the animals, and
yet which would speedily result in pronounced anzemia.

Ligaturing the arteries to the brain, before or simultaneously with
the beginning of the perfusion, brought on convulsions immedi-
ately. This was the case whether the ligatures were placed about
the arteries in the neck or in the chest. Even when the perfusion
had been begun shortly before the arterial blood was completely shut
off, it still remained impossible to prevent convulsions and quickly
ensuing death.

In Experiments 1-5 (see the table on page 135), the blood-vessels
in the neck were quickly tied as perfusion was begun. In Experi-
ments 6-10, they were tied just above the heart as perfusion was

On the Irritability of the Brain during Anemia. 135

instituted. Experiments 11-13 were carried out by the following
method.

Instead of closing the arteries to the brain, the abdominal aorta,
vena cava, and vena porta were ligated, and the heart’s action utilized
to pump the perfusion fluid through the brain. The warm solution
was directed into the heart by way of one jugular, and passed from
the brain by way of the other. With this method, anemia was
gradually though quickly induced, convulsions were entirely pre-
vented, and life was considerably prolonged.

In all cases, microscopic examination of the fluid pressed from the
brain showed the presence of red corpuscles.

TABLE II.

Cessation of

| reflexes. Time
| after beginning
the perfusions.

Total time of perfusion.

Lung oedema.

Solution used.

Pressure of perfused
of experiment.
fluid at end.

Weight of Animal.
Amt. of solution perfused.
Hemoglobin in exit

Heart beat.
Blood drawn at beginning

130-150
90

NaCl—1%
Calc. sach. — 0.035% §

a as ae ? 3 50 | 110-140 }..
Rabbit serum eas 110-150}..

{NaCl—1% |
(Calc. sach. — 0.035 % ely Sh es es Mey

100

NaCl—1%

Rabbit serum

136 Wilham J. Ges.

The disappearance of functions in these experiments was not at all
regular in the first ten. The events of each experiment transpired
so quickly that it was extremely difficult to note accurately the time
of cessation of each reflex. In the last three experiments respiration
ceased first in one, second in two; the “lid reflex” disappeared first
in two, second in one. In each of the last three experiments, the
‘nose reflex” was the third to disappear. Heart beat was always
fourth in the order of cessation.

Experiments on dogs. — Only two experiments were performed on
dogs. The first was by a method similar to that in the tenth experi-
ment with rabbits. The weight of the dog was 12 kilos. The
pressure of perfusion was 140-150 mm. Hg. The amount of blood
drawn at the beginning of the experiment was 47 grams. ‘The per-
fusion fluid was a 0.7 per cent solution of sodium chloride containing
0.03 per cent calcium saccharate. Perfusion was continued for forty-
two minutes. The volume of fluid perfused was 1125 c.c. The
amount of hemoglobin present in the fluid leaving the jugular vein
at the end of the experiment was 30 per cent of the normal content
in blood.

Reflex responses failed in the following order: (1) lid and nose
reflexes in twenty-six minutes; (2) respiration in forty minutes; (3)
heart beat in forty-two minutes.

There were no convulsions at any stage of the experiment.

In the second experiment, with a small dog weighing only 5 kilos,
200 c.c. of blood was taken, and an equal quantity of horse serum
immediately afterwards was transfused to take its place. This pro-
cess was repeated three times at intervals of half an hour. After the
fourth blood-letting, the dog ceased to breathe, and did not recover
when the new portion of serum was transfused. Aside from varia-
tions in heart action and respiration, no special functional changes
were observed until the end, when respiration suddenly ceased, and
the other functions came to an end about the same time. Death was
neither preceded nor accompanied by convulsions.

III. Summary oF CONCLUSIONS.

The more important conclusions of these preliminary experiments
are that when the brain is subjected to anemia by the process of per-
fusing solutions, such as Ringer’s, Schiicking’s, serum, etc., its func-
tions soon cease. When the anemia is induced rapidly, convulsions

On the Irritabtlity of the Brain During Anemia. 137

ensue. When it is brought about gradually, anaemia may be made
acute without causing the appearance of convulsions.

When anzmia of the brain is produced gradually by the methods
used in these experiments, the functions here referred to cease usually
in the following order:

(A) In cold-blooded animals: (1) respiration, (2) skin reflex,
(3) lid reflex, (4) nose reflex, (5) heart beat.

(B) In warm-blooded animals: (1) lid reflex, (2) respiration, (3)
nose reflex, (4) heart beat.

ON THE FORMATION OF GLYCOGEN FROM GLYCO-
PROTEIDS AND OTHER PROTEIDS:

By LYMAN BRUMBAUGH STOOKEY.
[From the Sheffield Laboratory of Physiological Chemistry, Vale University. ]

| cena since the original experiments of Claude Bernard on the
formation of glycogen from proteids, this theme has been a de-
bated one. The extensive literature on the subject has lately been
collected and discussed in a monograph by Cremer,! and hence need
not be repeated here in detail. Among recent investigators, Schon-
dorff? denies the possibility of direct glycogen formation in the
body from proteid substances which fail to yield a typical carbohy-
drate group on cleavage. In this class are included casein and gelatin,
which do not yield reducing bodies when they are decomposed with
mineral acids. Schondorff’s negative experiments were carried out
on frogs. On the other hand the experiments of Bendix? on dogs
have led him to the conclusion that in mammals glycogen formation
follows the feeding of proteids (like casein) free from a carbohydrate
constituent as well as of those (like ovalbumin) which contain the
latter in their molecule. In fact, it appears from his protocols that
the carbohydrate-free proteids are even better glycogen formers
than the glycoproteid ovalbumin; and Bendix suggests that the
amido-carbohydrates which are obtainable from the proteids on
decomposition, may be unsuited for utilization as carbohydrates in
the organism.

Although the formation of sugar from proteid in the organism has
long been indicated by the observations of both physiologists and
clinicians, and the possibility of glycogen formation from the same
source is strongly suggested by the older experiments of Naunyn, v.
Mering, Kiilz, and others, the problem can scarcely be regarded as

1 CREMER: Ergebnisse der Physiologie, 1902, i, 1, p. 803.
2 ScHONDORFF: Archiv fiir die gesammte Physiologie, 1900, lxxxii, p. 60;
1902, Ixxxviil, p. 339.
® BENDIX: Zeitschrift fiir physiologische Chemie, 1901, xxxil, p. 492.
138,

On the Formation of Glycogen from Glycoproterds. 139

conclusively solved.|. The renewed study of the proteids in recent
years has made it clear that these substances show far greater differ-
ences in chemical structure than was formerly assumed. For many
of them it has been possible to demonstrate the existence of a carbo-
hydrate nucleus which is apparently lacking in a smaller number?
Accordingly it was to be expected that a reinvestigation of the rela-
tion of the purified and better identified proteids to sugar formation
in the body might bring new suggestions. Thus the glyconucleopro-
teid of the pancreas yields a pentose, I-xylose; ? while chitosamin:has
been identified as a derivative of other compounds, such as the
mucoids, the proteids of cartilage, egg-white, and blood-serum. The
6-carbon carbohydrates which arise from these sources are for the
most part nitrogenous; indeed the biological distribution of amido-
sugar groups is apparently far more extensive than has heretofore
been assumed.

The experiments of Fabian‘ and of Offer and Fraenkel ® have indi-
cated that chitosamin introduced as such into the body is in great
part eliminated again unchanged. Glycogen formation was not ob-
tained. For the glycogenic function of the pentoses the evidence is
also still uncertain.® But it is conceivable that the behavior of these
different carbohydrates when built up in the complex molecule of a
proteid may be quite different from that of the isolated cleavage pro-
ducts. The transformation of chitosamin into the non-nitrogenous
dextrose is by no means a physiological impossibility. Furthermore
Loew’ has pointed out that, theoretically at least, the formation of
glycogen from proteid is conceivable independent of the existence of
any preformed carbohydrate group. scat processes may come
directly into play.

The present experiments by the writer were undertaken some time
ago to determine the influence of true glycoproteids (in the broader

1 Compare the criticisms of PFLUGER and his school.

2 Cf. LANGSTEIN: Ergebnisse der Physiologie, 1902, i, 1, p. 63.

3 NEUBERG: Berichte der deutschen chemischen Gesellschaft, 1902, xxxv,
p- 1467.

4 FABIAN: Zeitschrift fiir physiologische Chemie, 1899, xxvii, p. 167.

5 OFFER and FRAENKEL: Centralblatt fiir Physiologie, 1899, xiii, p. 489.

6 Cf. CREMER: Zeitschrift fiir Biologie, 1901, xlii, p. 428; SALKOWSKI: Zeit-
schrift fiir physiologische Chemie, IgoI, xxxii, p. 393; NEUBERG and WOHLGE-
MUTH: Zeitschrift fiir physiologische Chemie, 1902, xxxv, p. 1; FRENTZEL:
Archiv fiir die gesammte Physiologie, 1894, lvi, p. 372.

7 LoEw: Hofmeister’s Beitrage zur chemischen Physiologie, 1902, i, p. 567.

140 Lyman Brumbaugh Stookey.

sense) on glycogen formation. For this purpose the substances
under consideration were fed to animals made glycogen-free as far
as possible by starvation. A pentose-yielding nucleoproteid of the
pancreas, ovomucoid from egg-white, and the so-called ‘ chondrin,”
or hydrated cartilage, were used. In addition, experiments were con-
ducted with the syntonin of muscle, casein and its salts, and leucin.
The trials were all made upon fasting, full-grown hens. In view of
the difficulty in obtaining large quantities of thoroughly purified com-
pounds for feeding purposes, the selection of a relatively small animal
seemed desirable. The experimental studies of Prausnitz! and Her-
genhahn2 have indicated that, after four to six days’ starvation, hens
are practically free from glycogen — or at any rate that the liver
glycogen has usually for the most part disappeared. In the muscles,
glycogen may apparently persist for a longer period. How difficult
it is to provoke the disappearance of the last traces of glycogen by
starvation alone has again been emphasized lately by Pfliiger and his
co-workers. Indeed it has been suggested that a formation of glyco-
gen may be induced by increased proteid metabolism during starva-
tion. It is difficult, therefore, to make certain of the glycogen
content of any animal at the beginning of a feeding trial; and small
increases above the average figures for starving animals are to be
attributed to the experimental conditions only with great reserve.
On the other hand, negative results ought likewise to be inter-
preted with great caution, wherever the extent of absorption and
utilization of the material fed has not been ascertained by appropriate
investigations. It is well to point out these limitations of experiments
of the character to be described, since they apply to many published
researches on this subject.

Method of experimentation. — Following an old suggestion of Kiilz
that exposure to cold rapidly decreases the content of liver glycogen
in rabbits, the hens were kept without food in a very cold room for
at least four days. The substances fed were mixed with a little water
and moulded into small pellets of about one gram in weight, in which
form they were fed. Water and sand or crushed marble were always
available in the feeding periods, during which the hens were again
kept at room temperature. When the feeding extended over more

1 PRAUSNITZ: Zeitschrift fiir Biologie, 1890, xxvi, p. 377.

* HERGENHAHN: Zeitschrift fiir Biologie, 1890, xxvii, p. 215. KU1Lz found 0.9
per cent in some instances, however. See Centralblatt fiir Physiologie, 1899, iv,
p- 788.

On the Formation of Glycogen from Glycoproteids. 141

than one day, a salt mixture containing sodium chloride, sodium
phosphate, and ferric chloride was also fed in small quantities. The
glycogen was separated by the Briicke-Kiilz method, and, after hydra-
tion to dextrose with 2.2 percent hydrochloric acid, was estimated by
the Allihn gravimetric copper method. In a few cases, estimations
were made on muscle tissue removed from the breast; usually the
liver alone was examined. The alimentary canal, including the crop,
was inspected after the death of the hens, in order to ascertain some-
thing regarding the extent to which the materials fed remained
unutilized. Control trials indicated that under the conditions stated
the glycogen had practically all disappeared from the liver of the
starving animals, and only traces were left in the muscles of those
studied in this respect. The loss of weight during the starving period
was always noted in order to make certain that the animals had not
succeeded in obtaining food. They were killed at varying intervals
after the feedings, the duration of which was likewise varied. In
general fifteen to eighteen hours elapsed after the last feeding.

The preparation and nature of the substances fed may briefly be
outlined. Ovomucotd was obtained from the concentrated filtrates of
coagulated diluted egg-white by precipitation with alcohol. It was
redissolved in water, and reprecipitated, until the concentrated wash-
ings were sugar-free, as indicated by the absence of any reduction
with Fehling’s solution. The final sugar-free products readily
reduced the latter, after being boiled with mineral acids. Miiller and
Seeman were able to obtain thirty per cent of glycosamin from
ovomucoid.! The g/yconucleoprotetd of the pancreas was prepared by
comminuting the fresh pancreatic glands of sheep, boiling for fifteen
minutes in five times their weight of water, and filtering. From the
filtrate, fat was removed after cooling; and acetic acid was added to
precipitate the nucleoproteid. The latter was filtered off, washed
with water and alcohol, dried, and reduced to a powder. As already
mentioned, Neuberg has identified the carbohydrate derivative of the
pancreatic nucleoproteid as |-xylose. The “chondrin” fed was pre-
pared by hydration of the carefully cleaned tracheal cartilages of sheep.
The material was dried and reduced to a powder. The nature of the
reducing substance obtainable from cartilage is somewhat in doubt,
although it is generally regarded as identical with chitosamin.?
Syntonin was obtained from thoroughly washed muscle tissue (lean

1 Cf. COHNHEIM: Chemie der Eiweissk6rper, 1900, p. 266.
2 Cf. LANGSTEIN: Ergebnisse der Physiologie, 1902, i, 1, pp. 80-83.

Lyman Brumbaugh Stookey.

142

‘Aep yyanoj oy} uo [I sem
uay ayy, ‘uasooA[Ss Jo quad 19d /'Q pourey
-uod ajpasnu ayy, “Atuo uondsosqe [erased
peieorpur joes, Arejuaulje ay} Jo uonip
-uod 94} pue prajoid awos poureyu0o dog

‘dois ay3 ut ptajoid ON
‘ep jsey uo jjamun poaivadde
usy fulIpuoYyd s[qeiapisuod paureyuod do19

‘jewuou yns {Ayduia doy
‘Aep yse[ uo pjamun paivadde
uay fULIpUOYS s[qviapisuod pouteju0d doig

‘dois Ul pooj ON

‘dois ut pooy oN

‘doi ul pooj ON
‘pooj jo saov1} Aju pourezuod dog
*jeutiou ynd £ Aydwua ATjeooesd dor9g
‘doi WOAF PI1BAOIIL PLOONUIOAO “SUS XI
‘Aydura Jeuvo Arejuawiye pue doi9
‘Ayduia yeueo Arequsuye pue doi
‘do19
dY} WOIZ PeaI9A0I91 ploonuIoAO ‘swIs UdAaS

‘do1o
ay} WOIJ palaAode1 ploonwoAo ‘sws dA

*queo red

90R1}
STO
8+0'0
auou
S8+0°0
+0'0
FET 0
auou
180'0
2081}
juou
S00
a0ety
20R14

9OPR1}
“Smeg

O¢ET

ie
aa

=p y

SL

n
|
3
H
8)

3

90+

I

”

uruojuds

”

ulmpuoyd

”

prazo.1doajonu
svoiue

”

proonwoag

"SUOTJVAIBSYO

“IDATL 9U3
UL punoj uasood4][5

“OAT OY} JO IYSI9. NM

ye yy s1am-Apog

“yjeap pure Surpass}
yse[ U9aM4JAq [VAIOIUT

‘Sulpaaj jo uoneing

‘pay saourysqns
jo Aytjuenb [ejo 7,

“‘SUISeJ JO uOTLING

‘yys1am-Apoq [euly

‘yystda-Apoq [entuy

"P9F
sgourysqus

‘ua
-119dxa Jo
Joquin NT

43

I

zdas.

on of Glycogen from Glycoprote

On the Formati

‘uasooA[S
jo yuoo sad ZQQ peutezu0d aposnut ayy
‘doid 9Y} WOIF pd1aACIeI SEM UTNE, aWIOG

‘uasook[s Jo yuao
Jad sy paureyuoo aposnw ayy, ‘dos ayy
WOIF PIIBAOIAI 919M ULONIT Jo ‘swIs OT noqy

“uasoo4]S Jo JU90 Jad 7G paureyuod aposnw aq],
"70R1}

Ale} USULITe UT Puno; Sanpisa1 pooj ou {uss0o

-{S jo quad sad 69° pourejuO afosnut sy],

yJeMun usy fpasnun Aied poo
*pooy 9WIOS poulezUOd [Ys dors oy} fuasos
-AJS Jo quao 19d 969 pautezuo0. apsnw sy 7
*pooy aWIOSs pautezUod [Ys dor9 ayy {uasoo
-AJS jo quso rod 6 pautejuoD aposnw ou T,

‘paAIasqo senpisal pooy ON

*peAtgsqo sanpisat poof ON

*‘paariasqo senpiset pooj oN
‘punoj sanpisoi pooj oN
‘pazt[iyn AjjuapiAe poox {sno108ta uazy
‘poof Jo uondiosqe [enieg

‘do1o
Ul poof} ‘feayIeIp fuojdiosqe yooaziodwy
‘doo ut pooy ! ajatduoour uondiosqy
‘emyslieip ! uodiosqe ajatduioouy

‘yeuvo Arejusute ul poog Jo Ajuo saov1y,

‘uondiosqe jensed Aju

ouou
ouou

£90
ouou
euou
ouoUu
ZITO
959'T

90e1}

EDS
‘ep

20u0

ulna’

”)

|
ISOILYIOVS pur |

Ulased WINIPOS

ulasvo WINTpOS

”

”

ulaseg

IAXXX

AXXX
IITAX

IAX
AX
AIXXX
WIXXX
ITX XX
IX XX
XXX
XIXxX
ITIAXX
AXX
AIXX
ITIXX

IIXX
IXX
XX

144 Lyman Brumbaugh Stookey.

beef) by extraction with dilute hydrochloric acid. It was precipi-
tated from the acid solution by addition of sodium hydroxide, then
washed free from alkali and salts, and dried. On account of its
insolubility the syntonin was formed into pellets by the addition of a
little gelatin to hold together the fine particles. In the preparation
of casein (caseinogen) a commercial product was treated with hot
alcohol and ether to remove fat, and then repeatedly extracted with
hot water, until the concentrated washings no longer reduced Fehling’s
solution. The milk sugar was thus completely removed. To facili-
tate the formation of pellets for feeding, a few drops of dilute sodium
hydroxide solution were added to the powder. The large number of
negative and unsuccessful results with this product suggested the
preparation of a more soluble salt. Sod¢wm casein was accordingly
prepared by suspending purified casein in warm water, and adding
sodium hydroxide until solution was effected. An alkaline reaction
was avoided. The material was evaporated to dryness, pulverized,
and fed in the usual way. Casein fails to yield typical carbohydrate
derivatives on decomposition. The /eawczz fed was a crude crystalline
product obtained by the auto-digestion of pig’s pancreas.

Protocols of experiments. —— The following typical protocol] will
illustrate the character of the data ascertained in each experiment.

Experiment IV. Ovomucoid feeding. — Weight of the hen before starva-
tion: 1210 gm. Duration of starvation: 4 days. Weight after 4 days:
960 gm. Amount of ovomucoid fed over a period of 2 days: 30 gm.
The hen was killed 14 hours after the last feeding. Weight: tooo gm.
The crop was practically empty ; the condition of the alimentary canal
was apparently normal, and suggested that absorption had taken place.
Weight of the liver: 20 gm. Glycogen content of the liver: 0.05 gm.
= 0.25 per cent.

The more important data have been summarized in the table on
pages 142, 143.

Discussion of the results. — Earlier experiments! have indicated
that the glycogen content of the liver of hens reaches its maximum
within twelve to twenty-four hours after feeding, and then falls
rapidly again, so that at the end of thirty-six hours it may become
almost nil. This led in the present trials to the plan of killing the
animals within the periods selected above. The marked diminution
in body weight noted in every case, and the numerous experiments

’ Cf, PRAuSNITZ: Zeitschrift fiir Biologie, 1890, xxvi, p. 377.

On the Formation of Glycogen from Glycoproteids. 145

which showed only traces of glycogen in the liver, bear evidence of
the efficiency of the starvation process in removing this carbohydrate.
With reference to the food it may be added that particular care was
devoted to obtaining materials which should be free from contaminat-
ing sugarsorfat. The positive results cannot, therefore, be attributed
to the presence of the latter compounds.

The outcome of the feeding experiments with those substances
which yield carbohydrate cleavage products — ovomucoid, pancreas
nucleoproteid, chondrin — scarcely permits any positive conclusion
to be drawn. In some of the more satisfactory trials, where the
period of feeding was more prolonged, and the utilization of the food
was apparently better, appreciable quantities of glycogen were found.
These do not, however, exceed in amount the maximum glycogen
content (0.97 per cent) which has been observed in the liver of the
fasting hen by Kiilz;! although, like the writer, most other investi-
gators have noted very small quantities only. Prausnitz, for example,
found 0.06 — 0.13 per cent of glycogen in the liver of the hen after
four days’ fasting. The feeding of simple proteids — syntonin, casein,
and its sodium salts — failed to yield an increase of glycogen in the
liver when a single dose was given. But after trials lasting several
days, during which considerable quantities of casein were ingested, an
accumulation of glycogen too large to be attributed to any residual
store in the liver was repeatedly found. (Compare Experiments
XXII, XXVIII, XXXI, XXXII, XXXIII, XXIV.) The quantities
range from 2.3 to 5 per cent, and were usually obtained in those trials
in which the animal remained vigorous and a utilization of the food
given was indicated. These observations are thus in accord with the
results which Bendix? obtained after feeding casein to dogs.

It has already been pointed out that negative results in experiments
like the present ones do not necessarily indicate the incapacity of
the substances fed to promote glycogen formation. Nothing short of
an estimation, in each case, of the actual extent of absorption and
metabolism would permit a final interpretation in those cases where
glycogen failed to appear in the liver. The experiments (XV, XVI,
XVII) in which proteid and sugar were fed simultaneously, were
undertaken to see what effect would be produced by quantities of

1 KUxz: Centralblatt fiir Physiologie, 1890, iv, p. 788. Cf. note 2, p. 140.

2 BENDIX: Zeitschrift fiir physiologische Chemie, 1901, xxxii, p. 479; 1902,
Xxxiv, p. 544. Compare the critique of CREMER: Ergebnisse der Physiologie,
1902, i, 1, p. 874.

146 Lyman Brumbaugh Stookey.

carbohydrate comparable with those which might be liberated from
the glycoproteids in the amounts fed. In two of the three trials the
amounts of glycogen found in the liver (1.5, 1.6 per cent) are not in-
considerable, and agree in general with previous experiments with cane
sugar.! The negative results with the glycoproteids are interesting
in view of the similar outcome of the studies by others on chitosamin
and pentoses. .

With the possibility of sugar- and glycogen-formation in the organ-
ism from carbohydrate-free proteids made probable, it is of interest
to learn the intermediate stages in this transformation. What con-
stituent groups of the proteid molecule are the antecedents of the
newly formed carbohydrate? It has been suggested that the leucins
(C,H,,NO,) constitute the most important intermediary products ;
and it is known that the ordinary proteids yield a large proportion —
fifty per cent or more — of leucin on decomposition. The transfor-
mation of an amido-acid like leucin into a 6-carbon carbohydrate is
not a theoretical impossibility.2 To test the hypothesis, Cohn ? fed
leucin to starving rabbits. His observations are by no means conclu-
sive, although they point toward an increased glycogen content in the
liver; but Simon * has repeated the experiments with entirely negative
results. It has been pointed out by Bendix® that if leucin is an ante-
cedent of glycogen in the body, then all proteids which yield this
amido-acid in abundance ought to induce glycogen formation. This
is, however, not the case, as the differences between the results of
gelatin feeding and casein feeding indicate. The two experiments
presented above by the writer do not, at least, speak against the
possibility under discussion; nor do they justify any far reaching
statements. The physiologist must look forward to a more profound
acquaintance with the chemistry of the proteid molecule before the
final word can be spoken.

The writer desires to express his obligation to Professor Lafayette
B. Mendel for the suggestion of the subject for investigation, and for
criticism.

Cf. PRAUSNITZ: Loc. cit.

Cf. MULLER: Zeitschrift fiir Biologie, 1got, xlii, p. 549

Coun: Zeitschrift fiir physiologische Chemie, 1899, xxviii, p. 211.
SIMON: Zeitschrift fiir physiologische Chemie, 1902, xxxv, p. 315.
BENDIX: Zeitschrift fiir physiologische Chemie, 1901, xxxii, p. 492.

THE SHARE OF THE CENTRAL VASOMOTOR. INNER-
VATION IN THE VASOCONSTRICTION CAUSED
BY INTRAVENOUS INJECTION OF
SUPRARENAL EXTRACT.

By S. J. MELTZER An» CLARA MELTZER?!
[From the Rockefeller Institute for Medical Research.]|

LIVER and Schaefer? ascribed the vasoconstricting effect of

suprarenal extract to its action upon the blood-vessels directly
and not to an action upon the vasomotor centres. They based their
claim upon the following considerations: first, that a rise of blood-
pressure occurs even after section of the cord below the vasomotor
centre; second, that a vasoconstriction takes place in a peripheral
organ even after that organ is deprived of its nerves; for instance,
in the foreleg after the brachial plexus is cut. The validity of the
first consideration has already been a matter of discussion. Thus
Szymonowicz* and Cybulski claimed that in their experiments, section
of the cord did interfere with the vasoconstricting effect of the extract.
Furthermore, Velich* and others called attention to the fact that
simple section of the cord does not exclude the influence of all the
vasomotor centres of the spinal cord. To meet this requirement,
Velich as well as Biedl® have thoroughly destroyed or extirpated the
entire cord. Both authors report that even under these circumstances
suprarenal extract caused a rise of blood-pressure. Velich, however,
admits that in many experiments he did not succeed in completely
destroying the spinal cord. He had only three successful experi-
ments. In these, however, the effect was apparently not so prompt
as in the ones in which the spinal cord was only cut. According to
Velich, after destruction of the cord, a much larger dose of the supra-
renal extract is required to bring about a similar rise in the blood-
pressure, than after simple section of the cord. From this it would

Research Scholar of the Rockefeller Institute.
OLIVER and SCHAEFER: Journal of physiology, 1895, xviii, p. 230.
SzZYMONOWICZz: Archiv fiir die gesammte Physiologie, 1896, Ixiv, p. 97.
VELICH: Wiener medizinische Blatter, 1896, xv—xxi, p. 227.
BiEDL: Wiener klinische Wochenscrift, 1896, p. 157.

147

1
2
3
4
5

148 S. J. Meltzer and Clara Meltzer.

appear that the integrity of the cord is indeed a factor in the success-
ful action of the suprarenal extract upon the blood-pressure.

With regard to the second experimental proof of Oliver and
Schaefer, namely, that a vasoconstriction takes place in a peripheral
organ even after its nerves have been cut, here, too, the objection
was raised that not all nerves of these organs had been cut; for in-
stance, section of the brachial plexus does not exclude the entire
innervation of the foreleg. But above all, it seems to us that the
method which the authors employed is not capable of furnishing con-
clusive evidence. They enclose a peripheral organ, kidney or forearm,
for instance, in a plethysmograph, and assume that each registration
of a diminution in the volume of this apparatus means a constriction
of the blood-vessels of the organ within it. They themselves state,
however, that the constriction of the arterioles can often lead toa
dilatation of the larger and middle-sized arteries which will produce
an increase in the volume of the plethysmograph. In other words, a
constriction of the small arteries may lead either directly to a dimi-
nution in the volume of the plethysmograph, or indirectly, by subse-
quent filling up of the larger vessels, to an increase in the volume.
It would seem that under these circumstances we can hardly draw
positive conclusions as to the actual condition of the organ from
tracings obtained by the plethysmograph. In fact, of the thirteen
plethysmographic tracings of the forearm, given by Oliver and
Schaefer in their paper, eight show an increase, while only five show
a decrease in the volume after intravenous injection of suprarenal
extract.

The foregoing analysis shows that the evidence in favor of the theory
of Oliver and Schaefer is by no means complete. A further study
of this subject by other methods is therefore desirable.

Direct inspection of the blood-vessels has already been mentioned
by Oliver and Schaefer as one of the appropriate methods of study-
ing the effect of intravenous injection of suprarenal extract. It
occurred to us that the rabbit’s ear would be a favorable object for
such a study. The condition of the blood-vessels there is easily dis-
cernible, and the vasomotor influences of central origin can be
satisfactorily removed. Furthermore, the conditions of the blood-
vessels of the ear deprived of the vasomotors, can easily be compared
at any stage with those of the normal ear; and this can throw light
on the share which the central vasomotor mechanism may have in
the vasoconstriction caused by the suprarenal extract.

Vasoconstricting Effect of Suprarenal Extract. 149

Such a series of experiments we have made on a large number of
animals, and we shall here report our results.

We employed exclusively adrenalin chloride in various dilutions.
The injections were made either into the external jugular vein, the
femoral vein, the inferior vena cava, or into the marginal vein of one
of the ears. In the first series of animals, the cervical sympathetic
was cut or resected either on the right or on the left side, and the
injections were made either soon after the operation or some time
later. The following are a few abbreviated protocols of that series
of experiments.

Experiment XIV. April 238, 1902. — Brown, medium-sized rabbit.

5:30 P.M. Resected about 2 cm. of cervical sympathetic on left side;
vessels of left ear at once very much dilated.

6.28. Injected into marginal vein of right ear 1 c.c. of I : 10000
adrenalin.

6.30. Both ears completely blanched.

6.35. Right ear-vessels begin to dilate.

6.42. Same, fully dilated.

6.40. Left ear-vessels begin to dilate, progressing slowly.

7.00. Left returned to normal.

Experiment XTX. May 7, 1902. — Large, white rabbit.

4.00 P.M. Left sympathetic resected, both ears very congested. Right
femoral vein exposed.

4.23. Injected into vein 1.2 c.c. of r : 2000 adrenalin.

4.23.30. Right ear pale.

4.25. Left ear beginning to blanch.

4.28. Fine vessels of right ear begin to appear.

4.29. Vessels of right ear fuller than those of left.

4.36. Left ear still more blanched than before.

4-54. Right ear becomes paler again.

5:45. Nochange. On taking ether and struggling, the right ear-vessels
become very full, left ear remains pale.

Experiment XXVI. June 19, 1902. — Brown rabbit, 1500 grammes.
4.50 P.M. Right sympathetic cut.

5-00. Injected into marginal vein of left ear 1.2 c.c. of 1 : 5000
adrenalin.

5-00.30. Left ear pale.

5-01.30. Right beginning to blanch.

5.02. Entire right ear pale, except marginal vein.
5-05.30. Left vessels beginning to dilate.

150 S. J. Meltzer and Clara Meltzer.

Experiment XX VI— (continued).

5.08.30. Right central artery becoming very slightly dilated — dilata-
tion not increasing.

5.15. Left ear-vessels, after having been well dilated for about eight
minutes, become narrower again.

6.35. Nochange. Right ear distinctly paler than left.

8.00. No change.

June 20, 9.30 A.M. Vessels of right ear more dilated than those of left,
but not as fully dilated as on the previous day soon after operation.

Experiment XX VIL. June 19, 1902. — Black and white rabbit, 1500 grammes.
5.30. Left sympathetic cut.
5:40. Vessels of left ear very full, of right, medium.
5-41. Injected into right marginal vein o.5 c.c. of 1 : 5000 adrenalin.
5-41-30. Right entirely pale, except a small central vessel.
5:-43- Left beginning to blanch.
5-46. Right getting fuller.
5-47. Left, pallor increasing.
5:55. Left, pale except for large vessels.
6.08. Right vessels as much dilated as before injection.
6.30. Left, still pale.
8.00. Left ear redder than right.
June 20, 9.35 A.M. Left wider than right, but not as fully dilated as
after operation.

Many similar experiments were made. The results were the same
in every case. The degree of vasoconstriction which was caused
by the intravenous injection of adrenalin, was in all cases about
the same on the operated as on the normal side. Here we have
to mention that Velich reports a single experiment he made on a
rabbit in which the right sympathetic nerve was cut, and in which
the cadaverous pallor of the ears, following intravenous injection of
suprarenal extract, was equal on both sides. In our experiments,
however, we have observed some significant differences in the course
of the vasoconstriction of the two ears. But before entering upon a
detailed discussion of our observations gathered from the above-men-
tioned series of experiments, we have to report the results of other
series of experiments, the execution of which became necessary
through the following observation. A study of the effects of cutting
and stimulating the third cervical nerve revealed to us ! its considerable

1S. J. and CLARA MELTZER: This journal, 1903, ix, p. 57.

Vasoconstricting Liffect of Suprarenal Extract. 151

importance as a vasomotor nerve for the ear, which in many instances
exceeds that of the sympathetic nerve. Hence section of the cervi-
cal sympathetic nerve alone does not completely deprive the ear-
vessels of their central innervation.

The following abbreviated protocols are examples of a series of
experiments we made on rabbits, in which the sympathetic as well as
the third cervical nerves were cut on the same side.

LEixperiment XXIX. Sept. 20, 1902. — Large gray rabbit.

5:00 P.M. ‘Third cervical and sympathetic cut on the left side. Can-
nula in inferior vena cava. Animal lost some blood. Left ear-vessels
considerably fuller than right.

6.20. Injected through cannula 1 c.c. 1: 2000 adrenalin. In a few
seconds, right ear blanched completely.

6.21. Left ear begins to blanch — blanches slowly.

6.24. Right begins to fill; left still blanching, especially large central
vessel.

6.27. Left beginning to fill.

6.30. Both moderately filled, left, a trifle more than right.

6.35. Second injection —1 c.c. 1 : 2000. In a few seconds right
blanched.

6.36. Left begins to blanch — blanches slowly.

6.38. Both begin to fill.

6.45. Third injection, 1.2 c.c. of 1 : 2000. Immediate complete
blanching of right ear, followed closely by complete blanching of left ear.
Blanching on right side lasted six minutes, then faint filling, which increased
to normal within ten minutes; left blanching lasted ten minutes, vessels
filled in a few minutes, but not to their former width; blanched and filled
again several times. Fourth injection of 1.2 c.c. pure adrenalin (1 : 1000).
Complete blanching of right ear in a few seconds, of left, half a minute
later. Rabbit killed after fifteen minutes, before ears returned to normal.

Experiment XXXT. Oct. 5, 1902. — Brown rabbit.

5-30 P.M. Sympathetic and third cervical cut on the left side.

6.30. Injected through cannula in left jugular vein 1.2 c.c. adrenalin
(1 : 1000).

6.30.30. Right ear pale. Fifteen seconds later, left ear beginning to
blanch.

6.34. Right beginning to fill; left getting still paler.

6.35. Right, as full as before injection.

6.40. Right, fuller than before injection ; left, pale, slight rhythmical
dilatation in small area around bifurcation.

6.49. Right, paler again, condition as before injection.

152 S. J. Meltzer and Clara Meltzer.

Experiment XX XI— (continued ).

7.05. No change in either ear.

8.15. Left ear-vessels dilated again.

8.20. Second injection of 4.2 c.c. undiluted adrenalin. Right ear
blanched immediately, half a minute later, left beginning to blanch from
below upward.

8.25. Both ears begin to fill, left, slight rhythmical dilatation in the
central vessel, while the marginal vessels getting still narrower.

8.28. Right, as before injection.

8.30. Right fuller than before injection.

8.43. Right, again as before injection ; left, still pale.

8.58. Small vessels of left ear becoming visible.

9.45. Central vessel of left somewhat dilated, but not nearly as before
injection. Marginal vessel fills up from below.

10.00. Left, almost as full as before injection.

In view of the well-known statement of Claude Bernard! that the
excision of the superior cervical ganglion causes greater paralytic
effects than simple section of the cervical sympathetic nerve, we
deemed it desirable to perform a few experiments in which, instead of
simple section of the cervical sympathetic, the superior cervical gang-
lion was removed. The following experiment is an example:

Experiment XXX. Sept. 22, 1902. — Large gray rabbit.
5.00. P.M. Ganglion removed, and third cervical cut on the right side.
Cannula in inferior vena cava below the renal veins.
6.00. Injected 1 c.c. 1 : 10000 adrenalin.
6.01. Both ears moderately pale.
6.03. Left filling faintly.
6.04.30. Right, central vessel filling from below, rest of ear pale.
6.06. Both ears as before injection.
6.10. Second injection, 0.6 c.c. 1 : too0o adrenalin.
6.10.30. Left ear getting paler.
6.11.30. Smaller vessels in right beginning to disappear.
6.12. Entire right ear pale.
6.13. Left filling.
6.14. Right filling.
6.15. Right fuller than left.
6.19. Third injection, 1.2 ¢.c. I : 2000 adrenalin.
6.20. Left, entire ear pale ; right, smaller vessels getting narrower.
6.23. Right, central vessel disappearing.

1 BERNARD: Lecons sur la systéme nerveux. Paris, 1858, ii, p. 492.
~ J ; > ll, p

Vasoconstricting Effect of Suprarenal Extract. 153

Experiment XXX — (continued ).

6.23.30. Left ear getting redder.

6.24. Right, completely pale, left, already a little fuller than before third
injection.

6.26.30. Right, central filling slightly.

6.40. Right, not yet as full as before injection.

6.43. Fourth injection, 1.2 c.c. I : 2000.

6.44. Left paling.

6.44.30. Right paling.

6.45. Left slightly filling.

6.48. Right beginning to fill.

6.50. Left, as before injection.

6.55. Right, not yet as full as before fourth injection.

In a previous paper! we reported that in a few exceptional cases
we found that after cutting the branches which connect the third with
~ second and fourth cervical nerves, the corresponding ear became con-
siderably flushed. In a few experiments we have therefore also cut
these connections in addition to the removal of the ganglion and
cutting of the third cervical nerve. We shall give no protocols of
these experiments ; they read the same as the others already quoted.
Section of these connecting branches did not change the results in
any respect.

We may safely claim that after cutting the third cervical and remov-
ing the superior cervical ganglion, the blood-vessels of the ear are de-
prived of all vasomotor influences of central origin. We know, at least
for the present, of no other path by which vasomotors reach the ear.
The changes, therefore, which we have seen taking place in the ear-
vessels, after the elimination of the above-mentioned nerve and
ganglion, can be ascribed only to the influences of some peripheral
mechanisms. In all our experiments, without exception, an intravenous
injection of a sufficient dose of adrenalin caused a distinct constriction
of the blood-vessels of the ear on the side on which the ganglion was
removed and the third cervical cut. The constricting effect of ad-
renalin in these cases is, therefore, certainly of strictly peripheral ori-
gin. Furthermore, in all of the experiments in which a dose of
adrenalin was employed sufficient to bring out a distinct pallor in the
normal ear, there was no case in which the degree of the pallor on the
operated side was less than on the normal side. Even when small

1S. J. and CLARA MELTZER: This journal, Zoc. czt., p. 57.

154 S. J. Meltzer and Clara Meltzer.

doses were used, which caused only a moderate constriction of the
blood-vessels, lasting a very short time, the degree of the constriction
was in both ears the same, provided, of course, the injection was not
made into an ear-vein. In the latter case, the ear into which the in-
jection was made has shown the greater pallor.

Our experiments certainly justify the conclusion that suprarenal
extract can cause a strong constriction of blood-vessels, which are de-
prived of all extrinsic neurogenic influences. Is the constriction oc-
curring in normal animals due to the local effect,and to this effect
alone? Are the extrinsic vasomotor nerves and centres not affected
at all by the suprarenal extract? The fact that the degree of the
constriction of the blood-vessels has apparently been the same in
both ears, in all cases, would seem to indicate that the central vaso-
motor mechanism is not a factor in the attainment of the degree of
vasoconstriction caused by suprarenal extract. However, our experi-
ments brought to light also some differences in the course of the
vasoconstriction in the two ears. These deviations, we believe, are
capable of shedding light upon the effect which suprarenal extract
might exert upon the central vasomotor mechanisms, and the role of
the latter in the change of the blood-pressure, which is caused by the
extract in the normal part of the animal. The differences relate to
the onset, development, duration, and after-effect of the vasoconstric-
tion in the two ears.

1. The most striking point is the difference in the duration of the
constriction. In every one of our experiments the constriction of the
blood-vessels of the ear on the operated side lasted distinctly longer
than that of the ear on the non-operated side. While the duration of
the constriction on the normal side rarely exceeded six or seven min-
utes, and generally amounted only to four or five minutes, we found
the constriction on the operated side to last often for hours, and in
some instances the vessels did not reach their original size till the
following day. In both ears the duration of the constriction de-
pended upon the amount and concentration of the adrenalin we have
injected in each case. But even with small quantities and weak
solutions, as long as they were capable of producing any degree of
constriction, it lasted perceptibly longer on the operated side.

2. Another instructive point is the difference in the mode of the
disappearance and the after-effect of the constriction. In the normal
ear, after the maximum pallor has lasted a few minutes, the vaso-
constriction disappears evenly in a very short time, and is followed by

Vasoconstricting Liffect of Suprarenal Extract. 155

a moderate but distinct congestion of the ear. The blood-vessels are
then perceptibly wider than before injection. The dilatation lasts a
few minutes, its duration being mostly longer than that of the con-
striction. In the ear on the operated side, we see the maximum con-
striction, after having lasted for some time, at first giving way only
slightly, the larger vessels becoming faintly recognizable. In this
state the ear remains stationary again for some time, and then begins
to fill up gradually, and very slowly attains its original appearance.
In no case have we noticed the ear-vessels of the operated side be-
coming wider than their original size, not even in such experiments
in which the injection was made a long time after the section of the
nerves, and the vessels, before the injection of adrenalin, were there-
fore not much wider on the operated, than on the unoperated side.

3. In all of the experiments, the blanching of the ear on the oper-
ated side began distinctly later than on the normal side, provided the
injection was not made into the marginal vessel on the operated side,
or the dose employed was not too large. The latent period for the
normal side is very short, usually only ten to fifteen seconds; it varies
slightly with the strength of the dose, and also with the place of in-
jection ; for instance, as a rule the constriction appears much sooner
after the injection into the jugular vein, than after injection into the
inferior vena cava. On the operated side, the latent period was in all
instances distinctly longer, the difference varying from fifteen seconds
to two minutes.

4. Besides the difference in the length of the latent period, there
was invariably also a difference in the time during which the vaso-
constriction attained its maximum. In the normal ear there were in
all cases not more than a few seconds between the onset of the con-
striction and the complete blanching of the ear. In the ear on the
operated side, the development of the blanching was in nearly all cases
a very slow process. It was a frequent occurrence to see the blanch-
ing on the operated side still progressing, while on the normal side
all signs of a constriction had already completely disappeared.

The length of the latent period and the slow development of
the constriction on the operated side cannot be ascribed, at least
entirely, to the fact that before the injection the blood-vessels of
the ear on that side were more dilated than on the normal side, and
therefore more work had to be performed before the same degree of
blanching could be attained. The same slow onset and development
have been observed in cases in which the injection was made long

156 S. J. Meltzer and Clara Meltzer.

after the operation, and the dilatation was therefore no longer notice-
able. They were also present in experiments in which a second
injection was made while the ear-vessels on the operated side, as a
result of the foregoing injection, were not yet as wide as those on
the normal side.

The several differences in the course of the vasoconstriction of the
two ears can be ascribed only to the differences in their vasomotor
innervation, and indicate clearly that in the normal animal the central
vasomotor mechanism plays some part in the causation of constriction
of the blood-vessels by intravenous injection of suprarenal extract.
As to the interpretation of these differences, we shall discuss first the
divergence in the after effect and the duration of the constriction.

The fact that the vessels of the ear on the normal side, after the
disappearance of the constriction, become for a while wider than
they had been before the injection, can be explained by the assump-
tion that in the normal animal, at a certain phase of the con-
striction, vasodilating mechanisms are efficiently stimulated. That
such a dilatation is not observed on the operated side indicates,
furthermore, that the vasodilatation on the normal side is of central
origin ; hence, its absence on the operated side. This very assump-
tion also readily explains the difference in the duration of the
constriction. The constriction on the normal side is normally
abridged, we may assume, by the central stimulation of vasodilators.
This stimulation is absent on the operated side, hence the prolonged
constriction.

The fact that in many blood-pressure experiments no pressure-lower-
ing after-effect has been observed, is not a serious contradiction to
our assumption. The dilatation of the vessels of the normal ear
which we have observed as an after-effect was indeed comparatively
slight. The main activity of the vasodilatation might consist in
reducing to normal the vasoconstriction, which end is perhaps not
simultaneously attained in all parts of the body, on account of some
differences in the local conditions. In some parts, with an early
reduction to normal, there might even occur a slight abnormal dilata-
tion; but in measuring the blood-pressure as a whole, these slight
dilatations are liable to be counter-balanced by some still lingering
constrictions in other parts. Hence the absence of signs of vasodila-
tation when studied by the general blood-pressure from the carotid
artery, while ocular observation of the vessels of the normal ear

Vasoconstricting Liffect of Suprarenal Extract. 157

shows the existence of a moderate vasodilatation as an after-effect
in this peripheral organ.

What causes the central stimulation of the vasodilators at a
certain phase of vasoconstriction brought on by the suprarenal
extract? The hypothesis which appealed to us in the first place is as
follows: The suprarenal extract acts on the vasomotors simply as a
chemical stimulus. It circulates in the blood in a certain concentra-
tion, leaves the capillaries, and attacks the vasomotor centres. There
is no reason to believe that the extract has a special affinity for the
vasoconstrictors — that it attacks the centre for the vasoconstrictors
alone. On the contrary, it is more plausible to assume that it stimu-
lates simultaneously and with equal intensity both antagonistic centres,
the centre for vasoconstriction as well as the centre for vasodilatation.
It is, however, a well-established fact that when vasoconstrictors and
vasodilators are stimulated simultaneously, then a strong stimulus
_ favors the preponderance of vasoconstriction and a weaker stimulus
favors the preponderance of vasodilatation. Now a sufficient dose of
adrenalin, when first introduced into the circulation, represents of
course a strong stimulus, and favors vasoconstriction, hence the
immediate constriction of all the vessels of the body and the rise of
blood-pressure. After a few minutes, however, the dose of the
extract circulating within the blood loses its original strength, either
by oxidation of a part of the extract within the blood, or by elimina-
tion of a part through the kidneys, or by transudation and deposition
of some of the extract into the tissues. The remaining dose now
represents a weaker stimulus and favors the preponderance of vaso-
dilatation. We know now from numerous studies that during the
preponderance of the activity of one set of nerves, the stimulation of
the antagonistic nerves continues in full strength. When, therefore,
the dose of adrenalin becomes reduced, and the vasodilators com-
mence to be favored, their activity is immediately in full sway, hence,
the rather rapid termination of the constriction and the occasional
appearance of a local vasodilatation.

This hypothesis seemed to us to be accessible to an experimental
test. If adrenalin stimulate also vasodilators and weak stimuli favor
their preponderance, it was reasonable to expect to find a small dose
of adrenalin which would cause primarily dilatation of the ear-
vessels. We therefore made a large number of experiments in which
dilutions varying between 1 : 10000 and I : 20000 were injected in
quantities of 0.3 too.6c.c. The outcome of these experiments was,

158 S. J. Meltzer and Clara Meltzer.

however, not entirely satisfactory. We succeeded only in a very few
instances in obtaining a primary dilatation. The following abridged
protocol is an example of such an experiment :

Experiment XXXII. July 7, 1902.— Rabbit operated seven days before,

right sympathetic resected. .

11.58.30 A.M. Injected into marginal vessel of right ear 0.4 c.c. of
1: 20000 adrenalin. Right ear paled, left ear-vessels moderately dilated
and slowly filling up.

12.00. Left ear as full now as just before injection.

12.00.30. Vessels of left ear fully dilated. No initial constriction
noted.

12.01. Right ear-vessels slowly filling up again.

The right ear (operated side) became pale in this case through
the direct contact of the vessels with the injected adrenalin. The left
ear received the extract through the circulation, and the extract was,
therefore, very much more diluted. There was no initial constriction,
on the contrary, the vessels of the left ear (non-operated side) started
to fill up immediately.

In most of the experiments with small doses, however, there was
not such an immediate dilatation after the injection. Either there
was no change at all, or there was no change for a minute or two, and
then a dilatation followed, or finally there was a primary very brief
constriction, followed by a pronounced dilatation of the vessels.
This series of experiments, though not furnishing an absolute proof,
suggests the correctness of the premise in our hypothesis that a
smaller dose of the extract favors the preponderance of vasodilatation.

Our hypothesis is based on the assumption that the suprarenal
extract has no special affinity for either vasoconstriction or vasodila-
tation, but represents simply a general nerve stimulus. However,
Girber! as well as Hunt? stated that they succeeded in obtaining
from the suprarenal capsule a substance which causes only vasodila-
tation. Accordingly, we should have to assume that the extract
we employ does not represent one principle which stimulates both
centres, but contains separate principles for each of the antagonistic
centres. It would offer no difficulty to adapt our hypothesis to this
new conception, if it will find general recognition; but we shall
abstain from discussing these points for the present.

* GURBER: Miinchener medizinische Wochenschrift, 1897, p- 750.
* Hunt: This journal, 1900, iii, p. xviii.

Vaseconstricting Effect of Suprarenal Extract. 159

Regarding the differences in the onset and the development of the
vasoconstriction, we wish to emphasize again our contention that
these differences are not due entirely to the previous greater engorge-
ment of the ear on the operated side. Our experiments seem rather
to demonstrate conclusively that vasoconstriction which is brought
about exclusively by peripheral mechanisms, sets in later and develops
more slowly than in the case where the peripheral organ has the
prompt aid of the central vasomotor mechanisms. As to the reasons
for this difference, we can only offer some suggestions. In the first
place, it is possible that peripheral mechanisms at all times respond
less readily to stimulation than the centres in the cord. There is no
lack of analogies for such an assumption. Furthermore, the tonus of
the vessels is maintained by continuous excitation by normal stimuli
of the vasomotor centres, and not by stimulations of the peripheral
mechanisms, as is evident from the fact that cutting of the sympa-
thetic or third cervical nerve causes vasodilatation. The peripheral
vasoconstricting mechanisms do not take up the control of the tonus
till many weeks after the section of the nerves. It is probable, there-
fore, that in the normal animal artificial stimuli also affect, in the first
place, the central organs, while the peripheral mechanisms are stimu-
lated only secondarily and in lesser degree; hence the tardiness in
their response after the elimination of the vasomotor centres. It is
possible that the peripheral mechanisms gradually acquire a readiness
to respond rapidly to artificial stimulation, just as they gradually
acquire the readiness to respond to the continuous normal stimuli for
the maintenance of a vascular tonus.

Finally, it is possible that the suprarenal extract in the blood has
more obstacles to overcome in reaching the middle coat of the arteries
(the muscular layer or the hypothetical ganglia within or around it),
than in reaching the vasomotor centres within the cord. To reach
the latter the extract has only to pass the endothelial layer of the
capillaries which are normally arranged for such passages. While to
reach the middle coat of the arteries the extract has to penetrate the
complex structure of the arterial intima (endothelial, subepithelial,
and elastic layers), which is normally not constructed for purposes of
transudation of fluids. As to reaching the middle coat by way of the
vasa vasorum, we know nothing of the existence of such vessels for
the smaller arteries.

However this may be, it is a fact that in blood-vessels which have
recently become deprived of their central innervation, the constriction

160 S. J. Meltzer and Clara Meltzer.

sets in late and develops slowly. We have therefore sufficient reason
to assume that the rapid onset and development of the constriction in
parts with normal central innervation is due to a stimulation by the
suprarenal extract of the vasomotor centres; the constriction, however,
might be supported in its further course by the peripheral vasocon-
stricting mechanism which proves to be an efficient though slow
agent in the absence of the central innervation.

SUMMARY.

The removal of the superior cervical ganglion, and section of the
the third cervical nerve and its connecting branches on one side of
a rabbit, deprives the blood-vessels of the ear on the operated side of
all central innervations. The experiments with intravenous injections
of adrenalin into rabbits thus operated, brought out the following
results :

1. The degree of constriction which the blood-vessels attain in the
ear on the operated side is about the same as that of the ear on the
normal side.

2. The constriction, however, sets in later and develops more slowly
on the operated than on the non-operated side.

3. On the normal side, the constriction is usually followed by a
moderate but distinct vasodilatation. Such an after-effect is absent
on the operated side.

4. On the operated side, the constriction lasts considerably longer
than on the non-operated side.

From these observations, the conclusion is drawn that in the normal
animal the injected suprarenal extract stimulates, in the first place,
the vasomotor centres. It stimulates the constrictors, as well as the
vasodilators; but when the extract is present in the blood in a
sufficient dose, it favors constriction, which sets in quite abruptly
and develops rapidly. The further continuation of the constriction is
possibly also supported by the stimulation of the peripheral mechanism.
When, after a few minutes, the dose of the extract within the blood
becomes diminished, the stimulation of vasodilatation is now favored,
and the constriction therefore soon disappears, giving way in some
places even to some degree of vasodilatation. In the absence of
central innervation, the vasoconstriction is accomplished by the pe-
ripheral mechanisms, which react more slowly to stimulation by the
extract, but whose final constricting effect lasts for a considerable
time, since it cannot be interrupted by a central vasodilatation.

’

MUSCULAR CONTRACTION AND THE VENOUS
BLOOD-FLOW.

By R. BURTON-OPITZ.

[From the Physiological Laboratory of Columbia University, College of Physicians and
Surgeons, New York.]|

WO distinct groups of variations in the venous blood-flow are
recognizable. The first group embraces those variations in
the blood-volume which occur periodically, either with the changes in
intra-auricular pressure during each cardiac cycle, or with the changes
in intra-thoracic pressure during each respiratory phase.! The second
group embodies all those variations which are dependent upon acci-
dental, mechanical causes and do not appear at regular intervals. The
latter may therefore be termed “irregular” variations.

Leaving out of consideration those obvious changes in the venous
flow which result from external mechanical influences, the present
paper deals with only the most important class of variations of the
latter type ; namely, with those produced by the contraction of skeletal
muscles.

Briefly outlined, the method consisted in determining quantita-
tively the volume of the blood-flow, first under normal conditions and
subsequently during the different stages of muscular contraction.
The femoral vein was used in these experiments, because this vessel
is easily isolated and drains a complex of muscles, the nerves of which
are readily accessible to the electrodes. The blood-volume was
measured by means of Hiirthle’s recording stromuhr.?

The experiments were performed on medium-sized dogs in mor-
phine-ether narcosis. The nerves of the posterior extremity to be
experimented on were previously placed in covered electrodes. The
sciatic nerve was exposed where it leaves the pelvis, the obturator
nerve as it passes across the median surface of the adductor femoralis
magnus muscle, and the “crural” nerve at some point of its course
along the femoral vessels (saphenous nerve).

1 BuRTON-OPITZ, R: This journal, 1902, vii, pp. 435-459.

? HURTHLE, C.: Compte rendu du cinquiéme congrés international de physio-
logie, Turin, 1901; A short description of this instrument is also given in the

paper cited previously.
161

162 R. Burton- Opitz.

The animal was placed upon its back with the posterior extremities
slightly flexed and abducted. The toes of the leg experimented on
were loosely fastened to a flexible rod. ,

The femoral vein was isolated from the groin down to the entrance
of the vena femoralis posterior superior. In this preparation a very
small vein, draining the fatty tissue below the groin, was destroyed.
When present, another small vein entering the main vessel nearly
opposite the latter was ligated, but those veins found in the immediate
neighborhood of the groin were compressed only during the insertion
of the stromuhr. The stromuhr, filled with a warm normal saline
solution, was placed vertically. Its central cannula came to lie
close to the small veins near the groin, while its peripheral cannula
remained at some distance from the orifice of the vena femoralis
posterior superior.

In those experiments in which the effect of compression of the
femoral artery on the blood-flow in the corresponding vein was tried,
the artery was placed in a ligature opposite the central cannula of the
stromuhr; z.e., about three centimetres below the groin. The artery
was raised and compressed between the forceps.

Compression of the femoral vein, peripherally to the orifice of the
vena femoralis posterior superior, was also employod. The latter
vein drains the largest mass of the gracilis muscle, and thus, by stimu-
lating the obturator nerve, the effect of the contraction of a single
muscle on the venous blood-flow could be determined.

Upon the smoked paper of the kymograph were recorded the
curve of the blood-flow and the time-curve, written by a Jaquet
chronometer in fifths of seconds. The latter record served at the
same time as the abscissa of the former. The duration of the
stimulation of nerves was marked by an electro-magnetic signal. The
respiratory movements were recorded by a tambour communicating
with the left pleural cavity. In most of the experiments the venous
pressure was also recorded. For this purpose Hiirthle’s venous
manometer was connected by means of a T tube with the periph-
eral cannula of the stromuhr. The pressure was therefore recorded
between the muscles and the instrument.

THE NORMAL FLOW IN THE FEMORAL VEIN.

It seems advisable to consider first the normal volume of the blood-
flow and subsequently the changes which result in consequence of

Muscular Contraction and the Venous Blood-Flow. 163

muscular contraction. The average value being obtained, a more
ready comparison can be made between the normal blood-flow and
the flow during a muscular contraction.

The lever of the recording stromuhr of Hiirthle writes a continuous
curve, composed of upward and downward phases. The writing lever
passes in the former direction, if the piston in the central cylinder is
driven downward by the blood entering through the opening in the
roof of the stromuhr. But if by turning the disc below the floor of
the instrument, the blood is forced into the lower part of the cylinder,
the piston travels upward and the writing lever records in this case a
curve from above downward.

In Tables I to IX the value of the blood-stream is calculated for
about one half the total number of phases of each experiment. The
duration of each phase and the total quantity of blood propelled dur-
ing this time having been obtained, these values were reduced to cubic
centimetres per second. The latter figures were then employed in
calculating the average value of the blood-stream for each experiment.

In order to avoid all errors due to coagulation only about twenty
phases were included in the calculation, a number sufficient to obtain
a good average. Those variations in the curve, dependent upon the
cardiac and respiratory activity,! were wholly disregarded; only the
duration and height. of the entire phase were measured.

As those phases during which compression of the femoral artery
was resorted to, would have necessitated a different arrangement of
the tables, they are inserted separately after the experiments to which
they belong. Those phases, during which stimulation of the nerves
of the posterior extremity was tried, are indicated in the tables, but
will be considered separately in a later chapter. The last four experi-
ments of this series show the effect of nerve-section on the blood-flow
in this particular vein. A sufficient number of normal phases having
been recorded, the nerves enumerated above were quickly divided
with the scissors, and another series of phases written.

The other details, it seems to me, can easily be derived from
Tables I to IX.

1 Even at such a great distance from the heart the cardiac variations in the
blood-flow were often very conspicuous, but naturally their amplitude is less here
than in the external jugular vein. The same may be said regarding the respiratory
variations.

Weight of dog, 16} kilos.

Duration of
phase in
seconds.

R. Burton- Opitz.

TABLE

Total vol.
of blood dur-
ing phase.
Gc:

I. EXPERIMENT I.

Volume.
c.c. per
second.

Left femoral vein used.

Remarks.

Now 0 oot ©) Wo. WC. Mo. I SL ao

|
nunt wWwWnde OC

mw we d& WH
SN =

Del
5.0
8.6
7.9
12
8.0
7.0

8.3
Ue)
8.0

8.2
STi

Shi
7.2

7.8
8.2

8.1
7.2,
8.7
8.3
72
8.0
8.5

6.9
9.0

§.1

8.0
8.3
78

1.59
1.44
1.01
1.05
1.00
1.00
1.21

0.83
1.14
1.00

0.98
1.00

0.91
ae

1.06
0.95

Compression of femoral artery.

Compression of femoral artery.

Compression of femoral artery.

Stimulation of sciatic nerve, tetanic
>

current.

Compression of femoral artery.

EXPERIMENT I,

Highest value, 1.59; lowest value, 0.83; average value, 1.08 c.c. per sec.

Compression of the femoral artery.

| Duration of |

compres-
sion in
seconds.

Total quan-

tity of blood

during com-
pression.

c.c. per
second.

Volume.

Average
value.
Cicapen
second.

Dee
12.1
11.6

6.7

Av. value

of normal

blood-flow.
c.c. per
second.

Decreases
during com-
pression.
Per cent.

Muscular Contraction and the Venous Blood-Flow. 165

TABLE II. EXPERIMENT II.

Weight of dog, 13 kilos. Left femoral vein used.

2 Total vol.
Duration of aeiicatdue: Volume.

phase in c.c. per Remarks.

ing phase.
seconds. &P second.

c.C.

6.0 | 0.83
10.8 0.75
WES 0.87
ee 56 Compression of femoral artery.
10.1 0.79
10.8 0.72

9.0 0.66

OoaNA nA fPW NY

5¢ ac Compression of femoral artery.
9.0 0.72
Compression of femoral vein.

Stimulation of obturator nerve,
tetanic current.
Compression of femoral vein.

Stimulation of obturator nerve,
tetanic current.

1S

14.5

13.6

13:9

15.9

Stimulation of sciatic nerve, tetanic
current.

13.6
ZS

Highest value, 0.88; lowest value, 0.66; average value, 0.77 c.c. per sec.

EXPERIMENT II. Compression of femoral artery.

. | Total vol. : Av. value

No.of iene of blood dur-| Volume. eee of normal

phase he cs ing com- c.c. per eanye! blood-flow.
: pression. second. ae? ic per

seconds. sec s
C.c. ond second.

Decrease
during com-
pression.
Per cent.

8.1 1.3 0.16
10.7 2.2 0.20

0.18 0.77

166 R. Burton-Opitz.
TABLE III. EXPERIMENT III.
Weight of dog, 14 kilos. Right femoral vein used.
: Total vol. of
Nipcot Duration blood during Volume.
hack of phase in iece G6. per Remarks.
pnase- | seconds. eae ; second.
1 19.6 11.8 0.60
2 20.5 11.6 0.56
3 56 Dit ar Compression of femoral artery.
4 17.5 Valet 0.63
5 17.0 Weal 0.71
6 16.0 92
7a ae 5 e Compression of femoral vein.
7b we xe 5G Stimulation of obturator nerve,
single induction.
15.0 8.7 0.58
9 14.3 10.5 0.73
10 18.7 ore 0.49
1 ais He or Stimulation of sciatic nerve, sin-
gle induction.
We 16.1 11.0 0.68
13 16.9 11.9 0.70
14 i ae = Stimulation of sciatic nerve, sin-
gle induction.
15 14.0 10.9 0.77
16 15.3 O7 0.63
l7a Compression of femoral vein.
17b Stimulation of obturator nerve,
tetanic current.
18 18.2 10.6 0.58
19 17.2 10.0 0.57
20 fe Be ae Compression of femoral artery.
21 we Be a Stimulation of sciatic nerve, te-
tanic current.

22 | e: | - 3 Compression of femoral artery.
Highest value, 0.77; lowest value, 0.49; average value, 0.63 c.c. per sec.
BS SOR aS |
EXPERIMENT III. Compression of femoral artery.

Duration | Total vol. of Volu Average Av. value | Decrease
No. of | of compres- | blood during é ng value. of normal | dur. com-
phase. sion in compression. sag Cie anelG blood-flow. | pression.
seconds. CG. Saas second. ¢.c. per sec. | Per cents
3 5.2 1.3 0.25
20 5.4 1.4 0.26 0.23 0.63 63
22 5.3 ial 0.20

Muscular Contraction and the Venous Blood-Fiow. 167

TABLE IV. EXPERIMENT IV.
Weight of dog, 12} kilos. Left femoral vein used.

Total vol. of
blood during
phase.
Ch

Volume.
eicy per Remarks.
second.

Duration
of phase in
seconds.

14.0 Dull 0.65
11.8 9.6 0.81
te 38 as Compression of femoral artery.
11.5 9.6 0.83
11.6 8.2 0.70
Compression of femoral vein.

Stimulation of obturator nerve,
single induction.

Stimulation of sciatic nerve, sin-
gle induction.

Stimulation of sciatic nerve, sin-
gle induction.

Stimulation of sciatic nerve, te-
tanic current.

Stimulation of sciatic nerve, te-
tanic current.

| Compression of femoral vein.

Stimulation of obturator nerve,
tetanic current.

A 0.77

Highest value, 0.87; lowest value, 0.64; average value, 0.73 c.c. per sec.

EXPERIMENT IV. Compression of femoral artery.

Volume Aserace Av. value | Decrease
: 8 of normal | dur. com-

oe tier Al blood-flow. | pression.
: ag ame .cpermSeem|bercents

Duration | Total vol. of
No. of | of compres- | blood during
phase. sion in compression.

seconds. c.Cc.

3.2 : 0.73 71

168 LR. Burton- Opitz.

TABLE V. EXPERIMENT V-.

Weight of dog, 17 kilos. Right femoral vein used.

Pacneon Total volume
No. of ‘ of blood Volume.

: of phase in cp 2
phase. secre during phase. | C.c. per sec.
: Ge:

Remarks.

Tell

Stimulation of sciatic nerve,
single induction.

Stimulation of sciatic nerve,
| single induction.

Stimulation of sciatic nerve,
single induction.

Highest value, 1.38; lowest value, 1.03; average value, 1.18 c.c. per sec.

Muscular Contraction and the Venous Blood-Flow. 169

TABLE VI. EXPERIMENT VI.
Weight of dog, 14} kilos. Left femoral vein used.

RGration | Total volume
| of blood Volume.

of phase in | during phase c.c. per sec
seconds. a ae see ERI SES:

Remarks.

5.6
8.5
6.9

Stimulation of sciatic nerve,
tetanic current.
IWASEL

10.0 1.09

Normal flow: value, 1.29; lowest value, 0.79; average value, 1.06 c.c. per sec.

16-17 50 ne ais Sciatic, obturator, and crural
nerves cut.

18
19
20

2)

12.0
11.6
EA:
11.6

Flow after nerve-section: highest value, 4.0; lowest value, 2.4;
average value, 3.1 C.c. per Sec.

170

12

Normal flow:

R. Burton- Opitz.

TABLE VIE

Weight of dog, 12 kilos.

Duration
of phase in
seconds.

20.9

6.4
6.4
ab

Total volume
of blood
during phase.

C.C.

10.8

highest value, 0.68; low

|

EXPERIMENT VII.

Left femoral vein used.

0.52
0.49

0.58
0.68
0.56
0.46

0.55
0.62

0.51

Volume.
c.c. per sec.

est value, 0.46;

Remarks.

Stimulation of sciatic and ob-
turator nerves, tetanic.

Stimulation of obturator
nerve, tetanic.

Stimulation of sciatic and

obturator nerves, tetanic.
average value, 0.55 C.c. per sec.
|

Sciatic and obturator nerves
cut.

Flow after nerve-section: highest value, 1.76; lowest value, 1.26;
average value, 1.46 c.c. per sec.

Muscular Contraction and the Venous Blood-Flow. 171

TABLE VIII. EXPERIMENT VIII.

Weight of dog, 18 kilos. Left femoral vein used.

Duration Total volume
eo of blood Volume.

of phase in ;
phase. Aga during phase. | c.c. per sec.
>.
CC:

Remarks.

Stimulation of nerves, sin-
gle induction.

st ss 1.01

8.3 11.4 37)

Normal flow: highest value, 1.56; lowest value, 1.01; average value, 1.20 c.c. per sec.

16-18 ae ae | Dr | Sciatic, obturator, and crural
nerves cut.

1k,

Flow. after nerve-section: highest value, 3.73; lowest value, 2.92;
average value, 3.20 c.c. per sec.

172 R. Burton-Opitz.

TABLE IX. EXPERIMENT IX.

Weight of dog, 11 kilos. Right femoral vein used.

uate Total volume
No. of | of blood Volume.

| of phase in .
phase. es during phase. | c.c. per sec.
: c.c. ‘

Remarks.

o%

Stim. of sciaticand obturator
nerves, tetanic current.

0 10.8 0.49
4 12.0 0.47

|
Normal flow: highest value, 0.60; lowest value, 0.46; average value, 0.50 c.c. per sec.

8-9 os Ss ee Sciatic and obturator nerves
cut,
10

|
|
10.8 | 1.47

Flow after nerve section: highest value, 1.74; lowest value, 1.35;
average value, 1.58 c.c. per sec.

The most important facts derived from Tables I to IX are more
comprehensively arranged in Table X. This outline, however, also
includes the velocity of the blood-stream for five of the experiments,
this value being given in millimetres per second.

In obtaining the internal diameter of the vein, necessary in this
calculation, I have again employed the method which Tschuewsky !

' TscHuEwsky, F. A.: O Kposocnaésaenin Orgbabnxb Opranoss (On the blood-
supply of several organs), Charkow, Igo2.

Muscular Contraction and the Venous Blood-Flow. 173

made use of in his investigation on the blood-flow in different arteries.
Although far from being exact, it was the most suitable for these
experiments. The outside diameter of the vein having been deter-
mined by means of calipers, the vein was lightly compressed between
two narrow plates of glass. The thickness of the glass-plates being
deducted from this measurement gives the thickness of the vessel wall,
which in turn is deducted from the outside diameter of the vein.

TABLE X.

THE FLOW OF THE BLOOD IN THE FEMORAL VEIN.

Weight
of dog.
Kilos.

Experiment.
Normal volume
of blood
Volume during
compression of
| femoral artery.
Per cent of
decrease during
compression.
Volume after
section of nerves.
C.c. per sec.
Increase C.
times A.
Internal
diameter of vein
in mm.
Velocity of
blood stream.
mm. per sec.

io)
oy
oO

Ay.
values. |

Among the conclusions derived from the preceding table the fol-
lowing may be emphasized. We observe first that in spite of the
large calibre of the vein, the blood-volume is rather small. It varies
from 0.50 c.c. per second in a dog, weighing 11 kilos, to 1.20 c.c. per
second in a dog, weighing 18 kilos. The average value of the blood-
flow, as obtained in the above nine experiments, is 0.85 c.c. per
second ; the average weight, 14.2 kilos.

Although the correspondence between the weight of the animal
and the blood-volume is not brought out very strikingly in these ex-

174 R. Burton-Opitz.

periments, the table at least strongly suggests that in general the
volume of the blood-flow increases and decreases in proportion to the
weight of the animal.

If a comparison is made between the volume of the blood-stream
in the femoral and that in the external jugular vein, it is found that
the former is considerably smaller. The eight experiments! which I
made to determine the normal quantity of blood in the right external
jugular vein, gave the average value of 2.03 c.c. per second. The
femoral vein, therefore, carries less than one half this amount of
blood.

It cannot be assumed that this difference in the blood-volume of
the veins under consideration is due to corresponding differences in
body-weight, because the average weight of the animals used in deter-
mining the blood-flow in the external jugular vein was 12 kilos, while
that of the dogs used in the present experiments is 14.2 kilos. Al-
though smaller than the external jugular, the lumen of the femoral
vein is of considerable size, at least its large internal diameter is not
proportionate to the small quantity of blood traversing this vessel.

In the pamphlet referred to previously Tschuewsky gives a series
of experiments on the blood-flow in the femoral artery. The average
value of the blood-volume in seven dogs, ranging in weight from 12.5
to 14.5 kilos, is 0.63 c.c. per second. Two other dogs, weighing 37.0
and 51.0 kilos respectively, showed a flow of 1.2 c.c. and 1.0 c.c. per
second. If the average value for the venous flow (0.85 c.c. per
second) is compared with that of the arterial (0.65 c.c. per second),
it is found to be slightly greater. The difference, amounting to only
0.2 c.c. per second, could easily be explained by referring to the dif-
ferences in the weight of the dogs; those used in the present experi-
ments being the heavier. However, it is also noticed that in the vein
the maximal value of 1.0-1.2 c.c. per second was obtained already in
dogs weighing only 16 to 18 kilos. These facts suggest that even
normally the volume of blood traversing the vein is slightly greater
than that of the corresponding artery. It is, however, not absolutely
correct to draw these conclusions from two different sets of experi-
ments, or animals.

In support of the previous statement the following fact might be
cited: If the femoral artery is compressed opposite the stromuhr, the
blood-flow in the vein does not cease completely. The quantity of

‘ BurTON-Opitz: This journal, 1902, vii, Experiments 1 to 5, on page 439;
Experiment 2, on page 441; and Experiments 1 and 2, on page 442.

Muscular Contraction and the Venous Blood-Flow. 175

blood still propelled must therefore reach this vein in an indirect way
by means of anastomosing vessels. It is possible that even normally
a slight quantity of blood reaches the vein in this manner.

The reduction in the volume of the blood-stream, produced by the
procedure just mentioned, is not uniform. In the experiments given
above the decrease varied from 63 to 90 per cent. In another experi-
ment not inserted above, because the record was accidentally de-
stroyed, the decrease amounted to only 51 per cent. Thus, the
compression caused in this case a reduction in the blood-flow of only
about one half its former volume.

From what has been said regarding the normal blood-flow and the
diameter of the femoral vein, it can readily be concluded that the
velocity of the current is very slight. The five experiments inserted
above have given an average value of 61.6 mm. per second. If the
velocity of the venous current is compared with that in the corre-
sponding artery (134.4 mm. per second)! it is found to be about one
half as great.

After section of the nerves innervating the posterior extremity, the
phases written by the lever of the stromuhr immediately became
much steeper, indicating thereby that a greater quantity of blood
traversed the vein. In the Experiments VI to IX in which this pro-
cedure was tried the resulting flow was from 2.6 to 3.1 times greater -
than the normal (average 2.8 times greater).

A comparison between the quantitative determinations of the venous
and arterial blood-flow shows that the insertion of the stromuhr into
a vein is not such a serious procedure as might be supposed at first.
At least, if any impediment to the venous circulation is produced
thereby, it is not greater than that which results if this instrument is
placed in an artery.

THE CHANGES IN THE NORMAL FLOW PRODUCED BY MUSCULAR
CONTRACTIONS.

Tetanic current. — In the preceding paragraphs we have determined
the volume of the blood-flow when the muscles of the posterior ex-
tremity are at rest. The present chapter contains a consideration of
those changes in the blood-flow which ensue when muscular contrac-
tions occur.

1 See TSCHUEWSKY’s paper, page 76.

176 R. Burton- Opitz.

The sciatic and obturator nerves were stimulated while the curve
of the blood-flow was being written. The stimulation was confined
to one, or included both the nerves enumerated. Both tetanic and
single induced currents were used, their strength being varied so as
to produce either a strong, or a medium muscular contraction. The
phases during which stimulation of the nerves was resorted to are
indicated in the tables inserted previously.

Whether only one or both nerves were stimulated, the general
characteristics of the variation in the blood-flow remained always the
same for each kind of stimulus. Changes in the intensity of the cur-

NEG. PRESS L.PLEUD. CAVITY IRA pnnnamnu

TIME % SEC.

FLOW OF BLOOD
FEMORAL VEIN

mre rnpeoeeniergi erin eaierent —s

FicurE 1.— Two-thirds the original size. Variation in the venous blood-flow during
a tetanic muscular contraction.

rent also produced no alterations in the general outline of the curve.
The only difference noticed under all conditions was merely one of
one degree ; z. ¢., the more nerves stimulated and the stronger the
current, the more evident was the variation in the blood-flow.

It seems advisable to consider first the details of a variation pro-
duced by a tetanic muscular contraction. For this purpose Fig. 1 is
inserted in this place. A tetanizing current of medium strength (dis-
tance of coils, 13 cm.) was in this case applied to the sciatic nerve;
duration of stimulation, about six seconds.

We observe immediately that the curve of the blood-flow (from a’

Muscular Contraction and the Venous Blood-Flow. 177

to a) becomes very steep at a, suggesting thereby that a great in-
crease in the volume of the blood-stream has taken place. When the
ordinates are compared, this point is found to correspond with the
moment of stimulation. At 4 the opposite effect is noticeable.
The decrease in the flow beginning here continues to c, to the mo-
ment of breaking the current. Furthermore, it is evident that the
flow becomes greater than normal immediately on discontinuing the
stimulation; the curve showing for a brief period a greater incline
than the normal (¢c to d). Eventually, however, the blood-flow
returns to its normal value.

The period of great onward movement, occurring after the applica-
tion of the current (a to 4), is therefore synchronous with the period
of muscular shortening. It is noticed, moreover, that the quantity of
blood forced into the vein is greater during the first part of this period
than during its latter half, when the muscles have nearly reached
their maximal degree of contraction. Apparently, this increase in
the blood-volume is due solely to the pressure of the contracting
muscular substance upon the mass of blood contained in its vessels.

As soon as the point of maximal shortening has been reached (0)
the curve inclines strongly toward the abscissa, indicating thereby
that the blood-flow is less than normal (6 toc). The decrease in the
flow is most conspicuous during the first part of this period, while,
if the stimulation is continued for a longer time, a slight and gradual
increase above the previous value is noticeable during its latter half.
This slight rise becomes the more evident, the longer the muscles are
kept in the contracted state. The tetanic muscle therefore is an
obstacle to the blood-flow, but when in the course of a longer stimu-
lation the muscle becomes relaxed by fatigue, a steadily increasing
quantity of blood is enabled to pass.

The complete relaxation of the muscles on breaking the current is
followed by a brief period of increased venous flow (c to @), after which
the normal value is again slowly established. Considered quantita-
tively, this gush-like rise, after removing the hindrance to the blood-
flow, is always much smaller than that occurring during the shortening
of the muscles (a to 6).

The entire variation in the venous blood-flow produced by a tetanic
muscular contraction may therefore be divided into the following
phases:

1. Period of great flow, synchronous with the muscular shortening

B(@.fo0).

178 R. Burton-Opitz.

2. Period of slight flow, continuing during the contracted state of
the muscle (4 toc).

3. Short period of increased flow, following the relaxation of the
muscle (¢ to @).

The question how the venous blood-flow is altered by a tetanic
muscular contraction has previously been investigated by Sadler! and
subsequently by Gaskell. The latter author enlarged upon the work
of the former by bringing the changes in the blood-volume into rela-
tion with the alterations in the form of the muscle. He isolated the
vein, draining the largest mass of the musculus vasti and musculus
cruralis and measured by a special device the quantity of blood escap-
ing from this vessel, before and during the contraction of the muscles
mentioned. As far as the general characteristics of the changes
observed by Gaskell are concerned, the results obtained by means of
the stromuhr completely substantiate those found by the method just
cited.

To show that the retardation of the blood-stream following the
maximal muscular shortening (0 to c) is not due to the compression
of the larger vessels between the entire complex of muscles, but is
caused by mechanical obstacles within each muscle, the following
experiment was repeatedly tried. Both the arterial and venous trunks
were isolated from all surrounding tissues from the groin downward
to the musculus gracilis. The femoral vein was then compressed
peripherally to the orifice of the vena musculus gracilis, so that the
stromuhr recorded only the flow through this vessel. The musculus
gracilis was tetanized by stimulation of the obturator nerve.

The variations in the blood-flow accompanying the tetanization of
the gracilis muscle showed under these conditions the same outline
as those obtained when the total quantity of blood was measured
during the contraction of the entire posterior extremity. However,
as the volume of blood was in this case much smaller, it naturally
follows that the amplitude of the variations was much less.

In localizing the obstruction to the blood-flow within the muscle
the observations of Heilemann? prove very suggestive. This author
studied the circulation in the musculus submaxillaris of the frog under

‘ SADLER, W.: Arbeiten aus dem physiologischen Institute zu Leipzig, 1869,
pp- 77-100.

* GASKELL, W. H.: Arbeiten aus dem physiologischen Institute zu Leipzig,
1877, pp. 45-88.

* HEILEMANN, H.: Archiv fiir Anatomie und Physiologie, 1902, pp. 45-53.

Muscular Contraction and the Venous Blood-Flow. 179

the microscope. On tetanizing, or on stimulating this muscle by
single induction shocks, he found that the flow was greatly retarded
in those fine anastomosing venules which pass between the muscular
fibres and parallel to them. Even a complete cessation of flow was
observed at times, and if a strong tetanic current was used, an oscil-
lating motion and even a backward movement of the column of blood
resulted.

Differences in the force of the muscular contraction do not alter
the general character of the curve. This statement is illustrated by

N. PR.L.PLEUR. CAVITY

DNA) AANAAIAII

TIME % SEC.

FLOW OF BLOOD
FEMORAL VEIN

E.M. SIGNAL Bi ein ied Cc Ai ae c
——— = Te

Ficure 2.— Two-thirds the original size. Two successive variations showing different
amplitude (tetanic muscular contraction).

Fig. 2. In this case two brief tetanic contractions of different
strength were produced in quick succession. The sciatic nerve was
stimulated first with a current of medium strength (distance of coils,
15 cm.) and subsequently with a strong current (distance of coils,
7 cm.)

We observe immediately that, although the details of the curve are
the same in both cases, the changes in the blood-flow are more deci-
sive during the strong muscular contraction. The rise of the curve
on making the current (at a) is more sudden, steeper, and higher in
the latter instance; which indicates that the volume of blood pro-

180 R. Burton-Opits.

pelled during the shortening of the strongly contractec muscle is
proportionately greater.

The maximal shortening having been reached (at 0), tue curve in-
clines more strongly toward the abscissa in the latter case; the blood-
flow is therefore more effectively retarded by the strongly contracted
muscle (4 to c). Whether a cessation of flow can be produced by a
proportionate increase in the strength of the stimulation could not be
determined with accuracy, because under these conditions the mus-
cular movements were so forcible that the stromuhr was shifted out

ABER Sa:

THE CHANGES IN THE BLOOD-FLOW DURING A TETANIC MUSCULAR CONTRACTION.

Taken
from |
exp.

Nerves
stimulated.

Flow during

Normal
blood-flow.

Strength of
current.
Dist. of coils
cm.
Duration of
stimulation in
seconds.
c.c. per sec.
period of muse.
shortening.
c.c. per-sec.
Flow during
tetanic state
of muscle.
C.c. per sec.
Flow after
musc. relax-
c.c. per sec.

—
eal
Oo
eal
bo
I

sciatic

sr

=
ON

—_
i=)

{ sciatic
) obturator

“cc

of its normal position. The strongest stimulus applied was: distance
of coils, 6 cm.; two dry cells.

The after-effect of the tetanic contraction, consisting in the brief
rise above the normal value of the blood-flow (c to @), was generally
more conspicuous after the application of a strong current. In the
curve now under consideration this period of increased flow is well
marked in both instances, but a decided quantitative difference is not
evident.

Muscular Contraction and the Venous Blood-Flow. 181

To show in a general way the pronounced differences in the blood-
volume during the three principal phases of a tetanic muscular con-
traction, a ndmber of these variations have been calculated, as closely
as this is possible, in terms of cubic centimetres per second. The
periods measured are marked in the preceding figures by the letters
wero, 0 to co and. cto d.

Table XI also contains four instances in which the weak stimulation
was followed by a stronger stimulation. The greater prominence of
the changes resulting under these conditions (see Fig. 2) is clearly
betrayed by these quantitative determinations.

The pressure-changes in the femoral vein, occurring during the
tetanic contraction of the muscles of the posterior extremity, were re-
corded in most of the above experiments by means of a Hiirthle’s
membrane-manometer (venous). This instrument, as stated before,
recorded the pressure peripherally to the stromuhr.

On tetanizing the muscles the pressure quickly rose some millimetres
above the normal value. At about point J of Figs. 1 and 2 the pres-
sure decreased almost as rapidly as it had risen and kept subsequently
very close to zero, falling generally even below the abscissa during
the strong tetanic contraction. After the relaxation of the muscles
the pressure remained slightly above normal for some time. If the
tetanization was continued for a longer time, the pressure began to
rise even before the break of the current.

The pressure-changes were also recorded in several separate exper-
iments -by means of a soda-manometer (sodium carbonate solution,
specific gravity, 1.080) connected with the femoral vein by means of
aT tube. A float was not used, the values being obtained by read-
ing. The sciatic nerve was stimulated.

As is shown in Table XII, the pressure rapidly rose two to five
mm. Hg above normal during the muscular shortening (a to 3).
The end of this period having been reached (at 0), the pressure
again decreased almost as quickly, assuming a value slightly below
normal during the first part of the tetanization. Subsequently it
gradually rose, remaining slightly elevated for a short time after the
relaxation.

The only difference in the results of the former and latter methods
consists therefore in the height of the pressure during the second
period, the period of decreased flow (4 to c). The soda-manometer
indicated that the pressure does not drop to zero, but remains a few
millimetres above the abscissa, which, it seems to me, is the correct
result.

182 R. Burton-Opitz.

Table XII, shows, moreover, that a stronger tetanization causes a
greater rise in pressure during the period-of great flow, muscular
shortening (a to 6). We have seen previously that under these
conditions the blood-volume during the second period (0 to c) is even
more highly reduced; however, a correspondingly greater fall in pres-
sure during this period could not be ascertained definitely, because
the differences were so very slight.

TRACBIEE well
THE CHANGES IN PRESSURE IN THE FEMORAL VEIN DURING A TETANIC
MUSCULAR CONTRACTION.
(The corresponding values in mm. Hg are placed in brackets.)

|
Pressure at Pressure
Normal end of period during
pressure, of muscular | contr. state

Pressure at
relaxation of
muscle.

bonate. Mm. sod. car-|. Mm. sod. wis sod. car-
onate.

coils.
(Mm. Hg.) bonate. carbonate.
cm. (Mm. Hg.) |(Mm. Hg) | (“!™ He)

Strength |
of stim- |
ulus,

Distror Mm. sod. car-| shortening. of muscle.

Duration of
contr. in sec.

Experiment.
Wt. of dog.
Kilos

78[6.2] 100[7.9] 70[5 5] 78+[6 2+]
98[7.7] 140f11.1] 88[7.0] 110[8.7]
65[5.1] 112[8.9] 55[4.3] 78[6.2]
132[10.5] 60[4.7] 80[6.3]
145[11.5] 60[4.7] 80[6.3]
85[6.7] 58[4.6] 62+[4.9+]

Single induced current. — If single induction shocks are used, the
variations in the blood-flow accompanying the muscular twitches,
show a somewhat different outline.

Fig. 3 may serve to illustrate the general character of the varia-
tions obtained under these conditions. The sciatic nerve was stimu-
lated in this instance with a current of medium strength, distance of
coils, 10 cm, }

Both the make (@) and the break (c) of the current are followed by
a considerable increase in the blood-volume which continues during
the entire periods of rising energy of the muscles (a to 6 and ¢ to @).
Furthermore, this rise is directly proportionate to the strength of
the stimulus, and therefore also to the force of the muscular contrac-
tion. The stronger the twitch, the more conspicuous is the increase
in the venous blood-flow.

Muscular Contraction and the Venous Blood-Flow. 183

During the second period, 7. ¢., while the current is passing, the
blood-volume is not reduced as during the tetanization of the muscles,
but remains normal. The blood-flow resumes its normal value imme-
diately after the rise occurring during the make-contraction (a to 4),
and retains it until the break-twitch causes another period of increased
flow (c to d). Subsequent to the latter rise the blood-flow imme-
diately returns to normal.

The fact that the blood-volume is not reduced while the current
passes, greatly influences the total quantity of blood propelled in a

pees es il Sie
SIGNAL A C

FLOW OF BLOOD
FEMORAL VEIN

TIME Ys SEC. as
ae ie ca mbanet Nemaees uaa was

eee Cc

A C
V. PRESSURE i}
poe = J Ww

FiGuRE 3.— Two-thirds the original size. Variation in venous blood-flow accompanying
single muscular twitches (a, make, c, break of current).

given time. To illustrate: supposing the stromuhr has written three
successive phases of the same length. In the first phase only the
normal blood-flow is recorded. During the second phase a strong te-
tanic contraction has been produced, while during the third a single
induced current of equal strength and duration has been applied. In

184 R. Burton- Opitz.

the second case it will be found that the total quantity of blood pro
pelled during the phase is considerably less than the normal volume.
The decrease will be the greater, the longer the duration of the tetani-
zation and the stronger the current. Even if the blood-flow is greatly
increased by the muscular shortening, the reduction in the blood-
volume during the contracted state of the muscle far outweighs the
former effect. In the third case, on the other hand, the total quan-
tity will be greater than normal, the increase being proportionate to
the volume of blood propelled during the rising periods of the mus-
cular twitches.

The greatest increase in the venous blood-flow is therefore produced
by a series of single muscular twitches. The same result can be ob-
tained by successive tetanic contractions, but their duration must be
very brief, so that only the quantitative effect of the period of muscular
shortening can become evident. If continued for a long time, the
retarcation of the flow during the tetanic state of the muscle will
naturally cause a decided decrease in the total quantity of blood
propelled.

SUMMARY.

1. The value of the blood-flow measured in nine dogs varied from
0.50 c.c. per second to 1.20 c.c. per second. The average flow was
0.85 c.c. per second; average weight of dog, 14.2 kilos.

2. The velocity of the blood-stream varied from 48.5 mm. per sec-
ond to 74.7 mm. per second; average velocity, 61.6 mm. per second.

3. Compression of the femoral artery caused a reduction in the
blood-volume of from 63 to 90 per cent; average decrease, 75 per
cent.

4. Section of the nerves innervating the posterior extremity was
followed by an increase in the blood-volume of from 2.6 to 3.1 times
the normal value; average increase, 2.8 times.

5. The variations in the venous blood-flow accompanying a tetanic
muscular contraction may be divided into three periods: —

1. Period of great flow, synchronous with the muscular shortening.

2. Period of slight flow, continuing during the contracted state of
the muscle.

3. Short period of increased flow, following the relaxation of the
muscle.

Differences in the force of the muscular contraction do not alter the

Muscular Contraction and the Venous Blood-Flow. 185

general character of the variation, but only cause changes in the
amplitude of its various details.

6. If single induction shocks are used, a great increase in the flow
results during the periods of rising energy of the muscle. Between
the twitches the blood-flow resumes its normal value.

7. The venous pressure changes during a tetanic contraction in a
corresponding manner. There is a quick rise during the muscular
shortening and an almost equally rapid fall after the maximal con-
traction has been reached. During the first part of the tetanization,
the pressure remains slightly below normal, while during its latter
part the pressure gradually rises and continues slightly above normal
for a short time after the relaxation of the muscle.

is © al Ne
Fi i y :
“we , 7 ' +) ai

Ni sacaaubhe allg athe eee
ot Hitow , 1 i.

At

‘

tia") US Apt, oo ine

if 4 Ab) Te
' a ;
“ ta y’
’ q Pr or
*
Mer t tS

THE INFLUENCE OF FORMALDEHYDE ON THE
ACTION OF CERTAIN LAKING AGENTS AND
ON COAGULATION -OF -BEOOD:

By CHARLES CLAUDE GUTHRIE.

[From the Physiological Laboratories of the University of Chicago and Western Reserve
University. |

ie has been shown by Dr. G. N. Stewart, ! that where formaldehyde

is added to blood in sufficient amount to cause ultimately complete
fixation of the corpuscles, laking can be brought about by various
hemolytic agents, including saponin, for a certain period after the
addition of formaldehyde. The length of this period and the com-
pleteness of the laking depend, of course, on the amount of formal-
dehyde added, as well as on the strength of the laking agent. At
his suggestion I undertook a series of experiments to determine for
how long a period blood-corpuscles still remain susceptible to the
action of haemolytic agents after the addition of smaller amounts of
formaldehyde than were employed by him, and in particular to de-
termine whether such quantities of formaldehyde as were just neces-
sary to prevent putrefaction of the blood would exert a restraining
influence on the action of laking agents. If this question should be
answered in the affirmative, it seemed not impossible that a feasible
method of preserving blood for hzmolytic tests, at least for several
days, might be arrived at, ¢.g. by mixing blood with formaldehyde
and potassium oxalate, the former being for the purpose of restraining
putrefaction, and the latter, coagulation, —a point of some importance,
especially in connection with the investigation of biological hzemoly-
sins. The fact that formaldehyde fixation of hamoglobin does not
completely prevent the corpuscles from being acted on by some
hemolytic agents (¢. g. saponin), since their permeability for electro-
lytes is still increased by them, indicates that certain constituents of
corpuscles which are related to the laking process are not fixed by
formaldehyde. ‘The investigation of the relation between formalde-
hyde fixation and laking seems, therefore, calculated to throw light on

1 STEWART; Journal of physiology, 1901, xxvi, p. 470.
187

188 Charles Claude Guthrie.

the process of hemolysis. I had already been making some investi-
gations on the effects of intravenous injection of formaldehyde, in
the course of which certain phenomena were observed which, on
being reported to Dr. Stewart, led him to suggest that I should
combine with the experiments on the influence of formaldehyde on
laking agents, a series of experiments on the action of that substance
on coagulation of the blood. This accounts for the manner in which
the results are reported in this note. I hope to have an opportunity
to return to this subject.

The method of procedure was as follows: Test-tubes were graduated in
ro c.c. by means of narrow strips of gummed paper, labelled and arranged
in series inracks. ‘The reagent to be mixed with the blood was added by
means of a pipette, graduated in 0.01 c.c., immediately before adding the
blood. The animal which was to furnish the blood was anesthetized with
A.C.E. mixture, and a vaselined glass cannula, connected with a piece
of vaselined rubber tubing 1o cm. long, was tied in one of the common
carotid arteries. The blood was then allowed to flow into the test-tubes
up to the graduate mark or as near it as possible, the flow being regu-
lated by pressure on the rubber tube with the fingers. The tube was
then closed with the thumb, and vigorously shaken to thoroughly mix the
contents. ‘The time required to fill a tube and mix the contents was very
short, the average of a large number being eleven seconds. All tubes
were equally shaken, so that the contents of all might be equally agitated
or mixed, as the case might be. The formaldehyde solutions were freshly
prepared each time, excepting in the cases noted where solutions several
weeks old were used for the purpose of determining their activity.
Schering’s formalin, which has been shown by Dr. Torald Sollmann? to
contain 39.9 per cent formaldehyde, was added to o.g per cent NaCl in
sufficient quantity to make the solutions of the desired strengths. The
potassium oxalate solution was prepared by adding the salt to 0.9 per cent
NaCl solution in the proper proportion to make 1 per cent solutions.

The following are the results obtained:

1. Coagulation is retarded in blood drawn into solutions of for-
maldehyde in proportion to the amount of formaldehyde present in
the resulting mixture. (Table II.)

2. Formaldehyde in sufficient amount prevents coagulation indefi-
nitely, but the amount necessary to do this varies with different
animals of the same species. In eight experiments, seven dogs and
one rabbit being used, the amount necessary to prevent coagulation

' SOLLMANN: This journal, 1902, vii, footnote, p. 225.

Lnfluence of Formaldehyde on Coagulation of Blood. 189

ranged from one part of formaldehyde to 66.6 parts of blood to 1 to
400, and gave an average of 1 to 185.5. That varying amounts are
necessary to prevent coagulation in different animals may of course
be due in part to variations in the number of corpuscles per cubic
centimetre of blood.

In making such observations it is necessary to take into account
certain possible sources of error. The more common ones are men-
tioned below:

If to tubes containing equal amounts of formaldehyde, blood be added at
intervals, it will be seen that the more blood-clot the blood comes in
contact with before flowing into the tube, the shorter the coagulation time
will be. This is shown by Table II, Tubes 4, 17, 26, 35, and 47, as well as
by a number of other similar experiments.

By reference to Tubes 7, 8, 19, 28, 37, 49, and 57, it will be observed that all
but one of them (37) showed coagulation in greater or less degree. The
coagulation time gradually became longer until this tube was reached,
when it began to shorten and continued to become shorter to the end of
the series, Tube 57 coagulating in a shorter space of time than the first (7).
The reason of this variation in the amount necessary to prevent coagula-
tion in different samples of blood from the same animal, is not quite
clear. It may be due either to the formaldehyde retarding or preventing
coagulation by restraining the formation of fibrin ferment, or by a direct
action on already formed fibrin factors. The former explanation seems
the more likely, as blood drawn through a tube containing fresh blood clot
must contain more fibrin ferment than blood drawn through a clean tube.
My experiments on this point are still incomplete.

The coagulation time of normal blood is very materially shortened, as is well
known, by drawing it through a tube contaminated with fresh blood clot.
This is shown by Tubes 1, 10, 20, 30, 40, 50, 58, and 59. In this
experiment coagulation was not hastened merely by loss of blood, to any
appreciable extent, as shown by Tube 60, which only differed from the
tubes enumerated above, in being drawn from the opposite common
carotid and through a perfectly clean tube. It will be seen by comparing
this tube with 58 and 59, that it required almost four times as long to
clot, and slightly longer than 1. It has been shown by Arthus? that
coagulation is accelerated by bleeding. It appears, however, that this is
not so powerful a factor as contact with clotted blood.

Asphyxia is another important factor in retarding coagulation, which is worth
while bearing in mind, though if the blood be drawn through a cannula
containing clot, the effect is to a great extent lost. ‘This is shown in Tube

1 ArTHUS: Journal de physiologie et de pathologie générale, 1902, iv, p. 283.

190 Charles Claude Guthrie.

61. In one of my experiments, blood drawn by puncturing the right
ventricle and allowing the blood to spurt into a test-tube, ten to fifteen
minutes after the arterial pressure had fallen to zero (of course, plus residual
pressure), from asphyxia, coagulation was delayed for some hours, the
corpuscles having settled before it occurred.

That the dilution of blood is not sufficient to account for the delays noted on
addition of solutions of formaldehyde, is shown by comparing the series
29, 31, 32, and 33 with either of the series occurring just prior or sub-
sequent to it. The only difference in this and the two series mentioned
is that they contain formaldehyde in the amounts stated in the table,
while it consists of mixtures of o.g per cent NaCl and blood only. In the
NaCl tubes there is only a slight delay which is accounted for by the
dilution.

3. Amounts of formaldehyde too small to prevent coagulation,
prevent the clot from advancing beyond a delicate jelly-like stage, for
a period proportionate to the amount of formaldehyde present. With
amounts below 1 to 1400, contraction very gradually occurs, though
1 to 50,000 markedly delays contraction and prolongs the jelly stage.
On washing normal clots and jelly clots, and comparing the result-
ing insoluble substances, the normal clot shows for the most part
dense masses of fibrin, the fibres of which are easily seen under the
microscope, while the formaldehyde clots show brownish granular
masses or less gelatinous residues, having no gross or microscopic
characters of true fibrin. Treated with Loffler’s methylene blue and
examined microscopically, they are seen to be made up of irregular
granular masses and leucocytes, and red corpuscles more or less laked.

4. The onset of spontaneous laking! at room temperature is not
markedly delayed by small quantities of formaldehyde, but, once
begun, it proceeds much more slowly than in normal blood. This is
shown by comparing Tubes 1, 10, 20, 30, 40, 50, 58, 59, 60, and 61, with
the tubes containing formaldehyde. Larger quantities up to a certain
point greatly hasten laking. The amount necessary to fix the cor-
puscles with practically no laking varies with different specimens of
blood, owing probably to the variations in the number of corpuscles
per'c.c. of blood. As a rule, though not always, the amount of for-
maldehyde necessary to prevent coagulation also largely prevents
laking, the corpuscles settling to the bottom of the vessel and gradu-
ally assuming a chocolate-brown color. Corpuscles treated with for-

‘ It has been shown, e. g. by Dr. G. N. STEWART (Journal of physiology, 1899,
xxiv, p. 228), that sterile blood lakes when it is allowed to stand.

Influence of Formaldehyde on Coagulation of Blood. 191

maldehyde in proportions up to 1 part of blood to 4 parts of I per
cent formaldehyde for as much as forty-eight hours, and then washed
free from plasma with 0.9 per cent NaCl solution, are slowly but
markedly laked by water and solutions of sapotoxin.

6. The influence of formaldehyde on laking by foreign serum was
determined in the following manner. (Table I.)

Approximately two thirds of the total amount of blood was drawn from a dog,
and allowed to clot. 1 per cent solution of formaldehyde in o.g per cent
NaCl solution was then injected into a vein in about the proportion of 1
to 1500 of the remaining blood. Five minutes after the completion of
the injection, the remainder of the blood was drawn and allowed to clot.
Rabbit’s blood was then drawn and defibrinated. All the specimens were
then placed on ice and allowed to remain for sixteen hours, when they
were removed and the action of formaldehyde studied.

Rabbit’s blood and dog’s serum were treated separately with various pro-
portions of formaldehyde, for ten minutes, and then mixed with equal
amounts of blood or serum, as the case might be, formaldehyde rabbit’s
blood being mixed with untreated dog’s serum and vice versa. The
mixture of blood and serum was allowed to stand ten minutes, and was
then centrifugalized five minutes, and the supernatant serum examined.

(a) Rabbit’s blood plus formaldehyde, in the proportion of 1 to
1000, is moderately laked by dog’s serum in the time specified.
Larger amounts prevent laking. Blood treated with amounts too
small to prevent laking, are laked in proportion to the amount present.

(6) Dog’s serum plus formaldehyde in proportion of 1 to 2000
or less, still retains its power of laking rabbit’s corpuscles, the amount
of laking depending on the proportion of formaldehyde added. For-
maldehyde, in the proportion of 1 to 1000 or above, restrains all
laking for the time specified. It will be seen that a smaller amount
of formaldehyde will restrain laking when it is added to dog’s serum
previous to the mixture of the serum and the rabbit’s blood, than
when it is first added to the rabbit’s blood.

(c) Serum from blood drawn from the dog after injection of
formaldehyde, in the proportion of about 1 to 1500, lakes rabbit’s
corpuscles as energetically as serum from blood drawn before the
injection.

(d@) Mixture of formaldehyde bloods and sera, containing more than
enough formaldehyde to prevent biological laking, are rapidly and
strongly laked by water or sapotoxin solution in 0.9 per cent NaCl
solution.

192 Charles Claude Guthrie.

7. Spectroscopically no obvious change is discoverable in the
heemoglobin when formaldehyde is added to blood in the proportion
of 1 to 166, after a period of four days, or after the action of smaller
amounts for longer periods of time. With amounts adequate to
prevent contraction of the clot and putrefaction, the blood shows no
obvious spectroscopic change even after the lapse of weeks.

8. Blood can be well preserved for a number of days at room
temperature by adding potassium oxalate in sufficient amount to
prevent coagulation, and formaldehyde in the proportion of 1 to 1000
to 500 or more, to retard laking and putrefaction. In the series of
such tubes given in Table II, Tube 53 showed only a small amount
of laking after being kept in a warm room for ten days. On the sub-
sequent addition of water, rapid and complete laking occurred.

The amount of potassium oxalate necessary to add to blood was found to vary
considerably in different dogs. In four experiments, the largest amount
required was 1 part to 300 of blood; the smallest, 1 part to 700; the
average amount was 1 part to 534. Arthus recommends 3 parts to 1000.

Blood to which sufficient potassium oxalate solution has been added to prevent
coagulation, after standing some days at room temperature, will frequently
be found to have undergone coagulation, as has been pointed out by
Schafer." In no case observed by me has this solidification occurred
under five days from the time of drawing the blood.

In tubes containing a sufficient amount of formaldehyde to prevent clotting,
and in tubes containing a sufficient dmount of potassium oxalate to prevent
clotting, and to which small amounts of formaldehyde have been added,
brownish, and more or less gelatinous precipitates, which float or remain
suspended near the surface of the liquid, appear in periods of time,
ranging from a few hours to several days (depending on the amount of
formaldehyde present).

1 SCHAFER: Text-book of physiology, i, p. 169, footnote 7.

L[nfluence of Formaldehyde on Coagulation of Blood. 193

AA SIE)S, 1s

Proportion
of formal- Time.
dehyde Serum Result.
to blood acted.
or serum.

1% formal- Serum
dehyde. Se qus:

1.0 2.0 3c 30 10 min.| Very strongly laked.
0.1 1.0 2.0 10 Less strongly laked.
0.2 1.0 2.0 10 Not laked.

0.3 1.0 2.0 10
0.4 1.0 2.0 10
0.05 2.0 1.0 16 Slightly laked.
0.10 2.0 1.0 10D se Notilalked:
0.15 2.0 1.0 10 a3 e

form. 0.5%
0.10 2.0 1.0 10 Strongly laked.

0.05 2.0 1.0 10 More strongly laked.

Formaldehyde injection
into veins in propor-
tion 1-1500 of the
blood.

asm

| 2.0 | 1.0 at ore As strongly laked

Charles Claude Guthrie.

194

praynd Apystys
‘89 01 8b ‘peye] Ajaywsapoy] Sulyey Ou SOT se auieg) Cy OST ee OST
ous Ayal jos Tojoo ur g sv ates fAypal os | Est 2 Ve ReL G00
” ” tuntes ON § sv eutes tooo + ATTaf | 6EE WI OLT 0c OL
” ” wintas ON CU EECNN ETS I SyeilL Sel OS? 00'0T
“Te Weyy ssoTy uinies ON p Se oues | 122 tel SOs OlOl
‘Sh 0} 1Z prynd ‘payer 7 Surye[ ou fT seauesg]| oy eel 7s OVOL
Q se yep
ity Ayal aos | fpayey A[suors Sg uryzzo[9 a10WW | OFSZ | O8 Sel | 0901
yep
a a winias yonyy | At9a fpayey Apsuors {}0]9 weg | OZST OL Spey OO'TI
. 5 wintas ON Ayjal £9 uey} rayxaep fpayey | o¢e 09 Z6LI CLOL
= v wnias par doi Ayjel f payer $s ueyy 1oy1eq | Ozs os C'CQT 00'IT
- peaxr]
- winias ON | ftay1ep yonu Ajuo + se sues} Q¢z OF $'Z9Z OS‘O1
peyr]
5 i winias ON | fieyr1ep Apysis Ajuo ¢ se owes | COE o¢ CLLZ OIL
"1Z ueyy sso'T wnias ON layiep Ayysys Ajuo 7 sve atures | ¢9z 0Z ants 0Z ‘01
‘Ob 01 [Z wnies oN | Suryey10 winsas ou fyo;9 Ayal payy | 06Z OL SZIS SZ OL
SQ 0} Sb piynd ‘peyey | umaas raja ola yaep ‘Suryey ON | OTT 0 Re OO'0T |
“sinoq, *so08 “soos ‘oo
*£19}.18 OT
Ar suiuado |pooq oj
“‘SULYR] *sinoy 99 *sinoy [Z as Jayje | epXyap| ‘s-b
jo owty, IJV IOYV ne “| umerp | -[eur1o0f | poor |
TL | poolq jo
awty, juonniiq

TUL GETS aL,

aT
ho
% 1
199%

(qsa1j) “WA40F aba):

(ysoay) “W407 0°99

(ysoug)

“WIOJ ‘O'D +

(yso1j) “WI10J
(ysoaz) “WI0F %¢ *9°9 OT
(yso1) ‘ur0y YT 9°98
(yso1g) “W40F % O99 ZT
(ysazg) ‘wey % [99 9
(ysasj) “m4IOoF %¢ °9°9 80
(ysoay) “wit0y % 1 “9°9 +
(Ysa) “Wu0F %¢ 9°9 $0

(yserj) “wA0y Yl 99 Z

*syuasvay

xo

Dm

‘ON OQNL,

195

Lifiuence of Formaldehyde on Coagulation of Blood.

”

”

‘89 93 Sb

”

”

”

”

”

”

”

”

”

”

”

”

"89 01 Sb

”

Pr

”

”

”

”

”

”

”

”

”

”

‘1Z uey) sso’T

ge ueyy

pind pue pay] sso'y
piyjnd pue

poyey, Aysuoys A1aA

Aqjal 30S

wn19s ON

wnies ON

pind pur paxry o.10 py
piynd

pue poy] 1a5su014S
prand a1ow

pue poy] 1asuo01s
pryand

pue peye, tasuo0ijs

sutkjoajnd

pue suryry yysys

JO[D Jeg

Ayal yos

wnias ON

umdas ON

eZ uvy} pIny as0jy

61 sv ames

wuias ON

wintas ON

evuseid avajo !sajosndio09 pay

vuseyd swaps fsaposnd109 pay

Joo
OU 10 JYSITS {10[09 ur QZ se aWeS

17 sv ouieg
9z sv owes

6g sv awieg

Surv, ou
‘tunias arom ATuo ZE sv ouies
Sule, ou
Te sv awes
suryey ou
‘umes arom ATUO 6% se auwieg

fwmiaas a1our yuo

SuUIyL, Ou fT sv ames
Suryey ou { unies

a1ow Al}Yysys A[UOo OE sv owes
PITOs

a1ow Apyysiys A[UO $Z sv awes

pljos a1our ATUo ¢Z sve aures
Zz sv owes
[Tz sv ewes
Toy 1ep
Ayjysts £g se Aouazstsuod ouies
P 61 se
AQuaysIsuOd puke JOTOO UL owes

Joyiep Apystys ATUo JT sv owes

joyiep AQysips ATU QT sve aues

j

‘exo ‘ssvjod % T ‘99 ¢}:

U (ysarz) ‘ut107 % T -9°9 cQ'0s

S

*[exo ‘ssvjod ¥, [0:9 El

{ (ysary) 103 Y [ *9°9¢Z0'0S

(ysarj) “Ww10y % [9:9 8
(ysa1z) "WI0F % | ‘9°99
(ysaay) “WI0F % T *9°9 +
(ysaaz) “W107 % T 99 Z
ID®N %6'0 99 8
IO®N %60 99 9
IOXN %60 '9'9 +

ID®N %6'0 '9'9 Z
(ysoz) “WI0F % T°
(ysaagz) Wa0f YT ‘9:9 9
(Ysa) wt0oy YT 9°9 +
(yso.1z) ‘utt0y YT ‘9°90 Z

(plo) "w10f %,
(pfo) waoy % [9°99
(plo) ‘w40y % -9°9 $
(plo) ‘w40g % T *9'9 Z

ace TS ueyy Cp uey} ‘901d uMOIq pue r
payer, erow A[}Y SIS | eursejd a1ow {To sev owes IO[OD “i ZSb POLL | SLOT } idee ie Ear zs
Bury “1s | winds “jue [leu 9 ‘aod Lie
» » Ef ULY} pyr] sso] | pur (2) Aqpaligos ‘ep svowwes10joD | 9EE lly cIEL | OS‘OL 7 (ysazyz) ‘w0x % [ ‘9°9 SOs 1s
"89 93 8b peyey Surye] ou fT se oueg | Sp Ost a 00‘0T De AES eB tea a es ae

Sp uBY] WINIAS dI0] pros a10ur ATuo ye se awies

it one
%1 ‘999

6b

(ysougz) “W105

wmnias su0g Jayos AYss 10 9¢ sev aes Qt

(qsorz) ‘w10j

V2é.

Charles Claude Guth

196

”»

[Zz ueyy ssa’T

”

“Op ULY] SSa'T

"89 93 Sb

” ”

‘Sp uv Sso'T

"89 93 Sh

*smoy

“SULYL]
ZOROUILT,

wn.ias ON

winias ON

fp ueyy

peyxe, stow ApZYSI[S
cp weyy

peye, e1ow AY SILS
aqny *xo ‘jod ysappa.

‘Gp UY} Paxye[ ssoT

Ip wey} peye, ssoT
6£

uey} poyxry ssay Ap YSIS
prynd

pue payer, A[suoi3ys

ce sev ames

$¢ sv owes

eused ur oa1d a10u
yon pure +f} uryy 19x1ep APYSIS
ewsyyd ut ‘oad umoiq
d10W PUL Cf URYY AByIep A[WYSILS
eusvid ur ‘said “ywe [Teuls
pue iapper ApYsts. Zp se ours
peye, Apysis pur sajosnd1oo
lappea Ayysts ApUO T- sve awies
pexe] pue saposnd109
Jappar Apysys ATuo ge se autres

Sulyv] ou fT sv owes

Ofe

“$008

§
§
1(
S

(yso1y)
(ysozz)

(yso.y)
(qso.j)

(yso.y)

‘WO YT 99 $

HL ‘99%
ht '9"9 a

"WULOF

‘Texo ‘ssvjod

(ysoug) “wr0¥ %[ ‘9° 10

[exo Pee UTOONG Il
“wos YT ‘99 9:0 |
‘ssviod pes oe)
“u10j Y hl 2' $f
‘ssviod ¥/ Y 7199 Sl

‘wu0j 4, LT 2° 70S |
[exo eS WEE AED 2s

[exo

ee

(Ysa) “WIOF wT Pate) 03)

LY

| Ov

Sb
+r
eb
ch
It

‘sInoy 89
19qJV

‘SsInoy [Z
IJV

"JO]9 03
pormbe.
SUED

*A1oq1e
suruado
laze
uMeIp
poolq
oul |,

"( panu1uo?) —JTT ATAVL

OT

uoKnytC|

1} 0)
poold

‘ON 9qnNL

‘sjuasvay

197

L[nfiuence of Formaldehyde on Coagulation of Blood.

UONY[NSeOS ZYSIS Jou 10 JaYJaYA ATayUYyap Sururezsiaose Jo AyNoyIp ey} 0} Sutmo ‘pasn Ayosoo] Jayj}eI oie VIUsy7PP PUL MeNLAs SWI} OY],
‘AJIPIOSIA poyAVUL B JNq UOT}L[NSvOD on1] OU PaMoYs Eg pue TS seaqn
Kes |e : (IS ters 5 L

‘JO]D pooyq sy} Sururezuod aqnjz oy} YSnor1y} UMEIpP svA\ jT

WOIJ O19Z 0} UAT[LF pey Iinssaid oy} oye ‘SUTUT + JNOGE UMBIP SLM POO] sy} Jey} UT pouoUaT ysn{ sauO dy} WOY siazyip [9 equ,
‘aqn} uvapo & Yysnory} Ar1aj1v Surpuodso1109 ay} WOIT UMEIP SPM POOTq 9Y} ‘199}¥] BY
UT O[LYA JO]D poolq Sururezuos oqny & YSNoIY} UMLIP SA IAUIIOJ OM} 9Y} UL POoT ay} yeVY} Ur ATO AaHIp OO pure “6g ‘gg seqny, :aLON

‘Op Uv] SSA’T

”

89 °F Sb

‘[z ueyy ssavT

*soqn} 94} JO 9swWos UL patangs0

-erxAydse

pexey Apsuo33s
SULYL] OU 10 JYSITS
pe ihe t

PEP EAL

gg uey} WInIaS I10]
Ayal aos

wiIn1as BWIOG

WNnIIS ON

ZS UeYy sso] pue
[TS uvy} pay] a1out "[S

SULyR] OU faqny
« (ewiou,, Aue uey} winies sso'T
Sulyey ou
foqny reptuis Aue jo wines ysojq
Sg sv oules
Suryeyl ou ST sv ames
pl[os a1ow AjUo 6 sv owes
PIfOs o10ut ATUO QF sv oWIeS

Jp se ouieg

OF se aes

SUIye[ OU
£(z) umes atom A[UO ZS se auIeS

SII
66

LS¥
Ill

()SS9L

O0¢ ‘OL

So Ll

(ysazg) “W103
(qsaay) “u1103
(ysory) *WI10F
(yso1z) ‘W10J

*yexo ‘ssejod
(qysorz) *WUIOF

%1'F98
%1'999
Hl ‘VI
%1 99%

%T ‘99 T

VENOUS. PRESSURES.

By RUSSELL BURTON-OPITZ.

[From the Physiological Laboratory of Columbia University, at the College of Physicians
and Surgeons, New YVork.]

CONTENTS.

Page

Normal venous pressure). / sy 5 el cee a ne ae
Respiratory variations in venous pressure . .. . . : oo) 2 Ua
Changes in venous pressure resulting from certain cepeunieneal precedanee oan eee
Effect of compression of the right external jugular vein. . . .... =. . 204
Compression of both carotid arteries= 2). 7. 555) cee
Compression of the femoralvartery9 9: ~s)~- ‘= <0) os sneer
Effect of stimulation of the vagus’) 02 is) whe. ite | eee ee
Changes following section of the vagi . . seis Noes Ne
Changes resulting from opening the chest euepended respiracoaae Heiser alos | 2))2)
Summary . 2&6 uae Pe eee eh) cee

NORMAL VENOUS PRESSURE.

N this investigation dogs were used exclusively. They were placed
upon their backs with the posterior extremities slightly flexed and
abducted. The anterior extremities were fastened against the sides
of the chest and forced downward so that, by slightly depressing the
shoulder, more easy access could be had to the lower portion of the
external jugular vein. The head and neck were brought as nearly as
possible to the horizontal plane of the body.

All the animals experimented on were anzsthetized with ether.
Morphine was given only to the larger animals, and in comparatively
small doses. During the recording of the pressures, deep narcosis
was avoided.

The determinations were made with manometers (4.5 mm. tubing)
filled with a concentrated sodium carbonate solution of the specific
gravity 1,088. They were connected with the veins by means of T
tubes, the horizontal branches of which were very short and of the
same diameter as the vein. A stop-cock, interposed between the T
tube and the manometer, served to minimize the excursions of the
liquid. It need hardly be mentioned that the pressure, recorded in
this manner, is the average lateral pressure.

The level of the liquid in the different manometers having been

brought to that of the corresponding vein, all the clamps were quickly
198

Venous Pressures. 199

removed and the pressures read simultaneously during the next three
to five minutes. The average reading was subsequently derived from
these data.

The pressure was determined in the following veins: !

1. Left facial vein.

2. Left and right external jugular veins.

3. Superior vena cava, distal portion.

4. Superior vena cava, in the neighborhood of the right auricle.

5. Left and right femoral veins.

6. Left saphenous vein.

7. Right brachial vein.

The left facial vein was connected with the manometer peripherally
to the superficial laryngeal lymphatic glands, almost opposite the
angle of the jaw.

The pressure in the external jugular vein was recorded at the point

where this vessel leaves the surface and enters deeply into the groove

between the neck and the shoulder. A horizontal line drawn through
the tip of the manubrium marks the point of insertion of the T tube.
In some animals, however, it was necessary to place the tube two or
three centimetres above this level, because the shoulders were so
prominent that a correct adjustment of the manometer was im-
possible.

In many of the experiments a catheter was introduced into the right
external jugular vein by means of which the pressure was registered
in the distal and central portions of the superior vena cava.

For the sake of better orientation, the catheter was inserted diagon-
ally at first. Having reached the superior vena cava, 27.¢., a point
central to the bifurcation of the subclavian veins, a rise in the pressure
of the opposite jugular vein resulted, because the catheter interfered
with the influx of blood from this vein. The catheter being subse-
quently brought into the axis of the superior vena cava by placing
it against the side of the neck, the increase in pressure immediately
disappeared.

The entrance to the right auricle was determined approximately by
introducing the catheter, first, into the right ventricle, and subsequently
withdrawing it a sufficient distance. Ifa small catheter is used, and
the insertion is made in the axis of the superior vena cava, absolutely
no change in the level of the liquid in the other manometers results.

1 The nomenclature is the same as in ELLENBERGER and BAum’s Anatomie
des Hundes, pages 432 to 457.

200 Russell Burton-Opitz.

The T tube was inserted in the femoral vein midway between the
groin and the orifice of the vena postica superior, about 3 cm. below
the groin. In those cases in which the pressure was also recorded in
the left saphenous vein, the femoral vein was connected with the
manometer immediately below the groin, while the latter vein was
secured close to the median edge: of the gracilis muscle, near the tip
of Scarpa’s triangle.

The pressure in the brachial vein was obtained in but two experi-
ments, because of the difficulty of correctly adjusting the manometer
to this blood-vessel.

The distances between the different manometers were measured in
almost all the animals for the purpose of calculating subsequently the
average fall in pressure from the periphery to the centre of the circu-
latory system. Naturally, these measurements could be made only
approximately (air-line measurements).

The respiratory movements were recorded by means of a tambour
which communicated with the left pleural cavity. Although the
variations in intrathoracic pressure are of great importance in an
investigation of this kind, a detailed account need hardly be given.
As the experimental conditions were in all cases as normal as
possible, the changes in intrapleural pressure moved within normal
limits. The greatest negative pressure prevailed in Experiment 15,
the lowest in Experiment 7; in the former case it equalled 11 mm.
Hg, in the latter instance 6 mm. Hg.

The results obtained from eighteen experiments are comprehen-
sively arranged in Table I. To prevent confusion, and to have at the
same time a more ready means for comparison, the pressure is given
in millimetres of mercury. In transferring the values obtained with
sodium carbonate solution, the following formula was employed:

H (Na;CO; sol.) X spec. grav. (NagCO, sol.)
spec. grav. Hg

H (mm: He):

Table I shows very clearly that the pressure gradually decreases
from the periphery toward the centre of the circulatory system. If
the average value is taken, it is observed that the pressure in the left
facial is 4.6 mm. Hg higher than in the left external jugular vein, in
the vicinity of the chest. The manometers in these veins were on
the average 150mm. apart. The fall in pressure amounts, therefore,
to 1.0 mm. Hg for every 32 mm. distance.

(Ether-narcosis light in all cases.

Venous Pressures.

(AB ISH

VENOUS PRESSURE IN MM. Ha.1
Respiration normal).

ane |= | =| 2] as | $s KH +83
mm | 3 | 2 | $& | S& | Distal Central) & ”
2 8 Sn we port. port. |

4 kilos grams. oheeK,
1 13.5 | 0.08 | 5.1 | —0.48 | —0.4 | 5.8
2 11.5 | 0.08 | 4.6] 0.16 7.6
3 16.5 | 0.10 | 5.7| 0.80 | 5.2

4 14.0 | 0.08 | 5:7 | —0.4 242

5 16.5 | 0.10 | 6.0} 2.8* 0.6

6 20.0 | 0.08 | 65 | 0.7 2A | HA aie,

7 19.5 | 0.10 | 4.1 | —0.08 Cen er
8 18.0 | 0.08 | 36] 0.2 2.0 | -28 | 44
9 165 | 0.08 | 4.0 | —0.8 299% e909" | 516
10 17.5 | 0.08 | 5.6] 0.7 0.16 | 1.2 | 44
11 7.5 | 0.0 —0.4 PHT Wh 3 Abel a9
12 HO 10:12. |95:2 |) 2.0% 1A | 5.0
13 6.0 | 0.0 Oa" | “O16 | 5.6
14 6.5 | 0.0 08 2.4 63
15 24.5 | 0.15 2.4% =.4- 9) = 2:05) 6d
16 9.0|0.0 | 54) 1.2% OO) Waa ss
17 7.0 | 0.0 0.0 | —0.08 ays
18 13.0 | 0.08 ~0.5 1 VAG 64
Av. value | 14.1 5.12} 0.52 | —0.08 | --1.38 | —2.96 | 5.39
Highest v. G5") 98. | “O16.\ 16a nate 76
Lowest v. 3.6 0.8 0.4 2.4 4.8 | 32
Difference ZON HO |i O56) “3036s hoe

Right femoral
vein

|
|

5.6

4.0

5.42

6.8
4.0
2 Srl

Left saphenous
vein

| 6.8

8.4

TAZ

| 8.8
H Ned

Syl!

201

Right brachial
vein

4.2

57)

Sy)
2
SH

0.5

1 In this connection mention must also be made*of the determinations of the
venous pressure in sheep made by JACOBSON and reported in the Archiv fiir Anat-
omie und Physiologie, 1867, page 226.

202 Russell Burton-Opite.

In the larger dogs the distance between the manometers in the
facial vein and in the central portion of the superior vena cava, near
the entrance to the right auricle, measured approximately 290 mm.
The fall in pressure between these respective points is 8.08 mm. Hg,
or 1.0 mm. Hg for every 35 mm. distance.

The radius from this centre to the saphenous vein measured
approximately 380 mm._ The fall in pressure amounts to 10.38 mm.
Hg. The average decrease is 1.0 mm. Hg for every 36 mm.
distance.

The average fall in pressure from the left femoral vein to the
entrance to the right auricle amounts to 8.35 mm. Hg. The distance
over which this decrease takes place measured approximately 320 mm.
The average decrease in pressure is 1.0 mm. Hg for every 38 mm.
distance. From the foregoing data the average decrease in pressure
of 1.0 mm. Hg for every 35 mm. distance is obtained.

Experiments I, 3, 10, 13, and 17 show that the pressure in two
corresponding veins is very nearly the same. In fact, the differences
in the pressure of the left and right external jugular veins, on the one
hand, and between the left and right femoral veins, on the other, are
so slight that they might easily be due to the impossibility of adjust-
ing the zero-point of the liquid exactly alike in both cases. Experi-
ment 13 shows even a complete correspondence between the two
femoral veins, the pressure being 5.6 mm. Hg in each vessel.

If contrasted with the arterial pressure, the pressure in the entire
venous system shows only very slight variations in different animals.
This statement also implies that the pressure in a certain vein does
not vary greatly in different animals. The difference between the
highest and lowest pressure value is greatest in the left femoral
vein, but even in this instance it amounts to only 4.4 mm. Hg.

The difference in the pressure of the left external jugular vein
obtained from all the experiments, amounted to 3.6 mm. Hg. It
must be attributed, at least in part, to the fact that the T tube was
inserted in four cases (marked *) higher in the neck. Naturally,
the result was that the pressure was greater in these cases thanin the —
others. On the whole, it may be said that a horizontal line drawn
through the tip of the manubrium, marks the point central to which
the pressure in this vein becomes negative and peripheral to which
the pressure is positive’ The “danger line” of the surgeon lies,
therefore, in close proximity to the thorax.

Venous Pressures. 203

RESPIRATORY VARIATIONS IN VENOUS PRESSURE.

In the preceding enumeration of the experiments only the average
value of the pressure has been given, but we know that definite varia-
tions occur with every inspiratory and expiratory movement.

In Fig. 1, inserted to illustrate this statement, the respiratory
variations in venous pressure were recorded by means of a Hurthle’s
venous manometer, connected with the external jugular vein opposite
the tip of the sternum. The respiratory movements were recorded
by a tambour communicating with the left pleural cavity. The down-
ward stroke corresponds, therefore, to inspiration, the upward stroke
to expiration. The time-curve, written in fifths of seconds by a
Jaquet chronograph, served as the abscissa for the latter record, the
zero-line of the scale as the abscissa for the former curve.

. Mm. NA,C O,

io) S22 eee

(OS Se i CE Sey Se ae.

— * ae. SA m=. Oa:
we ri oa. J 2 U8 J ie J

TIME 4% SEC.

a ee ae ee OS ee eee ee Se eee A ee SS SS SSS SS ee
=]

-6 a F ee ee

2". Hae E E E 'E
FiGurRE 1.— The variations in venous pressure during normal respiration (external

jugular vein).

If a comparison is made with the calibration-scale, inserted at the
beginning of the venous pressure, it is seen that a considerable
variation occurs. In this case the pressure varied from +12 to —23
mm. Na,CO, solution, or 35 mm. in all (2.8 mm. Hg). It must be
remembered, however, that the inspiratory movement (J to E) was in
this instance deeper than usual; the greatest negative pressure at
the end of inspiration amounted to 11 mm. Hg.

Normally, therefore, the lowest pressure occurs at the end of in-
spiration, while the highest pressure is synchronous with the end of
expiration. The course of the respiratory variations in venous pres-
sure is therefore the reverse of those occurring in the arterial system.

For the purpose of determining the amplitude of the respiratory
changes in venous pressure, two manometers were employed, one of
which recorded only plus pressures, while the other was arranged in

204 Russell Burton-Opitz.

such a way as to indicate only the values below zero. Their simul-
taneous action was made possible by connecting the T tube in the
vein, first, with a Y tube, the branches of which led to the respective
manometers.

The pressure was recorded in four different dogs in the left exter-
nal jugular, facial, and femoral veins. Instead of the sodium-carbonate
solution, a normal saline solution was used. The values were sub-
sequently transferred into mm. Na,COs solution and mm. Hg accord-
ing to the formula givén previously. The respiratory movements
were normal in all cases. These four experiments gave the following
average results. In the left external jugular vein the pressure be-
tween its highest and lowest points showed a variation of 20 mm.
Na,COs; solution (1.6 mm. Hg). In the femoral vein the pressure
between its lowest inspiratory and highest expiratory value showed
a variation of 8 mm. Na,CO, solution (0.6 mm. Hg). In the left
. facial vein the respiratory variations ranged from 2 to 3 mm. Na,CO,
solution (0.16 to 0.2 mm. Hg),

The greatest variation in venous pressure during inspiration and
expiration is therefore shown most conspicuously in close proximity
to the chest.

CHANGES IN VENOUS PRESSURE RESULTING FROM CERTAIN
EXPERIMENTAL PROCEDURES.

Effect of compression of the right external jugular vein. — In some of
the experiments the right external jugular vein was tightly com-
pressed, while the pressure was recorded in the veins of the opposite
side.

The effect of this procedure always betrayed itself in a rise above
normal in the left facial and left external jugular veins. The increase
in the pressure was, however, more conspicuous in the peripheral
vessel, the facial vein.

In Table II the normal pressure is given in relation with the
pressure prevailing during the compression. It must also be men-
tioned that in this table, as well as in the succeeding, the experiments
are numbered in accordance with those dealing with the normal pres-
sure. The values are given in mm. Hg only.

Venous Pressures. 205

TABLE FT-

COMPRESSION OF THE RIGHT EXTERNAL JUGULAR VEIN.

Pressure during compression
(mm. Hg) in

Normal pressure (mm. Hg) in

Experiment.

Left external
jugular vein.

Left external

F : i in.
jugular vein. Left facial ve

Left facial vein.

3.6 0.2 5.2 1.4:
5.6 0.7 6.8 1.2
a2 2.0 6.0 2.6
at ee: 6.3 2.0

Compression of both carotid arteries. —In these experiments the.
common carotid arteries which had previously been placed in liga-
tures, were raised and compressed between the fingers.

The effect on the pressure in the left facial and jugular veins was
generally not immediate. Three or four inspirations were required
before the maximal decrease in pressure appeared.

Table III is arranged in the same manner as the preceding; the
normal pressure is inserted together with the pressure prevailing at
the end of the compression.

TAB TER WE:

COMPRESSION OF BOTH CAROTID ARTERIES.

Pressure at end of compression

Normal pressure (mm. Hg) in (mm. Hg) in

Experiment.

Left external
jugular vein.

Left external

: : Left facial vein.
jugular vein.

Left facial vein.

3.6 | 0.2 a2 —0.3

4.0 | —0.8 2.4 2.0

5.6 0.7 4.0 0.08
54 2.0 3.0 1.0
a 1.2 4.0

206 Russell Burton-Obitz.

Compression of the femoral artery. — lhe femoral artery was com-
pressed with the forceps opposite the T tube in the vein, z. ¢. about
three to four cm. below the groin.

Although a gradual fall in pressure always resulted, ne decrease
was surprisingly slight. Under these conditions the pressure ap-
parently assumes the value of the pressure prevailing in the next,
more central venous trunk, the iliac vein.

TABLE LY;

COMPRESSION OF THE FEMORAL ARTERY.

Normal pressure Pressure during
Experiment. (mm. Hg) in compression (mm. Hg)
femoral vein. in femoral vein.

ae
2.4
5.6

Effect of stimulation of the vagus.— The right vagus nerve was
stimulated in a number of instances with a tetanic current of suffi-
cient strength to cause a gradual cessation of the action of the
heart.

The effect of this procedure always betrayed itself in a rise in
pressure in all the veins. Attention has been called to this fact
previously by Klemensiewicz! who observed the changes in pressure
in the femoral artery and vein during the stimulation of this nerve.

The rapidity of the rise is proportionate to the strength of the
stimulus. If the heart is stopped slowly, the pressure increases at
an equally slow rate. If, however, the current is very strong, so that
an almost instantaneous stoppage results, the pressure rises very
quickly at first.

Naturally, the increase in pressure is also determined in a large
measure by the respiratory movements, but of course only in case
they do not cease in consequence of the stimulation. Supposing that
a rise of 30 mm. soda solution has taken place in the external jugular
veins solely on account of the backward stagnation of the column of

' KLEMENSIEWICZ, R.: Sitzungsberichte der kaiserlichen Akademie der
Wissenschaften. Mathematisch-naturwissenschaftliche Classe. Wien, 1886,
Sect. 3, pp. 66-84.

Venous Pressures. 207

blood from the right auricle, the next inspiration will cause a decided
fall in the level of the liquid, proportionate to the depth of this move-
ment. But, as soon as the expiratory phase sets in, the pressure
rises again, until its value is much greater than formerly. This
phenomenon will be repeated with every respiratory movement oc-
curring during the passive state of the heart.

It must be explained in the following manner: The inspiratory
movement draws a considerable quantity of blood into the central
venous trunks, but, as the blood is not propelled onward, this amount
is simply added to that already accumulated here. The pressure rises,
therefore, in a corresponding measure, not steadily, however, but by
degrees.

The heart was kept in the passive state for a comparatively long
time. In Experiments 8 and 10 the stimulation caused an almost
immediate cessation of respiration, with the chest in the expiratory
position. A weaker current being used in Experiments 9, 11, and 12,
at least one complete respiratory phase occurred after the heart had
ceaséd to beat. In Experiment 11, four respiratory movements took
place during the passive state of the heart. The results clearly show

e. TABLE V.

STIMULATION OF THE VAGUS.

Pressure at end
Experiment. Pressure in Normal pressure, | of stimulation,
| mm. Hg. mm. Hg.

Left external jugular vein
Superior v. cava (dist. p.)
Left femoral vein

STF Oo

ANDO WAD

Left facial vein

Left external jugular vein
Superior v. cava (dist. p.)
Left femoral vein

|

MNO BO
Mwbo

Left facial vein

Left external jugular vein
Superior v. cava (dist. p.)
Left femoral vein

|

ON
ANA
mor LY

Left external jugular vein
Superior v. cava (dist. p.)
Left femoral vein

|

WHS £OOM une
wwe eT ie Me) AnNnwao + ow

Left facial vein

Left external jugular vein
Superior v. cava (dist. p.)
Left femoral vein

Oe wur

oFOWN
Ce ease eC A
SHAD BWA

208 Russell Burton-Oprtz.

that the increase in pressure is particularly noticeable in those cases
in which the respiratory movements were not abated completely.

After the stimulation the pressure decreased gradually to normal,
every heart-beat causing a distinct fall in the pressure.

In the experiments contained in Table V, the normal pressure is
given in relation with the pressure prevailing at the end of the period
of stimulation.

Changes following section of the vagi.— This procedure was invari-
ably followed by a fall in venous pressure, but the decrease was not
equally evident in all cases. Experiments 14, 15, and 16 (Table V1)
show a very decided effect; in Experiments 13 and 18, on the other
hand, the decrease is less conspicuous.

TABLE VI.

THE CHANGES FOLLOWING SECTION OF THE VAGI.

Normal Pressure after | Pressure in left
Pressure in pressure. nerve section. | pleural cavity.
mm. Hg. mm. Hg. mm. Hg.

—3 to —18.5
Resp. slow.

Left external jugular vein
Left femoral vein

Left external jugular vein
Superior v. cava (dist. p.)
Left femoral vein

Resp. frequent.

wesw wd

—6.5 to —21
Resp. frequent,
deep.

AHS WS NWA

Left external jugular vein
Superior v. cava (dist. p.)
Left femoral vein

Left saphenous vein

—175
Resp. frequent.

Left external jugular vein
Superior v. cava (dist. p.)
Left femoral vein

Left external jugular vein
Superior v. cava (dist. p.)
Left femoral vein

Resp. slow, deep.

COR HOD WARS ADANWSO HWA

This slight inconsistency in the results cannot be ascribed to a
variable increase in the frequency of the heart, because, even if a
larger quantity of blood leaves the heart, an equal amount must be
returned to this organ. Evidently it must be attributed wholly to
differences in the rate and depth of the respiratory movements.

The deep inspiratory movements usually following section of these
nerves, naturally are accompanied by a very decided fall in venous
pressure. When the expiratory phase sets in, followed by the long

Venous Pressures. 209

pause, the pressure increases again and rises the higher, the longer
the interval between the successive respirations. In brief, the
respiratory variations in pressure are extremely conspicuous.

It may readily be assumed that the pressure must be lower in
those cases in which the respiratory movements are more frequent
and deep. The longer the duration of the respiratory pause, the
higher will the pressure become during this interval. These differ-
ences in the frequency and depth of respiration are always present
after section of the vagi, and undoubtedly are responsible for the
variable conspicuousness of the decrease in pressure observed in
Table VI.

The excursions of the liquid being slow, so that secondary oscilla-
tions could not occur, the highest and lowest pressure-values were
recorded, the average being obtained subsequently by calculation.
The pressures were read for about two minutes after the section of
these nerves.

Changes resulting from opening the chest (suspended respiration ). —
The normal venous pressure having been obtained, the chest was
opened, and the pressure again recorded about one minute after this
operative procedure. In Experiments 9 and 12 the tissues of two or
three intercostal spaces were cut and torn sufficiently, so that the
“reflex” respiratory movements, following the collapse of the lungs,
could no longer influence the level of the liquid in the manome-
ters. In the other experiments, the abdominal cavity was opened
first by an incision in the linea alba. The diaphragm was divided
subsequently along its anterior edge.

The heart-beats were counted before and after this operative pro-
cedure. They were found to possess normal force and frequency at
least for some time after the collapse of the lungs. The very
decided changes in pressure noticed after the suspense of respira-
tion, therefore cannot be attributed to differences in the action of
the heart.

The pressure in all the veins rose immediately with the first
rush of air into the chest. The increase continued gradually for
about thirty or forty seconds, when the maximal height was attained.
The rise, as is evident in Table VII, was often very considerable.
All the negative pressures disappeared, at least as far centrally as the
distal portion of the superior vena cava. The important question,
whether the negative pressure is also destroyed by this procedure
directly in and near the entrance to the right auricle, could not be

210 Russell Burton-Opttz.

determined by the method employed in this investigation, because
the correct adjustment of the manometer with the chest closed: is
impossible. An approximate adjustment is not permissible in this
case, because under these conditions even the most minute changes
must be taken into account.

These experiments clearly show, that the negative pressure normally
prevailing in the central venous trunks is wholly dependent on the
rate and depth of the respiratory movements (aspiration of the chest).
If the right auricle and ventricle are able to produce an active negative
pressure independently, this pressure must be confined to the limits
of the heart. As is shown in Table VII, the pressure in the distal
portion of the superior vena cava on opening the chest rose as an
average to +19.5 mm. soda solution,-or 1.55 mm. Hg. Considering
the distance between this point and the right auricle, and the fall in
pressure per millimetre distance (see page 202), we must conclude
that the pressure was positive even at the entrance to the right side
of the heart.

On opening the abdomen no decided change in pressure could
be observed. An increase, amounting to 5 mm. soda solution, was
clearly recognized in only one instance (Experiment 10). It took
place in the distal end of the superior vena cava.

Table VII contains the normal pressure in relation with the pres+
sure prevailing shortly after the chest had been opened and the

respiration suspended.
TABLE. VII.

THE CHANGES RESULTING FROM OPENING THE CHEST (SUSPENDED RESPIRATION).

Pressure after
Experiment. Pressure in Normal pressure, opening chest,
mm. Hg. mm. Hg.

9 Left facial vein 4.0 6.4
Left external jugular vein —0.8 4.8
Superior v. cava (dist. p.) Shs 2.4
Left femoral vein 5.6 12
10 Left external jugular vein 0.7 2.9
Superior v. cava (dist. p.) —0.16 1.6
ll _ Left external jugular vein —04 a0
Superior v. cava (dist. p.) =17 14
Left femoral vein ou, 3.2,
2 Left facial vein Be; 6.5
Left external jugular vein 2.0 3.6
| Superior v. cava (dist. p.) —l.4 0.8
| Left femoral vein 5.0 6.0

Venous Pressures. 211

If after this period of suspended respiration, the lungs of the
animak were: inflated artificially, another conspicuous change took
place. Normally the pressure falls during inspiration and rises dur-
ing-expiration (see Fig. 1). If artificial respiration is employed the
reverse phenomenon occurs, z. ¢. a rise on inspiration and a fall during
expiration. :

The upper line of Fig. 2 represents the expansion of the lungs by
the artificial air-current. The inspiratory phase continues from J to
E and the expiratory period from Eto J. The ordinates are indicated
by the same letters in the curve of the venous pressure. The respira-
tory phases were recorded by a tambour, connected by means of a T
tube with the tracheal cannula. In the case now under consideration

E E E jE E
eae gat bs phe 6 ee Sey ; sp Tras
| = —~_} oat / —
NATE CI SB Pe Se YW ap Yr YW UW
TiME Ys SEC.
go MM Na,CO, rn [{~ iene
(é(a)— DENN (4 RN Lr 4 Nin al M\n ANnar y i E Nami) ENS
40 ! ; Ea cy Re ; E
20 J E J J E J J
°
FiGurE 2.—The variations in venous pressure .during artificial respiration (external

jugular vein).

the stop-cock, interposed to lessen the excursions of the writing lever
of the tambour, was closed somewhat too tightly. The inspiratory
phase is therefore prolonged beyond its real duration. The venous
pressure was recorded as in Fig. 1, by a Hiirthle’s venous manometer,
connected with the external jugular vein in close proximity to the
chest.

As Figs. 1 and 2 were obtained from the same animal, it might be
well to compare’ the pressure-values of these two curves. In Fig. 1,
the pressure varied from —23 to +12 mm. Na,CO, solution. The
chest being opened, the pressure rose considerably above normal,
namely, to 63 mm. Na,CO, solution (5.0mm. Hg). Artificial respi-
ration being employed subsequently, the pressure varied from 58 to
83 mm. Na,CO, solution.

Thus, the lowest pressure value is reached in the latter case not at
the end, but at the beginning of inspiration. Ina similar manner,
the highest pressure which normally occurs at the end of expiration,
is in this case synchronous with the end of inspiration.

212 Russell Burton-Opitz.

The excursions of the liquid in the manometers are proportional to
the frequency and the force of the artificial currents of air. If a great
amount of air is forced into the lungs, a corresponding increase in
venous pressure follows. The effect of artificial respiration seems to
be purely mechanical. During the inspiratory phase, the pulmonary
vessels are compressed, while during expiration they are more free,
allowing during the latter period a greater flow from the right side of
the heart.

SUMMARY.

Among the conclusions derived from the preceding tables, the
following should be noted particularly :

1. The average venous pressure in eighteen dogs, varying in
weight from 6 to 24.5 kilos, was as follows:

Left facial vem =) 060 3) es ot eee nee pO elena eel es
Left external jugular vemos .9 2). =): ne) ean O cae
Right external jugular veins. 8. 202 es). 20 UGmmnne
Sup. vena cava, distal portion . . Siege ee llests;
Sup. vena cava, near entrance to eke antici 2.00 ae
Right brachial) ves <9 we |; ea co ne ot ee
Weftifemoralaviern. cies) ee) ee ener
Right-feémoralvein ace 0. eee ee Oe
Iteftrsaphenous vein (02) lee er 0 a ed

2. The pressure decreases gradually from the periphery toward the
centre of the circulatory system. The fall in pressure amounts to
1.0mm. Hg every 35 mm. distance.

3. The pressure in two corresponding veins is very nearly the
same.

4. The pressure in a certain vein does not vary greatly in different
animals. As compared with the arterial system, the variations in
pressure, in the entire venous system, are very slight.

5. In an animal with normal heart action and respiration, negative
pressure appears first in very close proximity to the thoracic cavity.

6. Compression of the right jugular vein causes a decided rise in
the pressure of the opposite facial and jugular veins. Compression
of both carotid_arteries is followed by a distinct decrease in the
pressure of the veins just mentioned. Compression of the femoral
artery causes a slight fall in the pressure of the corresponding vein.

7. Stoppage of the heart by stimulation of the vagus is followed by
a very decided increase in venous pressure, due to the accumulation

Venous Pressures. 213

of the blood in front of the right side of the heart. The rise is more
conspicuous in the central venous trunks.

8. Section of both vagi is followed by a fall in venous pressure,
which, if the respiratory movements are favorable, may become
extremely pronounced.

g. After opening the chest, the pressure rises far above normal in
all the veins. All negative pressures disappear, at least as far cen-
trally as the entrance to the right auricle.

10. Normally the pressure falls during inspiration and rises during
expiration. Artificial respiration being employed, the reverse phe-
nomenon results.

The respiratory variations in venous pressure are most conspicuous
in close proximity to the chest.

A STUDY OF THE PHYSIOLOGICAL ACTION AND
TOXICOLOGY. OF CASSIUM CHLORIDE:

By GA, PANE ORD:
[From the Sheffield Laboratory of Physiological Chemistry, Yale University.]

CONTENTS.

Page

Introductory . . : wR OR eee eo Ec
General effects of caesium piteiide ni OA SOME Sos he: th OR
Experiments on the frog «6 = fi be = a ancl poly
Eexpeximentsvontab bits easei rmeermeners
Experiments‘on-tats) ({.0..°0.- 4. 6G) so es al a) err
Experiments on dogs . . 2) 5 het. 60 8 605 GD ot ie ener
Metabolism experiments . . . Soe ea oe emg ee” en
General considerations on iactabeleds oe Sl ND
The elimination of czsium <i). -6 Yipes. cd. we
Summary . 0.0) je ee eS Mee ae ol errr

LTHOUGH cesium belongs to a group of elements of which
several members play an important rdle among the inorganic
compounds that occur in living tissues, its physiological action has re-
ceived little attention. There are a few researches on record in which
the comparative action of the various salts of the alkali metals have
been studied. Thus Richet! determined the absolute and relative
toxicity of the latter on the frog’s heart, and expressed their relations
in the following table:

Molecular Absolute Molecular
weight. toxicity. toxicity.

Cesium 133.0 100 0.74
Rubidium 85.4 43 0.51
Lithium 7.03 Pi 39
Potassium 39.15 26 0.67
Ammonium 17.04 25 1.4

‘ RicHET: Dictionnaire de physiologie, 1898, iii, p. 73, where the earlier
literature is cited. The figures express the quantity of metal per litre of which
four drops suffice to paralyze the frog’s heart.

214

Physiological Action, etc.,.of Castum Chloride. 215

From these figures it appears that equimolecular solutions of caesium
possess a toxicity comparable with that of potassium and rubidium,
the other alkali compounds standing in a group by themselves.

Brunton and Cash! made a comparative study of the action of the
alkali chlorides upon muscles and nerves. According to them, neither
rubidium nor czesium paralyzes the motor nerves until large doses are
employed. ‘The contractile power and also the duration of contraction
is increased by caesium. The numerous other details ascertained need
not be repeated here.

Ringer? has contrasted the action of casium and rubidium salts
with that of potassium salts by perfusion of the ventricle of the frog’s
heart. He concluded that rubidium, which is chemically closely allied
to czesium, is almost identical in its physiological effects with potas-
sium, from which cesium differs markedly. The chloride of caesium
showed a behavior more like the salts of barium and strontium. In
criticism of these results, Blake ® states that rubidium and cesium agree
with the other metals of the group in their action, potassium being the
single exception. He observed that ‘‘ when injected into the arteries,
the caesium salts cause a sort of cataleptic state of the voluntary mus-
cles, and a curious slowing of the reflexes, two or three seconds, for
instance, elapsing between irritating the cornea and closure of the
lids.’ He adds that “ this has not been observed in any other of the
monovalent metals.”

In an effort to establish experimentally a relationship between the
chemical properties and the physiological reactions of the alkali
metals, Botkin * studied the effects induced on blood-pressure by the
different alkali chlorides. In the case of rubidium chloride, intrave-
nous injection of 0.03 gram per kilo of body weight in the dog was
followed by a marked rise in arterial pressure frequently amounting
to one hundred mm. Hg, whether the animal was curarized or not.
The high pressure was maintained from one-half to four minutes,
according to the dose. The rate of the heart beat was not always
affected in the same way. The force of the cardiac contractions, how-
ever, was increased and remained so, even after the pressure fell again.
Frequently the rise in pressure was preceded by a slight initial fall.

1 BRUNTON and CAsH: Philosophical transactions of the Royal Society of
London, 1885, clxxv, p. 197.
? RINGER: Journal of physiology, 1883, iv, p. 370.
3 BLAKE: Journal of physiology, 1884, v, p. 124.
- + BoTKIN: Centralblatt fiir die medicinische Wissenschaften, 1885, xxiii, p. 849.

216 G. A. Hanford.

The injection of 0.04 gram per kilo into a curarized animal quickly
caused the heart to stop in diastole. These phenomena were likewise
obtained after section of the spinal cord, although a larger dose was
necessary. The rise in pressure was not prevented by section of the
vagi or by atropine.

With cesium chloride analogous results were obtained, although
almost double the dose noted for the rubidium salts was necessary.
In slowing the heart, caesium chloride was apparently more effective
than therubidium compound. Botkin also observed that rubidium and
caesium resembled potassium in their action on the excised frog’s heart,
caesium chloride being the least toxic.

Harnack and Dietrich! have also compared the action of rubidium
and cesium chloride. They report that these salts have similar actions
on the striated muscle of the frog, differences in intensity alone being
noted. Thus because czesium chloride diffuses Jess rapidly, its physi-
ological action is less marked. The essential feature is a paralysis
preceded by an initial increase in irritability and contractile power.
The results differ from those reported by Brunton and Cash,? who
observed no initial excitability, but only an increasing paralysis.

GENERAL EFFECTS OF CASIUM CHLORIDE.

In the present investigation, the physiological action of caesium
chloride has been studied on the frog, rabbit, cat, dog, and man. A
few observations have been made on the effects of this salt upon iso-
lated tissues and cells from the frog in comparison with the reaction
of equimolecular solutions of sodium chloride. In experiments with
the mammals, attention was directed to the toxicological effects of
czesium chloride, its action on the circulation, the mode of elimination,
and the effects on metabolism. The cesium chloride used in all the
experiments was obtained through the kindness of Prof. H. L. Wells
of the Sheffield Chemical Laboratory. It was prepared as follows:
Pollucite from Paris, Maine, was extracted with hydrochloric acid, and
precipitated with antimony chloride as a double salt, 3CsCl,SbCl,.
This compound was then decomposed with ammonia, and the result-
ant caesium chloride treated with nitric acid to form czsium nitrate.

1 HARNACK and DiETrRIcH: Archiv fiir experimentelle Pathologie und Phar-
makologie, 1885, xix, p. 153.

* BRUNTON and CASH: Philosophical transactions of the Royal Society of
London, 1885, clxxv, p. 197.

Phystological Action, etc., of Caestum Chloride. 217

With iodine and hydrochloric acid, this in turn gave a double salt of
cesium chloride and iodine (CsCl,I), which was recrystallized, and
upon ignition yielded pure cesium chloride. Spectroscopic examina-
tion indicates that the salt obtained is unusually pure.

Experiments on the frog. — Frog’s blood taken directly from the
heart was dropped into % solutions of pure czsium chloride and so-
dium chloride respectively. Microscopic examinations made at inter-
vals showed no marked difference in the behavior of the cells in the
two solutions. The same is true of ciliated epithelial cells taken from
the roof of the mouth. In some experiments the movements of the
cilia seemed to diminish in rapidity somewhat earlier in the caesium
chloride solution. Carefully isolated gastrocnemius or sartorius mus-
cles were immersed in freshly prepared % solutions of the two salts.
The spontaneous contractions described by Loeb! were quickly mani-
fested. They always continued for a longer time in the muscle
Immersed in sodium chloride. After the spontaneous contractions
had ceased, irritability to electrical stimulation disappeared first in the
muscle exposed to the czsium solution.

When two leg-nerve preparations were made from the same ani-
mal, and the attached sciatic nerves were exposed to one of the
solutions, the loss of irritability to electrical stimulation was evident
in the nerve dipped in % czsium chloride solution sooner than in the
one immersed in the sodium chloride solution. For example, in one
experiment the cesium chloride nerve failed to respond at the end of
three hours, while the nerve in sodium chloride still reacted faintly
after eight hours. When neither nerve was longer excitable, the gas-
trocnemii, which reacted to direct stimulation, were in turn immersed
in the corresponding solutions. Here again response to stimulation
failed first in the muscle exposed to the czsium chloride.

When cesium chloride is introduced subcutaneously in the frog, a
complete paralysis results. The dose necessary to produce this effect
was found to vary from one to two-and-a-half milligrams per gram of
body weight. The paralysis ensued gradually, and in no instance was
any preliminary excitation noted except such as might be referred to
a local irritation at the seat of injection. The paralysis usually first
manifested itself in the fore legs, which sometimes became stiff and
sometimes remained limp. The reflexes were usually better in the
hind legs than in the fore legs. Electrical stimulation of the exposed

1 LoesB: Beitrage zur Physiologie; Festschrift fiir A. Fick, 1899, p. 101 ; also
This journal, 1900, iii, p. 136.

218 G. A. Hanford.

sciatic nerve of a paralyzed frog failed to occasion a muscular response,
although, in the cases examined, the muscles were still irritable directly.
The symptoms elicited in the doses mentioned resemble more closely
those described by Brunton and Cash than those noted by Harnack
and Dietrich. The initial excitability described by the latter was not
noted in the present experiments.

The reflexes were studied by Tiirck’s method. In the pithed
normal frog, a response was obtained by immersion in 0.4 per cent
H,SO, for from four to eight seconds. Half an hour after the onset
of caesium paralysis, responses were delayed-half a minute, this latent
period increasing as the period of intoxication continued. In the few
experiments undertaken to locate the specific action of the salt, the
results seemed to indicate a reaction on both muscle and nerves in
accordance with the observations of Brunton and Cash.

Experiments on rabbits. — In this animal doses of caesium chloride
as large as one and a half grams per kilo of body weight, introduced
subcutaneously, failed to provoke any noticeable symptoms. This
quantity, as will be seen later, is considerably greater than the fatal
dose for the cat or dog. Within two hours, the urine removed by
catheterization showed the characteristic caesium spectrum; and in
one animal, with the dose mentioned, the spectroscopic test for caesium
could still be obtained in the urine after nine days. The fzeces also
gave the test.

A rabbit which received two hypodermic doses of one gram per
kilo of body weight died seven hours after the first injection. Soon
after the second dose was administered, the animal began to lose con-
trol of its hind legs, and severe diarrhcea ensued. An increasing
weakness, leading up to complete paralysis, was followed by death.
The post-mortem appearance of the organs was not unusual. Czesium
was detected in the stomach contents, various portions of the intestinal
contents, and the urine.!

Experiments on cats. /eeding trial.— A medium-sized cat was given
two hundred milligrams of cesium chloride in one hundred cubic
centimetres of milk each morning. This was continued for eight
days, when the animal was killed. The food was taken readily
and during the experiment no toxic or unusual. symptoms were
noted, except a tendency toward diarrhoea. Up to the time of

1 The stomach contents were carbonized and extracted with water. The ex-
tract was concentrated and then tested spectroscopically. The other materials
were tested directly.

Phystological Action, etc. of Caestum Chloride. 219

the animal’s death the urine each day showed a_progréssively
stronger cesium reaction with the spectroscope. The faeces also
gave the characteristic spectrum. The post-mortem appearance
of the organs was normal in every respect; they gave pronounced
czesium reactions when tested spectroscopically, showing that caesium
chloride readily passed into the various organs and tissues of the
body.

Subcutaneous injection. — The effect of subcutaneous injection was
next studied. In one animal, weighing two and one-half kilos, a dose
of one and one-quarter grams (one-half gram per kilo body weight)
proved fatal in seven days. Within a few minutes after the injection,
the animal became restless, and soon intensely excited, biting at
anything within reach. This lasted about two hours, until fatigue
supervened. These peculiar symptoms are probably referable to local
irritation at the point of injection, rather than to any specific action
of the czesium salt. Both the urine and the faeces gave good czesium
reactions up to the time when the animal died. For four days after
the injection, the cat improved, finally taking food previously refused,
so that its actions appeared quite normal. However, from this time
on, an increasing weakness was noted, showing itself in a disin-
clination to stand, and a tendency to favor the hind legs. Dur-
ing the last two days, the animal refused food, and seemed unable to
control its legs, staggering when attempting to move. This lack of
control gradually extended towards the head, until shortly before
death the animal was scarcely able to move any part. Its reflexes
showed a corresponding progressive disappearance. Thus the paral-
ysis evoked seemed to begin in the extremities, especially the hind
legs, and gradually to spread forward until it involved the entire body.
As will be seen later, similar progressive paralysis was observed in
the dog. Death ensued when the paralysis was complete, all muscles
being relaxed. On post-mortem examination, there was found con-
siderable cedema of the subcutaneous connective tissue. The internal
organs were carbonized, extracted with water, and the extract evap-
orated for spectroscopic examination. The liver and intestine gave
the strongest reaction; the kidney, spleen, and pancreas gave fair
reactions; the brain and lung showed the spectrum faintly. The in-
testinal contents also gave aa good reaction; it is thus evident that
caesium may be excreted through the gut.

In two other animals receiving the same dose, similar symptoms
were provoked, with the exception that recovery ultimately followed.

220 G. A. Hanford.

The fifst one of these (three and seven-tenths kilos in weight) having
received 1.85 grams, was killed at the end of a month. No caesium
reaction could be obtained from any part of the animal. The second
cat (three and four-tenths kilos in weight) received 1.7 grams; in this
animal the characteristic symptoms described were less pronounced.
The urine and faeces ceased to show the cesium spectrum after ten
days. Further details regarding the toxic symptoms produced by
cesium chloride are given in the descriptions of the metabolism
experiments.

Effects upon the circulation. — In a cat of two and eight-tenths kilos
body weight, arterial pressure was measured with a mercurial mano-
meter. Anzesthesia was maintained by administration of A. C. E.
mixture, after a subcutaneous injection of forty milligrams of chloral-
ose. A total dose of two and one-tenth grams (three-quarters of a
gram per kilo body weight) of caesium chloride was injected into the
jugular vein during the experiment. After receiving about two-thirds
of this amount, the animal breathed more deeply and slowly. Ina few
minutes both the pulse and breathing became very irregular. With
this irregularity in the pulse came also a slight fall in pressure.
Within seven minutes the latter had given place to a marked rise.
At the same time the heart beat increased in force, becoming irregu-
lar. After the pressure returned to the normal, the remaining one-third
of the caesium solution was introduced, resulting in a second marked
rise in pressure, and finally death due to cardiac failure. This took
place before the injection was finished. The urine (from the bladder),
the bile, the thyroid gland, and intestinal contents from different por-
tions, all gave good czesium reactions spectroscopically. The stomach
contents gave the reaction after carbonization.

Experiments on dogs. Swbcutancous znjecttons. — A young bitch
weighing eleven and three-tenths kilos received eight and one-half
grams of caesium chloride (three-quarters of a gram per kilo body
weight). The initial symptoms of local irritation were observed,
though they were not so marked as in the case of the cats described
previously. Food was refused, and within two days the animal be-
came indifferent to its surroundings, remaining in any position in
which it was placed. Thus, when placed standing in a corner, it
would remain so until fatigue intervened. On the fourth day the ani-
mal appeared very weak, and was unable to control its legs. Whereas
heretofore the symptoms seemed to indicate indifference rather than
fatigue, now the animal appeared weak and exhausted, and when

Phystological Action, etc., of Casium Chloride. 221

placed upon a table was unable to hold up its head. Diarrhoea and
bloody urine were symptoms of the last few days previous to death.
The animal died in complete paralysis, all muscles being relaxed.
Upon post-mortem examination the subcutaneous tissue was found
highly congested, as was also the spleen. The blood clotted nor-
mally. All the organs tested, except the brain, gave a pronounced
caesium reaction.!

In another experiment a dog weighing ten kilos received five grams
(one-half gram per kilo body weight) of caesium chloride; and after
three days, a second dose of seven and one-half grams (three-quarters
of a gram per kilo body weight) was injected. The first dose pro-
voked the usual initial symptoms of local irritation and uneasiness,
loss of appetite, and diarrhoea. Beyond this nothing unusual was
noted. After the second dose, however, the animal was more dis-
tressed, and refused food for two days, being quite sick and subject to
frequent defecation and vomiting. This animal recovered somewhat
after a few days, and would eat meat. The urine and faces showed
bright caesium bands when tested spectroscopically. At the end of
ten days, the animal was killed. During the two or three days previ-
ous, signs of weakness and loss of control already noted in the other
animals were evident. There was no cedema of the tissues, and the
internal organs presented no unusual appearance. The caesium spec-
trum was plainly shown by all the organs and important tissues, and
by both the urine and faeces.

From these injection experiments upon both the dog and cat, it
seems demonstrated that caesium chloride enters all the tissues and
organs, and is excreted by way of the intestine as well as by the kid-
neys. Although the foreign element soon appears in the urine and
faeces, complete elimination follows only slowly, the last traces not
disappearing until several days have elapsed. A more permanent
retention in specific tissues such as occurs with iodine in the thyroid
gland could not be detected. A marked disturbance of the gastro-
intestinal tract was noted in all animals to which larger doses were
administered by any channel.

Effect on the circulation. —To study the effect of the salt upon the
circulation in the dog, two experiments were made. In the first, a
bitch of seven kilos weight received three small doses and a large one.
Anesthesia was maintained by the injection of forty milligrams of

1 These organs were: kidney, liver, spleen, lung, intestine (with contents),
muscle, and brain.

rep: G. A. Hanford.

morphine sulphate and four milligrams of atropine, followed by A.C. E.
administration. Blood-pressure was measured by a mercurial mano-
meter connected with the carotid. The lymph flow was measured by
means of a cannula introduced into the thoracic duct. The urine was
obtained by catheterization. The injection of one-tenth gram into
the femoral vein caused no noticeable change of any sort. Aftera
few minutes, one-half gram more was injected, which caused a slightly
more rapid flow of lymph. The giving of a third dose of three-quarters
of a gram had no further effect. Finally two and one-half grams were
injected, which caused the animal’s death immediately from cardiac
failure. A few respiratory efforts were noted after the heart had
stopped beating.

The urine taken after the second injection (one-half gram) showed
a faint caesium spectrum. From this time on the reaction grew
stronger. After the second injection, the lymph also gave the spec-
trum test, which grew stronger as more was introduced. The liver,
kidney, and thyroid gland all gave good czsium reactions.

In the second experiment a dog weighing fourteen kilos was used.
Anesthesia was maintained by the injection of eighty milligrams of
morphine sulphate and eight milligrams of atropine, followed by the
administration of A.C. E. mixture. Cannulz were inserted into the
lymphatic duct, both Whartonian ducts, both ureters, the right fem-
oral artery, and the right jugular vein. The chorda-lingual nerves
were both cut. The lymph flow was normal, as was also the urinary
secretion. Adose of fifty cubic centimetres of a four per cent solution
(two grams) was injected, which brought out the usual symptoms,
though less marked than in the previous experiment. There was an
initial slight fall followed by a rise in pressure, great irregularity in
the heart beat, and deep, labored breathing. After the effect of this
dose had disappeared, a second one of twenty-eight cubic centimetres
of the same solution (one and twelve-hundredths grams) was injected,
causing similar effects. Pilocarpine hydrochloride was then introduced
into the circulation to provoke salivary flow. The second drop of
saliva from each submaxillary gland showed the spectrum of cesium,
as did the saliva dripping from the mouth and the secretion from the
nose. Within twenty minutes from the first injection czesium could
be detected in the lymph flowing from the thoracic duct. As the ex-
periment progressed, the rate of lymph-flow increased about fifty per
cent. Within about fifteen minutes after injection, the salt could be
demonstrated in the urine flowing from either ureter.

Phystological Action, etc., of Cesium Chloride. 223

The appearance of this salt in the lymph was more tardy than in the
case of some salts, as reported by other investigators. Thus, Tschir-
winsky | observed sodium salicylate in the lymph within four to seven
minutes; Cohnstein? found four to five minutes to intervene before
potassium or sodium ferrocyanide reappeared; and Mendel? noted
about the same length of time for sodium iodide.

METABOLISM EXPERIMENTS.

Methods. — The experiments were performed on healthy full-grown
dogs ranging in weight from five to fifteen kilos. The animals were
kept in cages suitable for the separate collection of solid and liquid
excreta. When possible, the animals were catheterized every day, and
always on the final day of a period. The urine for twenty-four hours
was combined, and its volume, etc., were noted. The fzces, when
dried and weighed, were analyzed for nitrogen and fat (ether-extract).
The dogs were fed upon a mixed diet of fresh lean beef, cracker dust,
lard, and water. The meat was prepared as described by Chittenden
and Gies,* and usually contained about 3.5 per cent N. The cracker
dust and lard were commercial brands, the former usually containing
1.6 per cent N, the latter 0.07 per cent N. The nitrogen content of
all the food and the excreta was carefully determined by the Kjeldahl-
Gunning process. In the urine total SO, was determined gravimet-
rically by periods in composite samples; P,O, and Cl were estimated
for each day, the former by titration with uranium nitrate, and the
latter by the Volhard-Salkowski method.

First experiment. — The animal used was a dog of twelve kilos. It
failed to attain perfect nitrogenous equilibrium; but the ‘experiment
may serve as a preliminary illustration. During the experiment,
which consisted of three periods of eight days each, the dog was fed
daily :

250 grams meat,

40 FS" lands
50 “ cracker dust,
300) Se) “water:

During the second period one-half gram of czsium chloride (about
forty-two milligrams per kilo body weight) was added to the daily

1 TSCHIRWINSKY: Centralblatt fiir Physiologie, 1895, ix, p. 49.

2 COHNSTEIN: Archiv fiir die gesammte Physiologie, 1895, lix, p. 509.
3 MENDEL: Journal of physiology, 1896, xix, p. 227.

4 CHITTENDEN and GieEs: This journal, 1898, i, p. 5.

224 G. A. Hanford.

diet. There was a deficiency in the nitrogen output as compared
with the intake. The weight of the animal remained the same
practically throughout the experiment, and no unusual symptoms
were noted. There was also no material difference in the amounts
of SO;, P,O;, and Cl excreted during the three periods. .aie fat
content of the feeces was not altered. A summary of the results
obtained is given below.

EXPERIMENT I. SUMMARY, AVERAGE PER DAY.

ToraL NITROGEN. URINE.

Periods,
8 days. ree ne
gested. | creted.

Balance.

grams. | grams. grams.

Fore | 9.689 | 9.475 | +0.214
Cesium | 9.670
After 9.650

Second experiment. — After two weeks, the dog of the first experi-
ment was used again, receiving the same diet as in the former trial.
A fore period, lasting seven days, preceded one of five days, in which
three grams of caesium chloride (two hundred and fifty milligrams per
kilo body weight) were given daily. At the end of this time the ani-
mal appeared languid, but otherwise no unusual symptoms were noted.
Then five grams were given, which caused the animal to vomit and
defzecate profusely. The dog appeared to suffer from intense gastro-
intestinal irritation, scarcely rising from the floor of the cage. Blood
was found in small amounts in the fzecal matter. From this time the
animal received no more cesium, and was allowed to recover. For
two days, food was rejected and the same symptoms of weakness, dul-
ness, and unsteadiness in the legs, which were noted in the injection
experiments, were shown. The animal now began to take food, and
gradually returned to a normal condition. No greater difference
in the amounts of nitrogen, SO,, P,O,, and Cl excreted were noted
than in the first experiment. The data are given in the following
tables :

Phystological Action, etc., of Caestum Chloride. 225

EXPERIMENT II. FORE PERIOD.

| Re-
Spec. action} Nitro- Total
grav. | to ‘ SOE: Oral:
litmus. |

grams.

1.022 | acid | 9.404) 1.613

5 | 1.026 118 1391
1.030 | 1.408
1.024 | "9. | 1.333 $ 99
1.020 | press 1.448
5 | 1.023 | | 8. 356 | 1.316
1.027 1.189

N of urine . 61.148
N of faeces. 5.049

Motals= =<

Av. per day

226 G. A. Hanford.

EXPERIMENT II. CASIUM PERIOD.

URINE.

Re-
.|action| Nitro-| Total Total

COM PEN SOx. Cl.
litmus.

| grams. | grams. grams. | gms.

1.027| acid | 10.6031 1.813)
1.025} “ | 6966 582.| 1.861

1.027 | | 9.598 $ 9.555 2.053 ¢ 71

1.024 | 8.298 1.986

1.021 10.571) 1.861 J

a]

Niseucne.. 46036)

INioffzeces’ = 3:62]

Totals =

Av. per day

EXPERIMENT II. SUMMARY, AVERAGE PER DAY.

TOTAL NITROGEN. URINE.

Periods.
In- Ex-

gested. | creted. Balance. SOs.

grams. | grams. gram. .c. 5 grams.

Fore 9.87 O431o Wea O45 sis) 2.223

Cesium whykey || SME Bs || =O bilyAll ‘ 207 leo

Phystological Action, etc., of Cesium Chloride. 227

Third experiment. —In this experiment a bitch weighing six kilos
was used. There were two periods, —a fore period of five days, and
a caesium period of three days. Approximate equilibrium was main-

tained on a diet of:
125 grams meat,
20 > land:
200 =) cracker dust.
100 “ water.

During the latter period a total of six grams was given, beginning with
one gram on the first day, and increasing a gram each day. This
amount caused no disturbance; but after the administration of four
grams on the fourth day, there appeared gastro-intestinal disturb-
ances so marked that further quantitative estimations of the urine
were discontinued. The symptoms were similar to those shown by the
preceding dog. No further excretion of the salt could be detected two
weeks later. The nitrogen balance during the caesium period was,
if anything, slightly greater than during the fore period. There was
no marked difference in the excretion of SO,, P,O,;, and Cl, except
for the small quantities of the latter which were introduced with the
caesium.
EXPERIMENT III. - FORE- PERIOD.

Foon. URINE.

Re-
Nitro- } . | action| Nitro-
gen. rav. | to gen.
\litmus.

grams. aCe grams.

4.856 .028 | acid | 4.826 )

4.856 Bile 0! sn Shr
4.856 : 4.886 | 4.776
4.856 é : 4.740

4.786 a 5.034 J

N of urine .

N of faeces .

228 G. A. Hanford.

EXPERIMENT III. CASIUM PERIOD.

Foop. URINE.

Re-
| action | Nitro- | Total | Total | Total
to gen. | SOs. | P,O;.) GLE

jitmus.,

| grams. | grams. grams.

grams.
1.025 | acid | 4.748) .... | 0:802)) 1.439

1.028} « | 49321 3.165 |0.875 | 1.719} 8.0 | 0.401
1.031] « | a ...+ [0.898 | 1.765

N of-urine . 14.575 |

N of feces . 0.401 |

| | !
‘Totals:..- | xe an a | 2.575 | 4.932

Av. per day | “ee =e ne 992 0.858 | 1.641

EXPERIMENT III. SUMMARY, AVERAGE PER DAY.

| ToraL NITROGEN. F ACES.
|
|

Periods. |
In- Ex- ; ‘ Nitro-
gested. creted. Ce se ‘ 50s | EO gen.

grams. grams. | AOS f Pat grams. gram.

Fore | 4.842 4.986 1.237 | 013860

Cesium | 4.786 4.992, ; 1.055 | 0.858 | 1.641 | 0.134

Fourth experiment. — About one month after the last experiment
the same bitch was again brought into approximate equilibrium dur-
ing a fore period of six days. The daily diet during this experiment
consisted of:

135 grams meat,

200 sa wlards

40 “ cracker dust,
100 <“ water.

During the second period, one gram of the czsium salt (about two
hundred and ten milligrams per kilo body weight) was given daily for
five days. The first four days brought out no unusual symptoms, the

Phystological Action, etc., of Cesium Chloride. 229

animal appearing well. On the fifth day, the animal seemed less well
and had less appetite. The next morning the further feeding of the
salt was abandoned, as the animal showed marked symptoms of its
toxic action. Food was taken only after coaxing, and then most of it
was vomited again. The feces were watery and light in color. The
animal was scarcely able to stand; the same evidence of loss of con-
trol being noted as has already been described, and the animal seem-
ingly having lost the power of co-ordination for its hind legs. It was
unable to hold the head erect. These symptoms continued to grow
more marked until the next night, when death ensued, the animal
being completely paralyzed. A few hours before death its reflexes
were tested. There was noted the same progressive loss already
mentioned in the injection experiments.

Until the marked toxic symptoms appeared, there was no difference
in the metabolism of the two periods, as reference to the tables will
show. The slight differences in nitrogen, SO;, P,O,;,and Cl are of no
moment. The animal’s weight was not greatly lessened. The experi-
mental data are given in the table.

EXPERIMENT IV. FORE PERIOD.

|
Re-

action - Total | Total
to EeOp |e Cl
litmus.

grams. |

1.027 576 ) 0.791 |

1.027 Were 0.904
1.025 | 4. 0.892
1.031 | « 426 | ossi |
1.029 (amet 0.847 |

1.026) « | 5.967. 0.898 |

N of urine . 33.180

N of feces. 0.536

Totals. . .| 33.098

Av.perday| 5.516

230 G. A. Hanford.
EXPERIMENT IV. CA/SIUM PERIOD.

Foon. URINE.

Re-
.|action| Nitro- Total | Dry
to gen. pOR: |e Clan ewite
litmus.

grams. -C. grams. grams. | grams.

1:0; || 2: .027 | acid | 5.714 1.362)

en) }1.030] “ | 5.691 (1.381

10" | 275" OAT} 5.590 1.439 + 8.5
10 | 202 |1.030] « | 4.843 1.458

|
1.0 | 1.031 | 4.880) 1.132)

f urine . 26.718

N of feces. 0.440 |

27.158 | 6.566

HAS2) 313

EXPERIMENT IV. SUMMARY, AVERAGE PER DAY.

FACES.

ToTAL NITROGEN.

Periods. | | eee j
; In- | Ex Balance. | Vol. | Nitro- SO3., |) Pe@rniemee Nitro-
gested.| creted. gen. eae gen.
: grams. | rans | “era i e.c. ae Dararnil| grams. "grams, gram.
Fore 5.516 5.619 —0.103 231 5030 27 869 | 1.272 | 0.089
-Czesium. | 5.647 | 5.432 +0.215 197 | 5.343 | 1.313°| 827°) 1.354 eee

lifth experiment. — For this a bitch of seven kilos was used. The
experiment lasted twenty-four days, consisting of a fore and after
period of six- days each, and a twelve-day cesium period between
these. Almost exact equilibrium was maintained during the fore
period by feeding:

175 grams meat,

20) 3" wilard®

40 “ cracker dust,
150 “ water,

Phystological Action, etc. of Cestum Chloride. 231

On the first eight days of the czsium period one and four-tenths
grams (two hundred milligrams per kilo body weight) of the salt
were added to the food. For the next two days, one and three-
quarters grams were given, —an increase to two hundred and fifty mil-
ligrams per kilo body weight. On the last two days of the period two
hundred and seventy-five milligrams per kilo body weight were given
amounting to one and nine-tenths grams. As the experiment pro-
gressed, there was only a slight change in the nitrogen balance. No
toxic or unusual symptoms were noted, the only effect of the largest
dose being a tendency to produce diarrhoea. The SO, excreted during
the last two periods was slightly less than the output during the fore
period, but this difference is too small to be of serious significance.
The nitrogen and P,O, output were practically unchanged, as was
the Cl,— when the amount added to the diet in the caesium period is
taken into consideration. Caesium was still being excreted three
weeks after the last dose was fed. The experimental results are
given in the tables:

EXPERIMENT V. FORE PERIOD.

Foop.

| Re- |
Spec. |action| Nitro- 'Total) Dry | Nitro-
grav.| to gen. | ClP i we. eens extract:
litmus. | |

Day of exp.
Body weight.

| grams. | grams.| gms. | per cent.
§! g

1.029| acid | 6.770 0.921)
6.495 | 1.036

6.219 | 0.959
. 9.361
6.495 | 1.017

6.701 0.940

6.609) 0.978)

N of urine . 39.389

Nof faces. 1.613 |

Totals. . 46 -» | 41.002 | 9.361

Av. p. day at a3 | 6.834 | 1.560
|

|

232 G. A. Hanford.

EXPERIMENT V. CASIUM PERIOD.

Re-
action | Nitro-

to gen.
\litmus.

Day of exp.
Body weight.

|

| acid
1.038
1.026
| 1.026 | 6. | 1.151
1.027 : | 1.381

11.025] §« 7 1.228

£52 0| 2.881
| 1.032 |

|

1.030

| 1.028
5 | 1.040 |
1.028} “ | 6.609

E035) 7.182

N of urine . 79.424
N of faeces. 2.881

Totals . |$3.726 Bibbs ofa: 11.429 15.347 | 52.0) 2.881

Average | 6 ¢ 54 OA ghee tn eal nes 239| 0.952! 1.278| .. |0.240

per day |
!

Phystological Action, etc, of Cesium Chloride. 233

EXPERIMENT V. AFTER PERIOD.

Re-
action| Nitro- Total Nitro-| Ether

to gen. | SOs. |P,O;.| Cl. gen. | extract.
litmus.

Day of exp
Body weight.

grams. . | grams. | per cent.

1.035 6.955 jal | 1.047

1.032 | 6.627 012 | 0.894

1.033] “ | 6811}6. 944 | 1.009 F
1.030 | | 6651 892 | 1.056
1.027 | | 4.131) | 0.773 | 1.094.

N of urine . 31.175
N of feces. 1.218

- | 31-393

EXPERIMENT V. SUMMARY, AVERAGE PER DAY.

ToraL NITROGEN.

Periods.
In- Ex-

gested. | creted. Balance. | . P,O;.

grams. grams. gram. Ce E gram. | grams.

Fore 6.903 6.834 +0.069 0.955 | 0.975

Cesium | 6.977 6.862 +0:115 | ‘ 0.952 | 1.278

After 6.417 6.278 +0.139 | 0.929

General considerations on metabolism. — From the foregoing experi-
ments it seems demonstrated that caesium chloride in the doses given
— forty to two hundred and seventy-five milligrams per kilo body
weight — does not seriously affect proteid metabolism. When the
dose becomes larger, a disturbing action on the gastro-intestinal tract
results, as is evinced by vomiting and diarrhoea, with an occasional

234 G. A. Hanford.

bloody faecal discharge. A dose sufficient to produce even such
toxic effects has no noticeable action on proteid metabolism. The
nitrogen output was not altered; at least such changes as occurred
were too slight to be of moment. In the first experiment there was
a fairly large progressive change, but this is the only one showing
such. In Experiment II, a plus balance of 0.413 gram changed dur-
ing the caesium period to a deficit of 0.171 gram. In Experiment V,
the average daily nitrogen balance in grams for the three periods was
respectively :
+0.069 ard) LIS +0.139:

When larger amounts of the salt were given the balance was de-
stroyed, because the animals refused food or vomited when fed by
force. Under such experimental conditions, the intake of nitrogen
could not be accurately determined, and estimations could not be
made.

The results given agree in. general with those reported by Krum-
macher! for sodium chloride (injected subcutaneously). Straub’ also
reported that feeding comparable amounts of sodium chloride did not
alter the nitrogen output. Pugliese,? on the other hand, observed
that this salt depressed proteid metabolism, whereas potassium chlor-
ide increased the exchange of nitrogen. Larger amounts of the salts
were used than those fed in the present experiments.

The excretion of sulphur and phosphorus, as might be expected,
ran parallel to that of nitrogen. There was no difference of any con-
sequence in the various periods. No evidence of excessive phos-
phorus metabolism indicating marked nuclear or cellular decomposition
was obtained. In this respect, the experiments agree with the results
obtained by the observers just mentioned, after giving the chlorides
of the other alkali metals.

During the cesium period, the amount of chlorides eliminated in-
creased to an extent about equal to the amount of chlorine added to
the diet with the caesium. Thus, in Experiment IJ, the animal re-
ceived three grams of chlorine in excess of that in the food. All
except thirty centigrams of this was eliminated during the same
period. In the last experiment (V), a total of 3.9 grams of chlorine
was added as cesium chloride. During this period there was an
excess eliminated over the first period of 3.6 grams; the after period

' KRUMMACHER: Zeitschrift fiir Biologie, 1900, xl, p. 173.
2 STRAUB: Zeitschrift fiir Biologie, 1894, xxxvii, p. 545.
® PUGLIESE: Archives italiennes de biologie, 1896, xxv, p. 17.

Phystological Action, etc., of Castum Chloride. 235

contained an excess over the fore period of 0.2 gram chlorine. The
chloride apparently is not stored up to any extent, but is for the most
part rapidly eliminated. Czesium could often be detected spectro-
scopically for several days after its administration. It must be remem-
bered, however, that this test is a very delicate one.

In those experiments CLEP IV, V,) in which the daily portions of
urine were carefully separated by catheterization, the elimination of
water by the kidneys was practically constant throughout the feed-
ing. No noticeable disturbing effect was obtained even with the
largest doses recorded. Pugliese reported similar results after feed-
ing the chlorides of sodium and potassium.

The elimination of cesium, —Czsium is excreted by way of the
intestine as well as through the kidneys. Its elimination follows
quite rapidly, although traces could be demonstrated spectroscopi-
cally in the urine of some animals for many days after the last feed-
ing. The salt disappeared from the feces before its presence ceased
to be demonstrable in the urine. Whether this is attributable to the
greater ease with which minute traces can be detected in the urine,
or whether, as a matter of fact, the last traces are eliminated by way of
the kidneys is not certain. In those cases in which the organs and
tissues were carefully examined spectroscopically at a period when
caesium ceased to be demonstrable in the urine, no evidence of the
element was obtained. When various parts of the alimentary tract
were examined spectroscopically after subcutaneous or intravenous
injection of caesium chloride, casium was detected throughout the
entire length of the intestine. In the stomach, the reaction was
usually less pronounced. If we recall that cesium was shown above
to be excreted by the salivary glands under such circumstances, it
may be questioned whether the reaction obtained with the gastric
contents was not in part or entirely due to admixture of saliva.
Functional differences in the secretory processes along the digestive
tract have been shown to exist in the case of various other substances
similarly examined. Portions of hair from some of the animals which
had received doses of czsium for some time were examined spectro-
scopically with negative results.

A few experiments on the elimination of caesium were carried out
on man. The individuals took doses varying from fifty to three
hundred and seventy-five milligrams dissolved in one hundred cubic
centimetres of water in each case. Only with the largest dose was
ceesium detected in the urine within the period of observation. In

236 G. A. Hanford.

this instance the spectroscopic reaction was evident at the end of an
hour. The faeces were not examined. The observations of Hiifner?
on the excretion of lithium in man are of interest in this connection.
The elimination of czsium through the intestine is interesting in
view of the fact that relatively few elements (Ca, Mn, Fe, Li) have
been shown to be excreted by that channel. Its presence in the saliva
recalls the work of Bernard,? Langley and Fletcher,? and Eckhard,
with potassium iodide, lithium citrate, and other salts. These inves-
tigators have made it clear that not all salts introduced into the cir-
culation reappear in the saliva. Certain of these, such as the
ferrocyanides, fail in this respect. For lithium Good® has recently
found in experiments on cats that when large quantities (one to two
grams) were injected hypodermically, very considerable amounts
were obtained from the stomach (by lavage), and from the bowel;
and the saliva also contained appreciable quantities. In fatal poison-
ing, more was found in the stomach and bowel contents than in the
urine. In experiments in which small doses were administered re-
peatedly, more lithium was excreted by the urine than by the ali-
mentary tract.

SUMMARY.

Experiments on frogs.—1. No marked differences between the
effect of % solutions of sodium chloride and cesium chloride upon
cilia and red blood-corpuscles were noted.

2. Isolated muscles contract spontaneously in such solutions. The
contractions cease sooner in the cesium than in the sodium solution.

3. Muscles exposed to cesium chloride lose their irritability toward
electrical stimulation sooner than those exposed to sodium chloride.

4. The irritability of nerves disappears sooner in caesium chloride
than in sodium chloride solutions.

5. Complete paralysis is produced by considerable doses of caesium
chloride (one to two and one-half milligrams per gram body weight).

' HUFNER: Zeitschrift fiir physiologische Chemie, 1880, iv, p. 378.

* BERNARD: Lecons de physiologie expérimentale, 1856, ii.

° LANGLEY and FLETCHER: Philosophical transactions of the Royal Society
of London, 1889, clxxx, p. 149.

* ECKHARD: Beitrage zur Physiologie, Leipzig, 1887, p. 13. Quoted in
Schafer’s Text-book of physiology, i, p. 504.

® Goop: This journal, 1902, vi, p. xx; American journal of the medical sci-
ences, 1903, Cxxv, p. 273. All of these observations on lithium are of interest in
comparison with those noted in this paper.

Phystological Action, etc. of Cesium Chloride. 237

6. No initial excitability was noted.

7. The cesium salt seems to act on both nerves and muscles.

Experiments on mammals.— 8. In the rabbit, cat, and dog, final
paralysis followed the subcutaneous injection of caesium chloride.

9. Two grams per kilo body weight (subcutaneously ) proved fatal to
a rabbit. One-half to one gram per kilo body weight usually proved
fatal to a cat or dog.

10. The symptoms elicited were those of intense gastro-intestinal
disturbance, vomiting, diarrhoea, loss of reflexes, and progressive
paralysis.

11. Intravenous injection produced an initial fall, followed by a
marked rise in blood-pressure in both cats and dogs. The lymph
flow from the thoracic duct was slightly accelerated.

12. Fatal doses (about three-quarters of a gram per kilo body

weight) caused death by cardiac failure.
- 13. Proteid metabolism in dogs was not noticeably disturbed by
doses of from forty to two hundred and seventy-five milligrams per
kilo body weight fed for several days. Sulphur and phosphorus met-
abolism were also unaffected. Chlorine elimination was unchanged,
except that the added amount fed with the cesium was rapidly
excreted.

14. No diuretic action was obtained.

15. When larger amounts were given per os, marked gastro-intes-
tinal disturbances resulted.

16. Elimination of the caesium by the intestine and kidneys was
comparatively rapid, no prolonged retention being noted.

17. In man, the cesium was detected in the urine within an hour
after ingestion of three hundred and seventy-five milligrams of caesium
chloride.

I desire to express my obligation to Professor Chittenden for sug-
gesting this research, and to Professor Lafayette B. Mendel for help
and criticism.

SOME FACTS CONCERNING GEOTROPIC GATHERS
INGS OF PARAMECIA.

By ANNE MOORE.

INTRODUCTION.

Te experimenting with Paramecia it is often found necessary as a
preliminary measure to wash them in distilled water. To ac-
complish this it has been customary to take advantage of the fact
that if they are removed from the culture to a test tube of distilled
water they will, after swimming about actively, come to rest at the
top of the tube. In following this method, I noticed some peculiarities
in their behavior which seemed to me worth recording and worth
further experimentation. The gathering at the top was apparently
an expression of negative geotropism.! I wished to ascertain whether,
having once gathered at the top, the Paramecia would change their
position to the bottom of the tube, or, in other words, whether the
sense of geotropism ever changes normally from negative to positive,
and, if so, to ascertain what factors influence the change.

The experiments were begun in the Physiological Laboratory of
the University of Chicago, and were afterward continued in San
Diego, California.

OBSERVATIONS AT RooM TEMPERATURE (18°-20° C.),

Small glass tubes half a centimetre in diameter and twelve centi-
metres long were used. These were filled to within two centimetres
of the top with distilled water, and then completely filled with water
from the culture containing Paramecia. The Paramecia were at first
carried downward with the stream of water, but soon began to swim
about independently. Although their orientation was not uniform,
there was a more or less steady movement downward. When the
bottom was reached, which occurred usually in about fifteen minutes,

' Although the gathering at the top seemed not always to be the direct effect
of an orientation, the word geotropism has been used for convenience, as indicative
of the general character of the phenomena.

238

Facts Concerning Geotropi Gatherings of Paramecia. 239

they remained massed there a few moments, and then formed about
the wall of the tube a well defined ring. This ring then began to
rise slowly, the individuals composing it remaining massed close
against the glass and against each other, indicating a stereotropic
reaction. Occasionally the ring was thicker on the light side of the
tube than on the other side, but this was not invariably true. The
upward movement of the ring seemed independent of active swimming
and of definite orientation. The rapidity with which the ring rose
depended somewhat upon the temperature ; the lower the temperature
the more slowly rising occurred. Ordinarily the minimum time in
which it might be expected to reach the top was about two hours.
If the tube was left for several days after the top was reached, the
Paramecia became scattered throughout the tube and finally collected
at the bottom. It sometimes happened that the Paramecia, instead
of behaving with one accord in this way, became divided into two
groups, one of which remained at the bottom, while the other rose
to the top, as described above.

As the rate of rising varies with the temperature, changes in
temperature must be an important factor in determining the position
of the animals, but as the sense of geotropism is not the same in all
animals taken at the same time from the culture, and as it does not
remain constant when external conditions are apparently unchanged,
changes of temperature alone are insufficient to account for this
behavior.

THE EFFECT OF CHANGES OF JEMPERATURE.

In Chicago, the experiments were tried during the winter. Several
tubes were left for three hours in a thermostat at a temperature of
26°-28° C. The Paramecia collected at the top in dense clusters.
On being removed from the thermostat to room temperature, they
immediately streamed toward the bottom. They were returned to
the thermostat, and left over night. The next morning they were
found at the top, and, on being removed, again streamed downward.
The tubes were put outside the window at a temperature of 2° C.
At first the Paramecia scattered, but in ten minutes they were all at
the bottom of the tubes. One tube was returned to the thermostat.
Fifteen minutes later a ring had formed, and had begun to rise, while
in the tubes outside the window there were no signs of rising. If
left outside, or in the thermostat for as long as two days, the Para-
mecia became scattered.

240 Anne Moore.

These experiments were afterward repeated in San Diego, Cali-
fornia, and results were obtained in harmony with those just described.
The following experiment is typical. Two tubes were similarly pre-
pared. Tube A was placed in a thermostat at a temperature of
26°—28° C.; tube B was placed in a larger tube filled with water and
surrounded by a mixture of ice and salt, the temperature being kept
as nearly as possible at 1° C. In ten minutes the Paramecia in tube
B were massed at the bottom, and two hours later were still
massed there. The tube was then transferred to the thermostat.
The ring formed as usual and rose steadily. In an hour, -how-
ever, the Paramecia scattered throughout the tube. If scattering
occurred, as a rule no ring was again formed until conditions were
again changed. Consequently a quarter of an hour later the tube
was removed to room temperature, 18° C. They remained scattered
for a time, and then collected in two groups, one at the top, the other
at the bottom of the tube. They were then returned to the ther-
mostat, and an hour later were collected at the top. In tube Aa
ring had formed in ten minutes, and had begun to rise. In half an
hour some of the Paramecia were at the top, but many were scattered.
When returned to a temperature of 1° C., the process as described
above occurred. The scattering of the Paramecia after having been
made to gather at the top or bottom by an unusual temperature is an
interesting adjustment to environment, and suggests the observations
of Yasuda,! who found that protozoa have a remarkable power of
adapting themselves to changes in the osmotic pressure of the
medium surrounding them. It suggests also the observations of
Jennings,” who shows that Infusoria, after reacting to a given stimulus
one or more times, may, if the stimulus is not a harmful one, cease to
react though the stimulus is repeated without change. Without
attempting an explanation, he suggests that there is a difference
in the physiological condition of the organism before and after the
stimulus. The inversion of the sense of geotropism indicated in the
experiments described above is a further illustration of the principle
pointed out by Loeb,® that by appropriate stimulation the sense of an
animal’s response may be inverted. He showed that changes in
heliotropism may be correlated with changes in temperature. Increase

' YASUDA, ATSUSHI: Journal of the College of Science, Imperial University,
Tokyo, 1900, xiii, p. Ior.

2 JENNINGS, H. S.: This journal, 1902, viii, p. 23.

° Loe, J.: Archiv fiir die gesammte Physiologie, 1893, liv, p. 81.

Facts Concerning Geotropic Gatherings of Paramecia. 241

in temperature, for example, causes the Copepod Temora longicornis
to change its sense of heliotropism from positive to negative; while
decrease in temperature causes the change from negative to positive.

&
THE EFFrect OF WITHDRAWING WATER.

An analogy has been noted between the effect of a lowering of the
temperature and of the withdrawal of water. Loeb,! for example,
showed that the larve of Polygordius and certain Copepods can be
made positively heliotropic by increasing the concentration of the sea-
water or by decreasing the temperature, and can be made negative by
decreasing the concentration of the sea-water or by raising the temper-
ature. He therefore came to the conclusion that raising the concen-
tration of the salt solution has qualitatively and quantitatively the same
effect as lowering the temperature. Greeley” has recently called at-
tention to the fact that this is true, because the cell loses water when
the temperature is lowered, as well as when the concentration of the
surrounding medium is changed. I therefore tried the effect of weak
salt solutions upon the behavior of Paramecia. The results were,
however, not very satisfactory. When first put into #4 NaCl, or
into weaker solutions of sodium chloride, there was a general scatter-
ing as in water, but after an hour they gathered more thickly near
the top. After several days they showed a tendency to go to the
bottom, In 74 NaCl, they went to the bottom almost immediately,
but shortly after died. In a solution of calcium chloride isosmotic
with #4 NaCl, they went immediately to the bottom. In this solution
water is withdrawn very readily.

In accordance with a suggestion of Davenport, Miss Platt? ap-
proached the problem of geotropism (geotaxis) by means of varying
solutions, in order to test a conclusion reached by Schwarz that nega-
tive geotropism is due to “the direct influence of gravity on the
organism which incites movement in the opposite direction.’”’ Her
plan was to obtain the specific gravity of the organism “ by finding
the density of a solution in which the animals, either dead or paralyzed,
remained suspended without rising or falling,’ and then to observe
their movements in solutions of varying density. For if geotropism
were a weight effect, a negatively geotropic organism should become

LOEB: “Loe. cz.
2 GREELEY: This journal, 1go1, vi, p. 122.
3 Piatt, JuLIA B.: The American Naturalist, 1899, xxxiii, p. 31.

242 Anne Moore.

positively geotropic in solutions of greater specific gravity than its
own, supposing the animal to be normally heavier than water. She
found the specific gravity of Paramecia to be approximately 1.017, but
she could come to no conclusion from its behavior, for in Cambridge
she could find no Paramecia with decided geotropic tendencies.

From experiments on tadpoles, however, she concludes that the
direct action of gravity, as expressed in the weight of the organisms,
does not act as the incentive to negative geotropism, and she suggests
that it is possible, both in tadpoles and in infusoria, that gravity may
act through the internal organization of the animal, and that in this
case the density of the surrounding medium might be expected to
effect little if any change in the direction in which the organism
moves. And indeed I find this to be true. In solutions of gum
arabic of greater specific gravity than the Paramecia, the animals
remain at the top or become scattered. If, however, changes in
density of the surrounding medium are such as affect materially the
amount of water in the protoplasm, as the movements of the organism
are ultimately dependent upon the organization of the protoplasm, @
priori, one would expect such changes to affect such reactions as the
negative response to gravity. In salt solutions, therefore, in which
water is withdrawn, the internal constitution of their protoplasm is so
altered that the Paramecia go to the bottom.

INFLUENCE OF Foop.

The fact that Paramecia were found at the bottom of the tube
after remaining several days in #4 NaCl may be not entirely due to
the withdrawal of water. It was found that when Paramecia were
left in a tube of distilled water for some time, without the addition of
food, they finally collected at the bottom and remained there. If
water containing food was added, they were found next morning ina
ring at the top of the tube. This indicates that the sense of geo-
tropism differs in Paramecia when deprived of food and when food is
supplied in abundance. It suggests a case mentioned by Loeb.!
The young caterpillars of Porthesia chrysorrhcea are positively helio-
tropic when hungry, but lose the sense of heliotropism when fed.

* LoEB: Comparative physiology of the brain, 1go1, p. 195.

Facts Concerning Geotropic Gatherings of Paramecia. 243

Tue EFFrect oF MECHANICAL SHOCK.

It was mentioned above that when the tube was lifted from the
thermostat the Paramecia streamed downward. This was noted on
several occasions, and was probably not due simply to a change from
the temperature 28° to the temperature 20°. Whether the Paramecia
were in a ring, or whether they were scattered throughout the tube,
it often happened that merely lifting the tube seemed to disturb them
sufficiently to cause a downward movement. This suggested the idea
that a more violent shock would cause a more marked gathering at
the bottom of the tube. The tubes were therefore shaken by hand,
with the unfailing result that the Paramecia collected at the bottom
in a mass much more closely aggregated than was noted under any
other circumstances, unless perhaps where the temperature was low-
ered to. 1”,

This effect of a mechanical shock is by no means new. Loeb!
showed that shaking made certain Copepods temporarily positive,
and indeed agitation of the water has been one of the most successful
means of producing an inversion of the sense of an animal’s response.
Towle? showed that even so slight a shock as contact with the mouth
of a pipette was sufficient to render Cypridopsis, which is usually
negatively heliotropic, temporarily positive. In spite of this, however,
it seems well to emphasize here the effect of mechanical shock upon
the geotropism of Paramecia, as it adds to the list of evidences of the
importance of mechanical shock in the life history of body and other
cells which of late have been forthcoming.®? In a recent article by
Matthews and Whitcher,* the statement is made that the most probable
explanation of the action of mechanical shock on the egg-substance is
that it ‘‘ causes a partial gelation of the colloids of the egg-substance.
It produces the same effect on the protoplasm as cold, and the two
processes accordingly supplement each other. This conclusion is, we
believe, strengthened by Mrs. Andrews’s observations on living proto-

1 Logs, J.: Archiv fiir die gesammte Physiologie, 1893, liv, p. 96.

2 TOWLE, E. W.: This journal, 1900, iii, p. 345.

8 For example, in the work of Meltzer on the Influence of Agitation on Animal
Cells and Bacteria, Zeitschrift fiir Biologie, 1894, xxx, p. 3; of MATTHEWS on Arti-
ficial Parthenogenesis in the Egg of the Starfish, This journal, rgot, vi, p. 142; and
of FISCHER on Artificial Parthenogenesis in the Egg of Amphitrite, This journal,
1902, vii, p. 301.

* MATTHEWS and WHITCHER: This journal, 1903, viii, p. 300.

244 Anne Moore.

plasm. She observed that a mechanical shock caused a distinct
change in the viscidity of the choano-flagellates, a very small jar
causing the collar to become rigid.” In her article’ Mrs. Andrews
shows that the “viscosity of protoplasm varies locally very swiftly.
From a very fluid state, it becomes rapidly so viscous as to resist
pressure, after the manner of a stiff bristle.’ It would be strange
indeed if such changes in the viscidity of the protoplasm were not
accompanied by changes in the reactions to external stimuli of such
animals as Paramecia.

SUMMARY.

I. Paramecia are sometimes positively geotropic, sometimes nega-
tively geotropic, sometimes without the sense of geotropism.

2. Positive geotropism may be induced in Paramecia by

(a) Mechanical shock. The effect of the shock takes place
quickly and is lost quickly.

(6) Reduction of temperature. At 1° C. the effect is marked,
and takes place quickly. In a comparatively short time, the animals
may adapt themselves to the low temperature and lose the sense of
geotropism.

(c) Increase in concentration of the surrounding medium. This
factor is not so constant as the other factors, and the effect not
so marked.

(2) Lack of food. The effect takes place slowly and is lost
slowly.

3. Negative geotropism may be induced in Paramecia conversely
by a plentiful food supply and by an increase in temperature within
limits.

4. The geotropic reactions of Paramecia to these influences is of
importance in their life history, for, as the positive reaction carries
them away from the surface, they would be protected from surface
agitations, from the effect of surface ice, and from the failure of
surface food-supply.

* ANDREWS: Journal of morphology (supplement), 1897, xii, p. 27.

SOME OBSERVATIONS ON THE. EFFECTS OF
AGITATION UPON-ARBACIA EGGS.

18 Sia dfs WINES ZANEA Re

[From the Marine Biological Laboratory, Woods Hole, Massachusetts. |

OR nearly twenty years I have studied at different times the
mechanical effect of shaking on unicellular organisms, While
spending my vacation last summer in Woods Hole and enjoying the
hospitality of the Marine Biological Laboratory, I was anxious to
study the effect of shaking upon the eggs of lower marine animals.
Loeb ! discovered that the simple increase in the osmotic pressure of
the sea-water is sufficient to start the process of segmentation in un-
fertilized Arbacia eggs. Mathews? has shown that in unfertilized
star-fish eggs this artificial parthenogenesis can be brought on by
the mechanical effect of squirting the eggs with a pipette. Similar
observations were made by Fischer® on the eggs of Amphitrite. In
Arbacia eggs no artificial parthenogenesis could be started by the
simple mechanical procedures employed by Mathews and Fischer.*
In my studies of the effect of shaking upon bacteria? I found that
different organisms differed in their behavior towards a certain degree
of vibration, that a degree of shaking favorable to one organism
might be indifferent to a second, and destructive to a third. As
the very mild mechanical action employed by Mathews proved to
be indifferent to the unfertilized Arbacia eggs, the question arose
whether other or more energetic methods of shaking would not prove
to be effective in one or the other direction upon the unfertilized as
well as upon the fertilized eggs of the sea-urchin. The methods I
have employed in my previous investigations upon blood-cells, as
well as upon bacteria, were principally of two kinds. The organisms
were either subjected to the finer continuous vibrations which exist

1 Logs: This journal, 1900, iii, p. 135.
2 MATHEWS: This journal, tgo1, vi, p. 142.
3 FISCHER: This journal, 1902, vii, p. 301.
4 Logs: Archiv fiir Entwickelungsmechanik, 1901-02, xiii, p. 481.
> MELTZER : Zeitschrift fiir Biologie, 1894, xxx, p. 3.
245

246 S. J. Meltzer.

in the neighborhood of strong rhythmical shocks, for instance, in all
parts of a building where large engines are continuously at work;
or they were subjected to the direct effect of the shocks, produced
by energetically shaking bottles, containing the organisms, to and
fro by the hand or by machines. In this latter method the effect
proved to be incomparably greater when small, corpuscular firm
bodies were placed in the liquids within the bottles, — glass beads,
for instance.

In the present series I have omitted the method of shaking gently
by hand without the addition of other foreign bodies, since I found
that Mathews and Whitcher! had started work on these lines. I
confined myself to the study of the effects of more violent methods of
shaking. The bottles which were shaken by free hand contained,
besides the eggs, some glass beads, about one-tenth of the volume of
the sea-water, the latter filling about one-third of the bottle. The
shaking was carried on by not less than one hundred and eighty
movements in a minute, mostly, however, two hundred and forty times
and more in one minute. For other methods of shaking I have
availed myself of the steam-engine and the dynamo located on the
grounds of the United States Fish Commission. Bottles with or with-
out beads were attached either to the piston of the steam-engine,
which made only thirty-six to fifty excursions in a minute, or to the
piston of the dynamo, which made steadily three hundred and sixty
movements ina minute. Bottles containing Arbacia eggs were also
placed on the large iron box of the steam-engine, or they were
attached to a high pole connected with this box. In either case the
eggs were subjected to a nearly continuous vibration emanating from
the distant shock; the vibration was more intense on the pole than
on the box, as could easily be felt and seen. The strongest effect of
the direct shock was of course produced by the movements of the
piston of the dynamo. The next strong effect was produced by
shaking the bottles by the hand. The mildest effect of the direct
shock was produced in the bottles attached to the piston of the steam-
engine on account of the slowness of the movements of the latter.
The temperature in the places where the bottles were attached was
only slightly higher than that of the air.

I shall now report briefly some of the observations I thus gathered.

1. Among unfertilized Arbacia eggs, which had been subjected to
any of the methods of shaking (except of course when attached to

* MATHEWS and WHITCHER: This journal, 1903, viii, p. 301.

On the Effects of Agitation upon Arbacia Eggs. 247

the piston of the dynamo), there were always a small number seg-
mented, while segmented eggs were absent in most of the controls.
The segmented eggs were mostly in the two and four cell stages,
rarely in the eight-cell stage, and never in a more advanced stage.
The usual precautions were taken to avoid contamination with sperm ;
this nevertheless does not absolutely exclude the possibility of an
occasional contamination. However, it seems to me that the very
fact that only a few eggs segmented, and that the segmentation did
not advance beyond the very earliest stages, speaks against the possi-
bility that the segmentation was due to a real fertilization which
accidentally always could have occurred in the shaken specimen and
rarely in the control.

Moreover, since it has been established that in the eggs of the
star-fish and Amphitrite the entire series of segmentation can be
started and brought to a finish by mechanical effects, it is surely
nothing unusual to find that strong mechanical effects can induce
some degree of cleavage, even in a few of the Arbacia eggs, though
the latter are in general very little susceptible to artificial mechanical
influences.

2. Shaking by hand was carried on with fertilized as well as un-
fertilized eggs for half a minute, one minute, two, three, four, and
five minutes. The predominant feature of the effect of this mode of
violent shaking was a greater or less destruction of the eggs. One
interesting result came out very distinctly in every experiment: it is
the fact of the superior resistance of the fertilized over the unfer-
tilized eggs. The little bottles containing the fertilized or the un-
fertilized eggs were prepared exactly alike, and in every experiment
a bottle of each kind was kept in the same hand, and shaken simul-
taneously for the same length of time. When examined under the
microscope, the fertilized eggs shaken for five minutes looked, with
regard to the number of disorganized eggs, like the unfertilized eggs
shaken only for two minutes; or fertilized eggs shaken for three
minutes looked like unfertilized, shaken for one minute and less.
Furthermore, in the progress of the destruction of the unfertilized
eggs, parallel with the disappearance of the eggs, there is a steady
increase of dust-like débris. There are very few shapeless eggs or
masses of large fragments present in the destroyed unfertilized eggs.

On the other hand, in the violently shaken fertilized eggs, we see
plenty of disorganized eggs of all shapes, and coarse fragments of all
sizes and forms, and see but little fine, dust-like débris.

248 S. J. Meltzer.

Unfertilized eggs are apparently more brittle ; they offer less resist-
ance to a mechanical shock, and break down more readily into dust ;
while the fertilized eggs are more elastic and coherent. I wish, how-
ever, to state expressly that by using the terms brittle and elastic, I
do not mean to offer an hypothesis as to the real nature of the
change of structure; these terms are only other expressions for the
facts I have observed. What I would be inclined to insist upon, is
that the change is of a physical character. Possibly the assumption
of Mathews and Whitcher,! that the difference in the resistance is
due to a change in the viscosity of the protoplasm, is correct. There
are, however, many other possible interpretations of the facts which
Mathews and Whitcher and myself have observed. Thus we may,
for instance, assume that the physiological molecules of the eggs,
which are stirred through the action of the sperm, and get into
readiness to enter into a new grouping so as to facilitate segmenta-
tion, are more separated from one another. It is possible, therefore,
that after fertilization each physiological molecule is surrounded by
a greater sphere of an elastic fluid than before, and it is the elas-
ticity of these spheres which acts as a buffer and dampens the effect
of the mechanical shock.

We should not omit to add that Morgan? in his studies on the effect
of centrifugalization upon eggs, found that it is more destructive to
unfertilized than upon fertilized eggs.

In experiments on unfertilized eggs, it appeared further, that the
longer the eggs were removed from the ovaries and preserved in sea-
water, the less was their resistance toward mechanical shocks. How-
ever, no detailed studies were made on this point.

When unfertilized eggs were shaken for one minute, a small number
of eggs remained uninjured, and a few have shown cleavage. In
many experiments, the mass contained bodies smaller than eggs;
they looked like small blastulae, and some of them were swimming
rapidly. Whether they were real artificial blastulz, or were foreign
invaders,’ etc., I was unable to determine. When the unfertilized
eggs were shaken three to five minutes, there were neither unseg-
mented nor segmented eggs, nor any swimming bodies to be seen.

Fertilized eggs, when shaken for one minute, even very soon after
fertilization, appeared very little injured; there were hardly any

1 MATHEWS and WHITCHER: Loe. cit.
* MorGan: Archiv fiir Entwickelungsmechanik, 1902, xv, p. 238.
* See LoesB: Archiv fiir Entwickelungsmechanik, 1902, xiv, p. 288.

Ox the Effects of Agitation upon Arbacia Eggs. 249

fragments to be seen. But there were present quite a number
of segmented eggs in an advanced stage, and even a few distinct
blastula were seen. When shaken for five minutes, a good many
injured eggs and large fragments were present, but the majority of
the eggs were still intact. Among them there were many seg-
mented eggs in all stages of cleavage.

Fertilized eggs shaken for five minutes (by hand over two hundred
times in a minute and with the addition of glass beads) contained
blastulz resting on the bottom of the vessel and a fair number of
swimming gastrulz, while the control contained nothing of that kind.
The fact that by violent agitation of fertilized eggs the cycle of seg-
mentation could be so hastened as to reach the gastrula stage ina
few minutes, is so unusual that now, writing from notes and not
being in a position to verify it, I hesitate to set it down unreservedly.
When the eggs thus shaken are left alone,a good many more develop
into blastule, but few into swimming gastrule, and none into plutei.
All soon undergo decomposition.

Violent shaking of the eggs within the bottles attached to the piston
of the dynamo thoroughly destroyed fertilized and unfertilized eggs
alike in the shortest time.

Fertilized eggs which were kept in bottles on the box of the steam-
engine, or were attached to the pole, and thus were subjected toa
moderate but nearly continuous vibration, have been influenced in a
double way. Eggs which were kept there from forty to eighty minutes
have shown in all instances a marked advance in the segmentation
over the eggs of the control. When in the control the four-cell stage
predominated, with only a few of the eight-cell stage among them, the
eggs subjected to vibration contained at this time predominantly
eight or sixteen cell stages, or had progressed even farther. Further-
more, the segmentation of the eggs in the bottles attached to the pole
(more vibration) was greater than that on the box, although the
temperature on the pole was lower than on the box.

Keeping the eggs in the mentioned places for longer than two
hours begins to affect their further development unfavorably, and
the longer the eggs were subjected to the vibration, the more dis-
tinctly they remained behind the control. The stage of plutei, if it
was ever attained, was reached ten or twelve hours later, all being
small and mostly crippled. The shaken eggs, as a rule, had only few
swimming gastrulz, most of them were lying on the bottom. Even
these, however, were, without exception, far in advance of those

250 S. J. Meltzer.

eggs which were shaken for a longer or shorter time, and then kept
under the same conditions as the control. For instance, thirty-six
hours or two days after the beginning of an experiment, while the
control would be teeming with well-developed plutei, the eggs shaken
for twenty-four hours would show a few crippled plutei, some swim-
ming gastrule, a good many resting blastule and eggs in all stages
of division ; the eggs which were made to vibrate for only one hour,
and placed near the control, than which they were then more de-
veloped, would be already decomposed and emitting a foul odor.

In other words, if fertilized eggs were once subjected to artificial
vibration, they became decomposed early, if the vibration were not
continued.

The last series of experiments brings out the following instructive
point. Mild vibrations of comparatively short duration, though they
affect somewhat favorably the first steps of segmentation of fertilized
eggs, upset their normal equilibrium to such a degree that further
development is almost checked, unless vibration be continued, and
even then the development of the eggs ultimately lags far behind
that of the uninfluenced control eggs.

Our experiments have shown that any kind of shaking might start
the first steps of cleavage in a few of the unfertilized Arbacia eggs.
Our experiments with violent shaking by hand have shown that some
fertilized or unfertilized Arbacia eggs might by violent agitation be
hastened in a very short time into an advanced stage of cleavage.
Finally, our experiments with mild vibrations have shown that the
first steps in segmentation of fertilized eggs can be slightly hastened
by these vibrations, but the experiments have also shown that these
vibrations are incapable of bringing the development of the egg to
a normal finish, and that they upset the normal equilibrium of the
fertilized eggs.

Shaking or vibration, at least in the methods we employed, has
proved to be incapable of producing artificial parthenogenesis in
Arbacia eggs. But the experiments have shown that even these in-
adequate vibrations are capable of profoundly influencing the finer
mechanism, underlying the process of segmentation, and that they do
not simply injure the eggs in a coarse, traumatic manner.

In unfertilized Arbacia eggs, cleavage can be brought about by
change in osmotic pressure, but not by simple mechanical means.
In star-fish eggs, cleavage can be induced by very simple mechanical
shocks, but not by the change in osmotic pressure. Since the dis-

On the Effects of Agitation upon Arbacia Eggs. 251

covery by Loeb of artificial parthenogenesis by the addition of chemi-
cals to the sea-water, it has been often stated that the entrance of the
spermatozoon into the ovum causes changes which might be similar
in character to those which are caused by an alteration in the osmotic
pressure of the suspension fluid. But why not also assume that the
great motility of the sperm is in every case an adequate mechanical
shock appropriate to start the normal cycle of cleavage, just as we
see that a simple mechanical shock may be the means of setting
up an artificial segmentation of the eggs of the star-fish? The arti-
ficial mechanical shocks at our disposal are capable of affecting the
eggs profoundly, but for the purpose of setting up a nearly regular
cleavage they are adequate only for eggs of one or the other species.
Might it not be that the motility of the sperm is specific, and there-
fore adequate and effective in each special case? Perhaps a study of
the different rates, characters, etc., of the motions of the different
spermatozoa will throw some light upon this question.

ON THE EFFECTS OF SUBCUTANEOUS AN]J#Cri
OF THE EXTRACT OF THE SUPRAKEN ye
CAPSULE UPON THE BLOOD-VESSE@S
OF THE BABB S.A

By S. J. MELTZER anp, CLARA MELTZER?

[From the Rockefeller Institute for Medical Research. ]
EFFECTS UPON THE EARS OF NORMAL RABBITS.

N a previous communication? we recorded the observation that
after an intravenous injection of adrenalin, the vasoconstriction
in a rabbit’s ear (on the non-operated side) is usually followed by a
vasodilatation exceeding that which existed before the injection. A
search in the literature on the subject satisfied us that our experience
of the vasodilating after-effect is not an isolated one. Langley,? for
instance, records that the submaxillary gland appears flushed after
the blanching of the gland, caused by injection of suprarenal extract,
vanishes. Regarding the general blood-pressure, Lewandowsky *
states that in cats it frequently sinks below the original level, after
the usual primary rise.

We have interpreted our observation by the assumption that the
reduced dose of adrenalin within the blood stimulates preferably the
central vasodilating mechanisms. On the basis of this hypothesis we
started out to determine a small dose of adrenalin which by intrave-
nous injection might be capable of causing a primary dilatation of the
ear-vessels. As we have stated in the above-mentioned paper, we
met with full success in only very few instances. In most of these
experiments an injection of the small dose brought either no change
at all, or there was no change for a minute or two, and then a per-
ceptible dilatation followed, or finally there was a very brief primary
constriction, followed by a pronounced dilatation of the vessels.

Research Scholar of the Rockefeller Institute.

S. J. and CLARA MELTZER: This journal, 1903, ix, p. 147.
LANGLEY, J. N.: Journal of physiology, 1901-1902, xxviii, p. 237-
LEWANDOwSkyY : Archiv fiir Physiologie, 1899, p. 360.

252 -

1
9
3
4

Lofects of Suprarenal Capsule on Blood-Vessels. 253

In a new series of experiments we studied the effect of adrenalin
upon the blood-vessels in the ear of the normal rabbit, when intro-
duced subcutaneously. These studies brought to light more clearly
the vasodilating effect of the suprarenal extract. We shall report our
results briefly.

These studies are not as easy a task as might appear at first thought.
The blood-vessels of the rabbit’s ear, as is well known, show a con-
tinual change in their volume. But these changes are not as regular
as might appear from the statements of Schiff,! who was the first to
observe them. Indeed, Becke van der Callenfels? published tables
showing that there may be great irregularity in these changes. We
can confirm his statements. Sometimes, indeed, constrictions of long
duration are interrupted by short dilatations, as Schiff reported; but
at times the dilatations last long and the constrictions are short, or
both are short, or both long. The duration of each phase can vary
from a few seconds to many minutes. It is obvious, therefore, that
it is very difficult to say whether a dilatation of long duration which
appears after an injection of adrenalin, is due to the injection or is
one of the normal irregularities; or whether it is due to some of the
numerous accidental conditions which favor dilatation. Warmth, for
instance, favors dilatation. Thus dilatations prevailed in animals, the
ears of which were observed by lamplight. Simple holding of the
ears for the sake of better observation favors dilatation. Any motion
of the animal causes paling of the ear, which is then followed by a
long lasting flush. There is also an individual variability; in some
rabbits constrictions, in others dilatations, preponderate. From our
experience, it would appear that younger and smaller animals are apt
to show more frequently pale ears.

However, after considering all possible sources of error, the fact
appeared to be unmistakable that subcutaneous injection of adrenalin
often favors a primary dilatation of the blood-vessels of the rabbit’s
ears.

Investigators? who studied the effect of the suprarenal extract upon
the general blood-pressure, have often made the assertion that sub-

1 ScuiFF: Archiv fiir physiologische Heilkunde, 1854, xiii, p. 523.

2? BECKE VAN DER CALLENFELS: Zeitschrift fiir rationelle Medizin, 1855,
p. 157-

3 GoTTLieB: Archiv fiir experimentelle Pathologie, 1896, xxxvili, p. 99;
LEWANDOWSKY: Loc. c7¢.; BorutTTrau: Archiv fiir die gesammte Physiologie,
1899, Ixxviii, p. 97.

254 S. J. Meltzer and Clara Meltzer.

cutaneous injection induces no vasoconstriction. From our studies
upon the blood-vessels of the rabbit’s ear, we can positively state that
a subcutaneous injection of a sufficient dose of adrenalin will cause a
distinct constriction of the vessels. Ina young rabbit, of a weight
less than 1000 grams, an injection of I c.c. of pure, commercial ad-
renalin (1 : 1000) will cause within ten minutes a distinct blanching
of both ears. Usually the animal then becomes exceedingly pros-
trated. In an animal of 1500 grams and more, a subcutaneous in-
jection of about 1.5 to 2.0 c.c. of adrenalin will cause a gradual, slow
but distinct narrowing of the blood-vessels, especially noticeable in
the larger ones; the central artery may finally appear as a fine line,
sometimes unevenly constricted, moniliform, but it never disappears
entirely.

When 0.6 to 1.0 c.c. of pure adrenalin is injected subcutaneously
into a medium-sized rabbit, in most instances it may be observed that
after a few minutes the periods of vasodilatation become steadily
longer, and those of constriction shorter, until about ten to twenty
minutes after the injection, when the dilatation remains practically
stationary for a period of from ten to thirty minutes. The central
artery stands out clearly in its entire length, and many very fine
vessels become visible. As a rule, however, the small vessels dis-
appear again soon, especially after the injection of a large dose, while
the central artery remains dilated for some time. The dilatation of
the vessels in these cases is not as intense as after an excitement or a
struggle of the animal; it is more comparable to the dilatations seen
generally after section of the sympathetic, which as a rule are also
not very intense.

When a smaller dose is injected, the effect is sometimes restricted
to an increase in duration and a more frequent appearance of the
phases of dilatation, but then an additional injection of 0.5 c.c. or
more will bring out a permanent vasodilatation. In very rare cases
a short period of constriction seemed to precede the vasodilatation.
We should add that injections of saline or water do not favor vaso-
dilatation.

The following are abbreviated protocols of some of the experiments.

Experiment 85. March 24, 1903.—10 A.M. Small gray rabbit, about 1200
grams. Ears watched for about half an hour ina cool room. Ears very
pale, occasionally vessels fill faintly but constrict again quickly. Injected
12 minims of pure adrenalin. In a few minutes the ear-vessels fill up
very well and remain full, except for very few and short intervals of con-

Liffects of Suprarenal Capsule on Blood-Vessels. 255

striction, for about half an hour, when the vessels become narrower again ;
but even after forty minutes the ear is not as pale as before injection.
March 25. Injected 15 minims of saline, no effect.
March 26. Injected saline, no effect. Half an hour later injected 12

minims of adrenalin, — very definite dilatation though not as marked as
on the 24th.

Experiment 96. April 5.— Large black rabbit watched for half an hour in a
cold room. Blood-vessels vary, dilate and constrict; at first average
duration of constriction five to eight seconds; of dilatation ten to fif-
teen seconds. Later the constriction periods become longer.

Injected 25 minims of adrenalin. During the first ten minutes slight
variation, but the ear did not become really pale again, and the con-
striction lasted only a second or two. After this period, the vessels
remained stationary in dilatation for 15 minutes, then became narrower
again. ‘Thirty-five minutes later vessels still somewhat dilated.

The results of this series of experiments would seem to have an
unfavorable practical bearing upon the subcutaneous application of
adrenalin for hamostatic purposes. As we have seen above, only
such doses of adrenalin as are large enough to cause a distinct and
dangerous general muscular paralysis of the animal, can induce by
subcutaneous injection a constriction of the blood-vessels. Only such
doses, therefore, come into consideration in the employment of ad-
renalin as a therapeutic measure. On the other hand, medium doses
of this drug cause a vasodilatation; therefore, instead of contracting,
the bleeding vessel would become dilated, and the bleeding would be
increased.

The results of the following series of experiments will throw some
light upon a mechanism by which the suprarenal extract might indeed
be capable of controlling certain forms of hemorrhage.

EFFECT OF SUBCUTANEOUS INJECTION AFTER ELIMINATION OF THE
AURICULAR VASOMOTORS ON ONE SIDE.

In our previous paper in this Journal! referred to above, we have
studied the effect of intravenous injection of adrenalin upon the
ear-vessels of rabbits after section on one side of practically all nerves
carrying vasomotors for the ear. We found that the constriction on
the operated side lasted considerably longer than on the normal side.

1S. |. and CLARA MELrzien: Loc, C77.

256 S. J. Meltzer and Clara Meltzer.

In a new series of experiments we studied, under the same conditions,
the effect of subcutaneous injections. - The remarkable results which
we uniformly obtained are chiefly as follows: In every adult rabbit
in which, on one side, the sympathetic was cut, the ganglion removed,
the third cervical nerve and its connections cut, or in which ganglion
and cervical nerves together were eliminated, a medium dose of ad-
renalin, when injected subcutaneously, brought out in every case a
constriction of the vessels on the operated side, while on the non-
operated side the vessels became more or less distinctly dilated. The
constriction is a gradual one; it begins a few minutes after the in-
jection, progresses slowly, and sometimes does not attain its maxi-
mum for half an hour. The constriction is rarely so great as after an
intravenous injection; but it may last for many. hours, — in fact, the
vessels sometimes do not assume their original width before the fol-
lowing day. The behavior of the blood-vessels of the ear on the
unoperated side differs but little from that reported above, when
adrenalin is injected subcutaneously into normal rabbits. We shall
only add that, if all conditions are equal, the vessels of the normal ear
in an operated animal generally show, as it seemed to us, a some-
what lesser tendency toward frequent dilatations than those of a
normal animal. Hence the dilating effect of adrenalin was more
easily recognizable and appeared to be more marked in the operated
animals than in the normal ones. It was in these animals that we
first noticed the distinctly dilating effect of the subcutaneous injec-
tions. The constricting effect upon the ear-vessels of the operated
side is present in all cases, whether the dilatation of the vessels was
due to the section of the sympathetic or to that of the cervical nerves.
In fact, the constriction is marked even if, before the injection, there
was no greater dilatation than the usual width in normal animals;
which is indeed often the case a few days after the section of the
sympathetic. It appeared, however, that in most cases when the
cervical nerves alone were cut, even if the dilatation consequent upon
this section was considerable, the constriction after the injection of
adrenalin set in later and appeared less pronounced than in the cases
in which the sympathetic was cut.

In a few cases in which either the sympathetic alone or the cervical
nerves alone were cut, the injection brought out either a constriction
in the centre and a dilatation in the periphery of the ear, or vice
versa, 2.é@., a dilatation in the centre and a constriction in the
periphery, thus showing in one and the same ear the dilating effect

Liffects of Suprarenal Capsule on Blood-Vessels. 257

upon vessels with central innervation and the constricting effect upon
those which are deprived of a central control.

In small and young rabbits, and by the use of very large doses, a
subcutaneous injection brought out a constriction even on the un-
operated side. But even in these cases the constriction set in later
and was markedly less than on the operated side.

In many experiments a leg was very tightly constricted and ad-
renalin injected peripherally to the ligature. When, some time later,
the ligature was removed, both ears blanched within a very short
time, even after the use of medium doses. But even in these cases
the constriction in the unoperated ear set in later and was distinctly
less marked than in the operated ear. The cause of the greater
efficiency of the subcutaneous injections in these experiments with
ligation of an extremity, we shall have a better opportunity to discuss
in a future paper on the influence of adrenalin upon the pupil.

The following are a few abbreviated protocols of the last series of
experiments.

Experiment 53. Dee. 1, 1902. — Large brown rabbit anesthetized. Ganglion
pulled out and third cervical nerve cut on the right side.

Dec. 4. Right ear-vessels full ; left pale.

2.45 P.M. Injected 25 minims of adrenalin 1 : 1000 under the skin
of the back.

2.50 P.M. Right ear paler; left variable.

3-10. Right ear pale; left full and stationary.

Dec. 18. Right ear pretty well dilated; left pale. Injected 15 minims as
above. Left soon became full, and remained so for several minutes, then
gradually became paler. Right ear continued to pale gradually.

Dec. 23,6 p.m. Injected 10 minims. Ina few minutes vessels of left ear
dilated ; right ear pale.

g. P.M. both ears returned to their previous states.

Dec. 24. Right hind leg tied. Injected below ligature 20 minims. Liga-
ture removed after five minutes. Right ear immediately becomes com-
pletely pale ; left ear a little fuller than before injection.

Lixperiment 54. Dec. 1, 1902. Rabbit anesthetized. Left cervical nerve
cut.
Dec. 15. Left sympathetic and also right third cervical cut.
Dec. 16, 8.30 p.m. Injected 30 minims adrenalin 1 : 1000 under the skin
of the neck. In 6 minutes both ears pale, left entire ear, right-central artery
still visible. Animal thoroughly prostrated ; found dead next morning.

258 S. J. Meltzer and Clara Meltzer.

Experiment 59. Dec. 11, 1902. Large gray rabbit anzsthetized. Left third
cervical nerve and connections and left sympathetic cut.
Dec. 17, 8.20 p.M. Left ear very full; right ear pale.
8.27. Injected under skin of neck 25 minims adrenalin.
8.29. Left ear paler; right fuller. Redness of right continues, while
left continues to get paler.
8.45. Right ear still moderately full; left pale.

Experiment 62. Dec. 16, 1902. Large white rabbit anzesthetized. Right
ganglion removed and third cervical nerve and its connections cut.
Dec. 17. Vessels of right ear fully dilated ; left ear very pale. Injected
under skin of neck 20 minims adrenalin. In two minutes vessels of right
began to contract ; in 14 minutes right ear quite pale ; left somewhat full.

Dec. 23, 6 p.m. Hind leg tied with Esmarch bandage, injected distally 30
minims ; toxic symptoms while bandage still on ; bandage removed, both
ears very pale.

g P.M. Left ear pale ; vessels of right dilated. Animal recovered, and

condition of ears as before injection.

Dec. 24. Injected subcutaneously 15 minims; vessels of left ear became
fuller ; right ear pale.

Jan. 14, 1903, 5.30 P.M. Right leg tied very firmly ; injected below liga-
ture 20 minims. No effect noted while ligature on. .

7 P.M. Ligature removed; right ear blanches immediately; left
blanches a little later.

Both our series of experiments establish beyond a doubt that the sub-
cutaneous injection of a medium dose of adrenalin causes a dilatation
of the vessels of the ears when they are under the control of central
vasomotors, but it induces a distinct constriction of the ear-vessels
when the central innervation is essentially eliminated. These medium
doses are, however, far in excess of those which give in intravenous
injections a maximum constriction. Above a certain minimum the
smaller the dose the more distinct the dilating effect of the subcuta-
neous injection upon the vessels of the normal ears.

These results will be better understood when we recall the hy-
pothesis which we offered in our last paper! in explanation of the
facts recorded there, which hypothesis was the starting point for
the present series of experiments. We have there put forward the
assumption that the suprarenal extract, when reaching the central
nervous system through the circulation, stimulates there the vaso-
constricting as well as the vasodilating mechanisms. But in conform-

1S. J. and CLARA MELTZER: Loc. cit.

Lifects of Suprarenal Capsule on Blood-Vessels. 259

ity with the mode of action of all stimuli upon the central vasomotor
mechanisms, stronger stimulation by adrenalin favors vasoconstric-
tion, and weaker stimulation favors vasodilatation. A large dose
of adrenalin represents a strong stimulus, and a small dose a weak
one. If a large (efficient) dose is introduced into the circulation it
at once favors constriction. When, however, later on, by oxidation
and elimination (and neutralization ?), the original dose becomes con-
siderably reduced, the reduced dose still stimulates both mechanisms,
but it now favors dilatation; hence the rapid disappearance of the
constriction and the appearance of a dilatation beyond the normal, as
an after-effect. When, however, the central influences are eliminated,
there is no longer a central dilating force present to overcome the
stimulation of the peripheral constricting mechanisms, for which the
small dose is a sufficient stimulus; hence the prolonged constriction
of the vessels in the ear deprived of its vasomotors.!. If we were
right in our assumption, there was then a possibility of finding a small
dose which would primarily cause dilatation. In the experiments
with intravenous injections, we were not very successful in finding
such doses. Our attention was then turned toward the method of
subcutaneous injections in the expectation that in this method only
very small doses at a time will be absorbed into the blood. We have
seen the positive result.

Our explanation of the results obtained in the present series of
experiments is then as follows: When even a comparatively large
dose of adrenalin is injected subcutaneously, only a very small frac-
tion of it is absorbed at a time into the blood; some of it is soon
eliminated and thus put out of service. The amount of the substance
which is actually present in the blood is very small. When the vaso-
motor nerves are intact, the small dose of adrenalin in the blood favors
vasodilatation. The result is that either the dilatation only exactly
neutralizes the (central and peripheral) constricting effect, and there
is consequently no change in the appearance of the vessels; or it
more or less overcompensates the constriction; the effect is then a
more or less distinct dilatation of the vessels. However, if the central
innervation is eliminated, the small dose within the blood stimulates
the peripheral constricting mechanisms without any antagonistic
influences of central origin ; hence the constriction of the blood-
vessels of the ear on the operated side.

1 For further particulars of this hypothesis we refer to the above-mentioned
communication.

260 S. J. Meltzer and Clara Meltzer.

We wish to add a few remarks. Our experiments show in the first
place that the suprarenal extract is very effective upon the blood-
vessels even when injected subcutaneously. It has been generally
assumed, as mentioned above, that subcutaneous injection of the
suprarenal extract exerts no influence on the blood-pressure, and this
was explained by the assumption that the extract of the suprarenal
capsule becomes oxidized in the lymph spaces before it can reach the
circulation.! Recently? some reports were published to the effect
that when very large fatal doses were used, the blood-pressure be-
came affected even by subcutaneous injection. This, however, might
be interpreted by assuming that the surplus of oxygen in the lymph
spaces was insufficient to destroy all the adrenalin. Our experiments,
however, show unmistakably that even comparatively small doses of
adrenalin are effective upon the state of the blood-vessels from the
subcutaneous tissues, the effect being either a dilating or a constrict-
ing one, depending upon the presence or absence of central vasomotor
influences. That the suprarenal extract does not become oxidized
by the tissues is best shown by the experiments with the injection
into the tightly ligated legs. Although the substance remained in
intimate contact with the tissues for hours, there appeared to be not
even the slightest impairment of the drug after the ligature was
removed.

The following consideration is of general interest. If we have the
constricting property of the extract in mind, our experiments have
taught us the fact that subcutaneous injection of rather large doses
has no effect on the ear-vessels when the central innervation is
normal, but it exerts a considerable influence upon the vessels as
soon as the central innervation is eliminated. Now our knowledge
of the effects of all drugs, alkaloids, toxins, or metabolic products, is
mostly derived from a study upon normal animals or organs. Are
the effects the same when the organs are deprived of their normal
innervation? As far as we know, this question has as yet hardly
been seriously raised. Our experiments have demonstrated that the
effect on pathological organs can be diametrically opposite to that on
the normal ones!

Finally we wish to recur again to the use of adrenalin as a hamo-
static. Above, we raised the point that if a permissible dose of

1 LEWANDOWSKY and BoruTtTau: Loc. cit.
2 AMBERG: Archives internationales de pharmacodynamie et de thérapie, 1902,
X1, Pp. 57.

Liffects of Suprarenal Capsule on Blood-Vessels. 261

adrenalin, when injected subcutaneously, dilates the vessels, the use
of adrenalin would obviously be somewhat harmful. Another objec-
tion has often been made of late against the use of such drugs in
hemorrhages as cause a contraction of all the blood-vessels in the
body, that the general rise of blood-pressure will favor the escape of
blood from the diseased or injured vessel. The harm would be greater
than the good done by a possible moderate local contraction of this
vessel. In our experiments, however, we have seen that a subcuta-
neous injection of a permissible dose of adrenalin causes a dilatation
of the vessels in normal parts, while vessels which are deprived of
normal innervation become constricted. Now if a hemorrhage occurs
in a diseased part, there is reason to assume that the vasomotor inner-
vation of these parts is already essentially injured. In this case a
subcutaneous injection of adrenalin could have an ideal hzemostatic
effect: it would lower the blood-pressure in the normal parts about
the lesion, which would divert the influx of blood from the bleeding
point, while it would cause at the same timea direct constriction of
the bleeding vessels.

SUMMARY.

Subcutaneous injection of suprarenal extract exerts a distinct effect
upon the blood-vessels of the ear. In normal animals a large dis-
tinctly poisonous dose causes blanching of the ears; a medium dose
causes a moderate but distinct dilatation of the blood-vessels. In
rabbits in which the vasomotor nerves were cut on one side, a sub-
cutaneous injection of a medium dose of adrenalin induced a distinct
constriction of the vessels on the operated side and a dilatation on
the unoperated side.

Apparently adrenalin is but very little oxidized in the subcutaneous
tissues.

DIFFERENCES OF POTENTIAL BETWEEN BLOOD
AND SERUM AND BETWEEN NORMAL
AND LAKED BLOOD.

BY G. NesSTEW ALE,
[From the Hull Physiological Laboratory of the University of Chicago.|

T is well known that, in general, the contact of solutions of dif-
ferent electrolytes, or of solutions of the same electrolyte of
different strengths, gives rise to electromotive force. This was first
explained by Nernst as due to the unequal velocity of diffusion of
the ions. With the view of throwing further light on the properties
of the envelopes of cells as regards their permeability for electrolytes,
I have endeavored to determine, by Poggendorff’s compensation
method, whether measurable differences of potential exist between
defibrinated blood, or sediments rich in corpuscles suspended in the
serum of the same blood, and the serum separated by centrifugaliza-
tion or obtained from the clotted blood; and between defibrinated
blood and blood laked in various ways. The idea in the first case
was to see whether the presence of the corpuscles, in virtue of the
unequal permeability of their envelopes for different ions, would cause
any electrical difference, and in the second case to see whether the
liberation of electrolytes and hamoglobin from the corpuscles, and
the alteration in the permeability of their envelopes would develop
a difference of potential. I had intended also to study the differences
of potential between standard solutions of electrolytes (2. g. 0.9 per
cent NaCl solution or serum) and suspensions of blood-corpuscles
in the same solutions, plus definite quantities of electrolytes known
to penetrate the corpuscles, and to compare those differences with
the differences of potential between the same liquids when the blood-
corpuscles are absent. These latter observations, however, I have
not yet been able to make.

Method. —'The two liquids occupied the two limbs of a U tube, into which
dipped a pair of unpolarizable electrodes (ZnSO, clay mixed with 0.9
per cent NaCl solution). The glass tubes of the electrodes were straight,
but tapered at the immersed ends so as to enter the U tube easily and to

262

Differences of Potential between Blood and Serum.

Combination.

in volts.

Remarks.

Blood, water
Blood, water

lood, water

Sediment of blood, se-
rum

Sapotoxin-laked blood,
control

Blood, 0.9 per cent NaCl
solution

Blood, 0.9 per cent NaCl
solution

Blood, sapotoxin-laked
blood

Blood, sapotoxin-laked
blood

Blood, sapotoxin-laked
blood
Mixed somewhat

Blood, heat-laked blood

Sapotoxin-laked blood,
serum

Defibrinated blood, se-
rum

Blood, sapotoxin blood

Blood, 0.9 per cent NaCl
solution

Water-laked blood, de-
fibrinated blood

Defib. blood, water
Defibrinated blood, sap-
otoxin blood

Mixed-up blood in U

tube
Putrid blood, fresh blood

0.0215
0.0057

0.0028
less than
0.0002
less than
0.0002

0.0183
0.0065

less than
0.0002
0.0036

less than
0.0001
less than
0.0001
less than
0.0001
less than
0.0001
0.0002

0.0003

less than

0.0001

Blood —, water-+. Laking only at interface.

Partially mixed up blood in one limb of
tube, the other remaining full of un-
altered blood.

Mixed up still more.

To 5 cc. of defibrinated blood, 0.9 c.c. of a
2 per cent sapotoxin solution in 0.9 per
cent NaCl solution was added, and then,
to make sure of maximum laking, a little
sapotoxin in substance. The “control”
consisted of blood to which as much 0.9
per cent NaCl solution was added as
was added of the sapotoxin solution to
the laked blood.

Blood—, NaCl solution+. The defibrin-
ated blood was drawn 48 hours before.
Same blood, but after standing 48 hours

more.

Same blood. Enough sapotoxin was added
in substance to cause complete laking.
0.0575 gm. sapotoxin in substance was
added to 20 cc. of defibrinated blood
drawn 24 hours previously. Laking is
complete in 2 minutes at room tempera-
ture. This observation of E. M. F. was
taken 31 minutes after mixture of the
blood and sapotoxin. Defib. blood +,

sapotoxin blood —.

Later on.

For the defibrinated blood, A (50) = 39.73.
For the sapotoxin blood, 4 (50) = 63.22.
For the serum, A (50) = 85.93. For serum
in 20 c.c. of which 0.068 grams sapotoxin
was dissolved, A (50) = 85.08.

For defibrinated blood, A (50) = 39.73. For
heat-laked blood, A (50) = 37.84.

Many Hb crystals in the sapotoxin blood.

Blood —, NaCl solution +.

Two volumes water were added to the blood.
Water-laked blood—, defib. blood+,
E. M. F. seemed to decrease after a few
minutes.

Defibrinated blood +, water —.

Sapotoxin blood+, defibrinated blood—.
For the sapotoxin blood, A (50) = 63.22.
For the defibrinated blood, A (50) = 41.09.

E. M. F. same as before.

The putrid blood was not completely laked.
Putrid blood +, fresh blood —.

263

264

G. NV. Stewart.

fit snugly. The clay plugs were flush with the narrow ends. Before
each observation, the difference of potential between the electrodes was
determined by immersing them in a U tube filled with 0.9 per cent NaCl
solution. Usually it was very small (less than 0.0002 volt), and unless
this was the case, the electrodes were set up again. When the same
electrodes were to be used for successive observations on more than one
combination, their tips were immersed in test-tubes filled with 0.9 per
cent NaC] solution. The U tube for the measurements and the test-tubes
for the washing of the electrodes were always slipped into position, while
the electrodes were separately clamped, and remained undisturbed through-
out the experiment. When scrupulous care is exercised in preparing the
electrodes at the start, good results can be obtained in this way without
the necessity of setting up fresh electrodes for each fresh combination,
results which, for the present purpose at least, can hardly be surpassed by
those yielded by the so-called normal mercury-calomel electrodes. The
resistance of the compensating wire (represented by a plug rheostat) was
never less than 10,000 ohms.

It will be observed that even after the action of such drastic
hemolytic agents as sapotoxin, in doses sufficient to produce marked
liberation of electrolytes from the corpuscles, the differences of poten-

tial
nite

between the laked and unlaked blood are very small. No defi-
differences could, in general, be made out between defibrinated

blood, or a blood sediment very rich in corpuscles, and the serum
separated from it.

tiie wero y. OF RENE,

By OTTO FOLIN.
[from the Chemical Laboratory of the McLean Hospital for the Insane, Waverley, Mass.]

THe TotTaALt ACIDITY OF URINE.

OTWITHSTANDING the large number of investigations

which have been published regarding the acidity of urine, no
method has yet been found for determining this factor with even
approximate accuracy. The reason for this condition is the fact that
no successful attempt has been made to overcome the two chief
difficulties which make direct titrations of urine unreliable: namely,
first, the occurrence of calcium in the presence of monobasic phos-
phates, and, secondly, the presence of ammonium salts.

The older methods of directly titrating the acidity with tenth
normal sodic hydrate in the presence of various indicators have there-
fore been given up as worthless by most investigators, and the more
recent methods have been directed toward the determination of the
monobasic phosphates on the assumption that to these is due prac-
tically all of the acidity of urine.

These later methods have, however, been shown to be inaccurate
for the determination of the phosphates, and are, moreover, certainly
open to the fatal objection that a considerable part of the acidity of
urine (sometimes more than one-half) is due to organic acids.

Two years ago Nageli' took up again the method of directly
titrating urine with tenth normal alkali, using phenolphtalein as
indicator, and attempted to prove that this gives fairly accurate
results. But it can easily be shown that the two most impor-
tant experiments with which Nageli supports his conclusion give, if
properly performed, results contradictory to those found by him.

The experiments referred to are the direct titration of monocal-
ciumphosphate, and the titration of any acid in the presence of
ammonium salts. That monocalciumphosphate cannot be titrated
with phenolphtalein and sodic hydrate under the conditions present

1 NAGELI: Zeitschrift fiir physiologische Chemie, 1900, xxx, p. 313.
265

266 Otto Folin.

in urine is seen at once when neutral calcium chloride is added to a
titrated monopotassium phosphate solution. The addition of the
calcium salt will in every case materially increase the apparent
acidity of the solution. Ndageli seems to have assumed that the
amount of calcium present in urine is too small to produce any
noticeable error in the titration of its acidity. In this he is mistaken.
The error due to the calcium of urine, as will be shown below, may
amount to more than ten per cent of the total acidity found.

Nageli’s experiments with ammonium salt solutions are even more
unsatisfactory. The apparent acidity of neutral ammonium salt solu-
tions in the presence of phenolphtalein is so pronounced that phe-
nolphtalein cannot be used as indicator in the presence of even
such limited amounts of ammonium salts as occur in normal urine.
Seventy milligrams of ammonium chloride, for example, containing
about 13 c.c. 77 NH, (an amount by no means rare in 25 c.c. concen-
trated urine) require in 100 c.c. of distilled water about 3 c.c. tenth
normal sodic hydrate to give a fairly distinct pink coloration.

Nageli has, therefore, by no means shown the old objections raised
against the direct titration methods to be invalid, and it is clear that,
in order to make such direct titration of the acidity of urine possible,
the difficulties mentioned must be overcome.

The disturbing effects of the calcium salts of urine can be com-
pletely removed by means of a reaction discovered by Liebig.
Liebig found that both di- and tricalcium phosphate can be “ dis-
solved” by means of potassium oxalate. That this reaction is
admirably suited to the purpose is seen from the following: A di-
sodicphosphate solution is, owing to dissociation distinctly alkaline to
phenolphtalein. If to such a pink colored solution is added calcium
chloride, the pink color disappears, calciumphosphate is precipitated,
and the solution is strongly acid. Adding a few crystals of neutral
potassium-oxalate and shaking rapidly restores the original pink
color of the solution, and the first flaky phosphate precipitate is
replaced by a finely granular one of calcium oxalate. This trans-
formation is quantitative, as can be shown by first adding potassium
oxalate, and then calcium chloride, to a titrated monopotassium
phosphate solution, — notwithstanding the presence of the calcium
it will now be found, on titrating, that the acidity of the phosphate
solution has not changed.

The principle of the above method for counteracting the tendency
of calcium to form basic phosphates, as has already been mentioned,

The Acidity of Urine. 267

was discovered by Liebig over twenty years ago. L. de Jager! in
his paper on the determination of calcium and magnesium in urine
has used sodium oxalate for the same purpose.

In addition to the calcium, there are present in urine certain smaller
amounts of magnesium salts which are also generally supposed to
interfere with the direct titration of phosphoric acid. This is, how-
ever, not the case. Pure magnesium salt solutions when added toa
disodicphosphate solution produce no precipitate, and do not change
the reaction of the solution.2, The addition of potassium oxalate to
urine before titrating the acidity is therefore amply sufficient to
eliminate any error that would otherwise arise from the formation
of earthy phosphates.

The presence of ammonium salts in urine, on the other hand,
seemed at first to offer almost insurmountable difficulties. If the
ammonium salt simply increased the acidity, one could find how
much this increase would amount to for a given quantity of ammonia,
and then after determining this constituent make a corresponding
correction in the acidity of the urine. This would indeed largely
correct the error. But another objectionable feature is the fact
that the ammonium salts give such a slowly increasing color with
phenolphtalein and tenth normal alkali, that the end-point of the
titration becomes exceedingly uncertain.

These difficulties due to ammonium salts cannot both be removed
quantitatively, to be sure, but they can be lessened to the point where
the error becomes negligible, by means of the same reagent which
was used above, — potassium oxalate.

A neutral ammonium chloride solution was prepared, 1 c.c. of
which contained 3.8 c.c. # NH;. 2 c.c. of this solution + 23 c.c.
distilled water + 1 drop } per cent phenolphtalein solution required
(at 24°) 1.2 c.c. 7% NaOH to produce a fairly pronounced pink
coloration.

The experiment was repeated as above, except that before titrating
rather more powdered potassium oxalate was added than could be
dissolved on shaking one minute. In this case 0.4 c.c. 7 NaOH
produced about the same depth of color as in the previous experiment.

The 11.4 c.c. 74 NH, contained in each of the above experiments

1 JAGER: Centralblatt fiir die medicinischen Wissenschaften, 1902, p. 641.

2 JAGER also overlooked the fact that phosphates can be titrated in the pres-
ence of magnesium salts without the production of basic magnesium phosphates ;
this renders his method for determining the magnesium in urine worthless.

268 Otto Folin.

represents fully as much ammonia as is found in 25 c.c. normal urine
having a specific gravity of not over 1.023.

5 c.c. of the above ammonium chloride solution (containing 19 c¢.c.
i> NH,) required in 2§ c.c. of water 2.7 c.c. 74 NaOH to produce
the same pink coloration as above.

After adding an excess of potassium oxalate and shaking for one
minute, 0.6 c.c. 7 NaOH sufficed to produce about the same end-
point in the titration.

Since 19 c.c. 7“ NH, represents more ammonia than, occurs in
25 c.c. of normal human urine, the result of the last experiment may
be said to represent the upper limit of the error that can arise from
the ammonium salts in normal urine on titrating the acidity of the
latter in the presence of an excess of potassium oxalate.

On titrating an acid solution, the error due to the ammonium salts
might be thought to be greater because of the dilution that occurs on
adding the alkali. As a matter of fact, this increase is very small in
the presence of the oxalate, and is moreover just about neutralized in
urine by the tendency of the dibasic phosphates formed during the
titration to dissociate and give a distinct pink coloration with phe-
nolphtalein a little before quite all the primary phosphate has been
converted into the secondary phosphate.

To illustrate:

20 c.c, standard monopotassium phosphate containing 100 mg.
P,O; with a theoretical acidity of 14.1 c.c. 7/9 required, after adding
5 c.c of the above ammonium chloride solution, and 15 gm. powdered
oxalate, 14.7 c.c. 7 NaOH to produce a distinct end reaction.

For determining the total acidity of normal human urine, I there-
fore propose the following simple method :

With a pipette transfer 25 c.c. of urine into a small Erlenmeyer
flask (capacity 200 c.c.). Add one, or at most, two drops one-half
per cent phenolphtalein solution, and 15 to 20 grams powdered potas-
sium oxalate. Shake about one minute, and titrate a¢ once with tenth
normal sodic hydrate until a faint yet distinct pink coloration is pro-
duced. The flask should be shaken during the titration, so as to
keep the solution as strong as possible in oxalate.

The following determinations taken from a recent metabolism
experiment may be cited in order to illustrate the values involved
on titrating the acidity of urine with and without the use of
potassium oxalate:

Lhe Acidity of Urine. 269

Total acidity in c.c. {5 in 25 c.c.

P,O; in | NHsg in c.c.
mg. in of 75 in

251 GE. DOECIC: Without
oxalate.

With only
enough oxalate
to prec. the Ca.

| With excess
of oxalate,

10.8
1Z9
104

It will be noted that the above directions for titrating the acidity
of urine have not been applied to pathological urines containing ex
cessive amounts of ammonia, such, for example, as are found during
and preceding diabetic coma.

My intention was to work out a somewhat different procedure for
such urines, but after having made a preliminary test on a urine con-
taining 38.2 c.c. 7 NH, in 25 c.c., I came to the conclusion that the
method described above gives with certain precautions results that
do not vary enough from the true values to make a separate method
necessary.

The test was made as follows: 25 c.c. urine shaken with about 20
gm. potassium oxalate was titrated as above, except that a second
Erlenmeyer flask containing 25 c.c. of urine and an excess of oxalate
was kept as a standard, and the titration was continued only until the
titrated sample showed a very faint darkening when compared with
the standard. The acidity thus titrated = 9.2 c.c. (4.

25 c.c. of the same urine were then transferred to a round-bottomed
litre flask. 50c.c. tenth normal sodic hydrate and 50 c.c. methyl
alcohol were then added, and the ammonia was distilled off at 50° C.
in a vacuum, according to the Boussingault-Shaffer method.! 3.82
c.c. 79 NH, was found on titrating the distillate. The remaining alka-
line liquid in the litre flask was cooled, and after adding 25 c.c. 44 HCl
and a little potassium oxalate, it was titrated in the presence of phe-

1 SHAFFER: This journal, 1903, viii, p. 345.

270 Otto Folin.

nolphtalein with tenth normal sodic hydrate. The titration required
222 CEG NAOT:

In this experiment the acidity of the urine is therefore 50 + 22.2 —
(38-25 25) = Sicies.

MINERAL ACIDITY OF URINE.

By the term mineral acidity is meant the excess of the combining
equivalence of all the mineral acids of the urine above that of the
known bases, sodium, potassium, calcium, magnesium, and ammonia.
The only available method at present for determining this excess is
to determine separately each of the mineral acids, and each of the
bases, and then to subtract the sum of the combining power of the
bases from the sum of the combining equivalence of the acids. It
need scarcely be mentioned that to determine all these constituents
with the necessary degree of accuracy is a task which few investiga-
tors have cared to undertake.!

The importance of making a large number of observations on nor-
mal and pathological urines in regard to their mineral acidity or
mineral alkalinity is, however, recognized.?

My own purpose in attempting to determine the excess of mineral
acidity or alkalinity, was to try to discover whether there is any ab-
normality in the metabolism of the insane, and whether the urines of
the insane show any indication of the occurrence of any unknown
organic bases or acids.

As a result of considerable experimentation, I have come to the
conclusion that the method which I am about to describe, and which
has been in use for some time in this laboratory gives the true excess

' After the writing of this paper was completed, I received the April number of
Beitrage zur chemischen Physiologie und Pathologie, which contains an article by
R. HOBER on the acidity of urine viewed from the standpoint of the ion-theory.
Such a theoretical discussion of the subject was purposely left out of the present
paper, but it will be found, I think, that the method here described is based upon
a correct application of the ion-theory as it affects the dissociation of acids, salts,
and indicators. H6per’s deduction that phenolphtalein must, of necessity, be the
most suitable indicator for titrating the acidity of urine, is only partly correct. He

forgets that phenolphtalein can ordinarily not be used in the presence of am-
monium salts.

* See BuNGE: Lehrbuch der physiologischen und pathologischen Chemie,
1898, 4th ed., p. 348; HERTER and WAKEMAN: New York University bulletin of
the medical sciences, 1901, i, p. 7; HERTER: Journal of experimental medicine,
1901, v, p. 617; FR. SOETBEER: .Zeitschrift fiir physiologische Chemie, 1902,
XXXV, p. 96.

The Acidity of Urine. 271

of mineral acids or bases in urine with as great accuracy as can be
obtained with the present state of knowledge concerning the com-
position of urine.

The principle of the method is simple and by no means new. It
has been used for determining mineral acids in vinegar and for deter-
mining the hydrochloric acid in stomach juice, but as far as I know
has never been applied to urine. The method is as follows:

0.3 to 0.6 gm.of pure, dry, granular potassium carbonate is accurately
weighed into a platinum dish, and 25 c.c. of urine are measured into
it. The resulting alkaline solution is evaporated on the sand bath or
electric oven to dryness, and when perfectly dry, the contents of the
dish are burned below red heat over a so-called radial burner, giving
a flame wide enough to heat the entire bottom of the platinum dish.
The burning should continue for about an hour after all visible ammo-
niacal fumes have ceased to come off. At the end of an hour the
flame is removed, and 10 c.c. of hydrogen peroxide water is added.
The dish is then covered with a watch glass, and gently warmed until
the peroxide is decomposed. The watch glass is then taken off,
and the sputterings rinsed into the dish by means of a little water.
The contents of the dish are again evaporated to perfect dryness, and
are then again heated over the radial burner as before for about an
hour.

The burning residue is now dissolved in water with the help of an
excess of tenth normal hydrochloric acid (75 or 100 c.c. 79 HCl, de-
pending on how much carbonate was taken), and is rinsed into
an Erlenmeyer flask, boiled to drive off the carbonic acid, and
cooled. The excess of acid is then titrated with tenth normal sodic
hydrate in the presence of a small amount of potassium oxalate
and two drops one-half per cent phenolphtalein solution.

Since the amounts of alkali and of acid added to the urine in the
above procedure are known, the final titration gives the data for
calculating the apparent excess of mineral acids or alkalies origi-
nally present in the urine. Before the final result is obtained, certain
other factors must, however, be taken into account.

The following factors must be determined: (1) thealkaline strength
of the potassium carbonate; (2) the acidity of the hydrogen perox-
ide; (3) the SO, content of the hydrogen peroxide; (4) the pre-
formed ammonia in the urine; (5) the inorganic SO, of the urine,
and, finally, (6) the total SO, found in the titrated solution of the
urine residue.

272 Otto Folin.

All these determinations are so easily executed as to require no
description, and since both potassium carbonate and hydrogen perox-
ide can be kept unchanged for months in well-stoppered glass bottles,
the first three determinations enumerated above need not be made
more than once (for any given sample of carbonate and peroxide).
The determination of the mineral acidity therefore requires : —

(1) The burning and titrating of the urine as described.

(2) One ammonia determination.

(3) Two sulphate determinations.

The calculation of the results is not complicated. The preformed
ammonia, the acidity of the hydrogen peroxide, and the acidity due to
the organic SO, of the urine, all in terms of tenth normal acid, must
be subtracted from the apparent excess of acidity found on titrating
the burned urine residue. The acidity (in c.c. of #4) of the organic
SO, is obtained by subtracting the sum of the SO, of the hydrogen
peroxide and the inorganic SO, of the urine from the total SO, of the
urine residue and dividing the amount so obtained in mg. by 8.

To illustrate:

25 c.c. of urine were burned with 0.5287 gm. potassium carbonate
(7.76 mg. of which contained 1 c.c. 7% alkali). The burned residue
was boiled with 75 c.c. + HCl, and the titration required 19 c.c. 7 NaOH.
An ammonia determination gave 5.2 c.c. 74, NH; in 25 c.c. of urine. The
total SO; = 59.9 mg.; the inorganic SO; = 42.8 mg. (10 c.c. of the
hydrogen peroxide used contained 8.8. mg. SO; and o.5 c.c. 7 acid).

0.5287 gm. K,CO; = 68.14 c.c. #4 NaOH ) _ Sy.1 Ce. 7 Namen

NaOH added =19 on Rye ayes Pe

HCl added ==). 5 0.

Apparent acidity of urine = 12.1 Cc. 1 HCl.

Ammonia in 25 C.c. Se BE? 1C.C. aa

Acidity of H,O, = 05 “ “ | —6.7cc.2;HCl.
Acidity of organic SO; = a92 a == Ue rae

Mineral acidity in 25 c.c. =. 5:4 6.6. ee

A few explanations of the above somewhat condensed directions for
determining the mineral acidity may be added.

1. The chief difficulty in burning urine residue completely free
from organic compounds containing ammonia is one that is frequently
overlooked, and the urine literature contains many instances in which
this oversight has undoubtedly led to erroneous results. This diffi-

The Acidity of Urine. 273

culty is due to the urea and to a smaller extent to the alloxur bodies.
- Although urea is both unstable and volatile at higher temperatures, it
is by no means an easy task to burn pure urea completely into car-
bonic acid and ammonia. . When heated even to a red heat in a plati-
num crucible, it is converted into white stable cyanic and cyanuric
acid derivatives which look like mineral matter. That this residue
is actually figured as mineral matter in determinations of mineral
residues of urine according to the methods given in textbooks is cer-
tain. In Neubauer and Vogel’s “ Analyse des Harns,” for example
(1oth edition, 1898, page 703), one is particularly warned against
heating too high, as otherwise abundant fumes of “alkali chlorides ”
are driven off. The alkali chlorides are, as a matter of fact, very
little volatile, and the fumes warned against are nothing more or
less than ammonium carbonate.!

On account of these difficulties encountered in attempting to drive
off urea quantitatively by means of heat, it is necessary to heat a
Iong time as is prescribed above, and to use a flame that will heat
the entire bottom of the platinum dish at once. If only a Bunsen
burner is used, the cyanogen compounds will melt and flow away
from the hot part of the dish, and thus escape decomposition. The
errors that arise from such incomplete decomposition of the nitrogen
compounds may be so large as to make the determination worthless,
nor can it be corrected by determining the remaining ammonia,
since the cyanuric acids also have a strong acid reaction. Where
there is any doubt as to whether the decomposition of nitrogenous
compounds has been complete, the final filtrate should, however, be
distilled with alkali and the ammonia determined, as this will indicate
approximately the amount of the error.

2. The burning process described above does not always yield a
perfectly white ash. The small amount of carbonaceous residue left
behind can of course be filtered off, but this is not necessary, as it is
never enough to obscure the end-point of the titration.

1 The above statement does not mean that more intense ignition would give
the correct amount of mineral residue. In my opinion it is, on the contrary, from
the nature of the case, impossible to isolate the mineral constituents from human
urine, or from the urine of carnivorous animals, by direct evaporation and ignition.

What has been said regarding the determination of the total mineral residue of
urine is also true of potassium determinations, and it looks to me very doubtful
whether more than a very small portion of the potassium determinations in urine
recorded in the literature are even approximately correct on account of the con-
tamination with ammonia.

274 Otto Folin.

3. Pure, dry, granular potassium carbonate seems the best alkali
to use when evaporating and burning the urine. It is much less
hygroscopic than the ordinary powdered sodium carbonate, and can
be kept for months in a small glass-stoppered bottle without chang-
ing its alkaline value. The alkaline hydrates of sodium and potas-
sium are not suitable; first, because if standard solutions are added it
takes much longer to evaporate to dryness, and, secondly, because the
hydrates are much more volatile than the carbonates.

The potassium carbonate can be kept very conveniently in a “ spe-
cific gravity” bottle. The amount of carbonate used for each deter-
mination should be, as is mentioned above, from 0.3 to 0.6 gm.
Larger amounts can of course be taken if it should seem necessary,
but numerous experiments have convinced me that nothing is gained
by adding larger amounts, and to do so entails an enormously in-
creased tendency to bumping, with consequent loss of alkali, during
the evaporation.

4. The hydrogen peroxide is added to oxidize the thiocyanates and
any small amounts of sulphide which may have been formed during
the burning. Before deciding on hydrogen peroxide, several other
oxidizing agents were tried, such as potassium chlorate, sodium
peroxide, and oxygen gas; but hydrogen peroxide was finally chosen
as being the most suitable. For the suggestion of using this reagent
I am indebted to H. Wislicenus.!_ Wislicenus used perfectly pure
hydrogen peroxide in his experiments, but this seems scarcely prac-
ticable on account of the difficulty of obtaining it. The commercial
product is good enough. The necessary corrections for the SO, and
the free sulphuric acid it contains are very easily made.

5. It may not be unnecessary to point out that in burning the
urine residue as directed above in the presence of potassium carbo-
nate, there is a certain danger that the gas may contain sulphur? In
my experiments this danger was avoided. The gas used is made
from gasolene,. and is free from any appreciable quantities of sulphur.
This was tested by heating potassium carbonate alone for two hours,
and then testing for sulphate. Since the flame used is very low there
may not be much danger, and the sulphate determinations would soon
indicate if any sulphur dioxide were absorbed from the gas ; but where
this is the case, an alcohol burner must be used.

‘ WISLICENUS: Zeitschrift fiir analytische Chemie, Igor, xl, p. 441.
2 See CARL TH. MORNER: Zeitschrift fiir physiologische Chemie, 1897, Xxill,
Desi

The Acidity of Urine. 275

6. A few words must still be added in explanation of the correction
used for the organic sulphur of the urine, including the ethereal and
the so-called neutral sulphur. I am well aware that it is arbitrary to
figure exactly one half of this sulphur as inorganic acid, and the
other half (of the chemical equivalence) as “ organically ” combined.

According to the present conceptions of the chemistry of the car-
bon compounds, all of the above sulphur equivalence is in fact ‘‘or-
ganic,” because of the partial combination with organic radicles.
Since sulphur is such a pronounced acid-forming element, and since
the acidity of such compounds as theethereal sulphates and sulphonic
acids is so clearly due to the SO; group of such compounds, it would
seem more logical in this connection to class the acidity derived from
them with the mineral acidity.

Having accepted this classification, it still remains arbitrary to
assume that the organic sulphur of urine actually is there as mono-
basic acids. In such compounds as the alkyl sulphides found in
dog’s urine, we know in fact that we are dealing with neutral and not
with acid compounds.

The term “ neutral sulphur,” as used at present, is however, in my
opinion, a misnomer, since we do not know that more than an exceed-
ingly minute fraction of the organic sulphur in human urine is there
in neutral form. Ethereal sulphates, chondroitin-sulphuric acid, and
the sulphocyanides we know, on the other hand, to be monobasic
acids, and it seems quite as probable that most of the other organic
sulphur compounds existing in urine will be found to have the sul-
phur with a free monovalent acid equivalence as to have it “ neutral.”’
Small traces of sulphur occurring in the mucoid and similar sub-
stances may, on the other hand, be spoken of as neutral.'_ The error
involved in assuming that the organic sulphur of the urine figures as
monobasic acids is, therefore, it seems to me, in all probability so
small as to be negligible.”

1 Urines containing appreciable quantities of albumen must be heated to boil-
ing after acidifying with pure acetic acid and the clear filtrate taken for the
mineral acidity determination. Very small quantities of albumen do not contain
enough sulphur to interfere to any noticeable extent with the mineral acidity
determination.

2 In order to avoid some criticism perhaps the “ organic phosphates ” should
also be mentioned. These have not been overlooked, but the determinations
recorded in the literature, as well as a number of determinations made in this
laboratory by Mr. Philip Shaffer, indicate that the organic phosphates are

present, if at all, in such minute quantities as to have no determinable effect on
the acidity of urine.

276 Otto Folin.

The following experiments are added in order to indicate how the
accuracy of the method was established.

I. 0.4560 gm. potassium carbonate was weighed into a large platinum crucible
weighing 39.8760 gm. 0.895 gm. perfectly pure urea was added, and the
crucible was then heated on the electric oven till no ammoniacal fumes
were visible. The crucible was then heated over a radial burner one
hour. After adding and evaporating 5 c.c. of water, the crucible was again
heated to just below redness for one hour.

At the end of this time the crucible and contents weighed 40.2797 gm.

The crucible was heated for another hour, and now weighed 40.2798 gm.
40.2797 — 39.8760 = 0.4037 gm. (the weight of the carbonate made abso-
lutely anhydrous). The theoretical weight of 0.4560 gm. of the potassium
carbonate (7.76 mg. of which = 1 c.c. 7%, HCl) was 0.4054 gm.

The heated carbonate was then dissolved in water and boiled with 75 c.c.
~~ HCl. After cooling, the titration required 16.5 c.c. {4 NaOH. ‘The
value of the carbonate remaining was, therefore, 75—16.5 c.c. = 58.5 c.c.
7; NaOH.

Theoretical = 58.76 c.c. 1, NaOH.

2. 0.3796 gm. potassium carbonate (corresponding to 48.93 c.c. 4, NaOH)
was weighed into a platinum dish. About 1 gm. of Kahlbaum’s pure
lactic acid (sp. gr. 1.21) and some water were added. ‘The mixture was
evaporated and burned two hours. In order to burn the charred residue
more completely, small quantities of distilled water were twice added
and evaporated, and the burning continued for fifteen minutes after each
evaporation.

The residue was then dissolved in water and boiled with 60 c.c 74 HCl.
The titration required 11.2 c.c. {4 NaOH.

The value of the carbonate recovered was therefore 60 — 11.2 = 48.8 c.c.
7, NaOH.

3. 25 c.c. urine burned with 0.6096 gm. potassium carbonate gave a mineral
acidity of 3.4 ¢.c. 7% in 25 C.c.

50 c.c. of the same urine burned with 0.5513 gm. potassium carbonate
gave a mineral acidity of 7.05 c.c. 7% im 50 c.c.

10

OrGANIc AcIDITY IN URINE.

Having described in the preceding pages how to titrate the total
acidity of urine, and how to determine the mineral acidity of the
same, but little need to be said in regard to the determination of the
organic acids of the urine.

By subtracting the mineral acidity from the total acidity is ob-
tained the organic acidity, or rather the total equivalence of organic

The Actdity of Urine. ar |

acid, whether free or combined. This deduction is true even for
pathological urines containing excessive amounts of combined or-
ganic acids such as are found in diabetic coma, only in these cases

SERIES No. l.

eres Specific Total Mineral Organic
: gravity. acidity. acidity. acidity.
1 1.025 12.9 4.7 8.2
iz 1.020 10.4 3.1 7.3
3 1.029 19.3 8.4 10.9
4 1.0305 22.0 4.9 7a
5 1.0245 12.6 a 8.2
SERIES No, 2.
1 1.0175 8.0 SZ 4.8
2 1.0185 10.0 343 4.6
3 1.018 9.0 See 5.8
4 1.027 15.0 8.5 6.5
5 1.027 16.4 US) 5.5
6 1.0175 7.0 2.0 5.0
SERIES No. 3.
1 1.021 9.8 —3.4 13:2
2 1.021 11.0 —4.3 5:3:
3 1.020 9.0 —3.8 12.8
4 1.022 9.0 0.4 8.6
5 1.0225 9.5 0 9.5
6 1.020 9.0 11.5 10.5

the mineral acidity is algebraically negative, 7. ¢. is replaced by
mineral alkalinity, and the original total acidity may be said to be due
wholly to organic acids. It is also true for urines which are directly

278 ; Otto Folin.

alkaline to phenolphtalein, as becomes evident by expressing the
original alkalinity in terms of algebraically negative acidity.

To illustrate: A mixture of sodic acetate and acetic acid would
have an original acidity corresponding to the amount of uncombined
acetic acid. The total amount of organic acid in such a mixture
would be obtained by making a mineral alkalinity determination
and adding the result so obtained to that obtained by the direct
titration.

It has been very generally assumed that the acidity of urine is
almost wholly due to the presence of monobasic phosphate, and that
the organic acids of normal urine play but a very subordinate rdle.
The extent to which the actual acid H ions of urine are derived from
phosphoric acid depends on the amounts of the organic acids present,
since these are in general much the weaker acids and therefore com-
bine to a correspondingly smaller extent with the bases present.

It is practically impossible to determine the exact point of equi-
librium in such a complex unknown mixture of acids as is contained in
urine, nor is this point of any very great importance. For practical
purposes it is, therefore, far nearer the truth to assume, as is done in
this paper, that the organic acids are free, and that the excess acidity
only is due to “acid phosphates.” If this is done, it will be found
that the acidity due to the phosphates varies within very wide limits,
but that the greater part of the acidity of urine is in most cases due
to organic acids. This is illustrated by the following acidity deter-
minations taken from some recent metabolism experiments. All
these acidity determinations were made in 25 c.c. of urine.

It will be noticed that the urines in the last series (No. 3) con-
tained no mineral acidity, but on the contrary a rather pronounced
mineral alkalinity. These urines were highly unusual in that 20
per cent or more of the total nitrogen was present as ammonia. Not-
withstanding this extraordinary high percentage of ammonia nitrogen,
the acidity determinations proved at once that oxybutyric or other
organic acids were not present to any extent that could be considered
pathological.

meohUDY OF THE REACTIONS AND- REACTION TIME
OF THE MEDUSA GONIONEMA MURBACHII
TO “PHOVICG. STIMULI

BysROBERT M. YERKES,
WITH THE ASSISTANCE OF JAMES B. AYER, Jr.

CONTENTS.
Page
Rroblems . . -. SR liek ed eee LOM e's, Eee I Cet co eee PATS
Reactions to light Sah saeisy Tepe ey merge ee aking: SPR Cale ame gl 200)
RMA ETT SDE S Ny a le Ach os os 0 coe alae Snel or te eae ren inis sen AZO
The directive influence oflight . . . 3 Cae eae ee LOL,
Effects of increase (light) and decrease (darletess) a aHOne seendianon ya kere LOO
The reaction time of Gonionema to photic stimuli . . ........ .. «288
Miethoday cas @ .. pA ; Soh! Epes en OS
The influence of the intensity of Genes on te time ieee 4 95 6 ese ALY
iRelatronrormeaction time toMight tO)Size aee-s fle tere eA oo
Relation of reaction time to light to pigmentation . . . . .... . 293
Relation of reaction time to light to sexual condition . . . . .. . . 294
The influence of changes in temperature on the time of reaction . . . . 294
Localization of sensitiveness of Gonionema to photic stimuli . . . . .. . . 299
MMMEMRCE TICE ONICIUISIONS 7. Pisin ky ih) a er! Genre! och eR BA Se San ee caer Pees wo MeR
PROBLEMS.

HE experimental study which is the subject of this paper? is a
continuation of work begun at Woods Hole three years ago.
Certain of the earlier results have already been published in this
journal.2 In my first paper, which dealt with the sensory reactions
of Gonionema, some of the organism’s reactions to light were briefly
described, and attention was called to their probable significance in
the life of the medusa.!

1 The experimental work of this paper was done at Woods Hole during the
summer of 1902. I desire to express my thanks for the assistance rendered me
by the authorities of the Marine Biological Laboratory.

2 This paper is one of a series of Comparative Reaction-Time Studies, of
which one dealing with the reaction time of the frog has already been published
in the Harvard Psychological Studies, 1903, i, pp. 579-638; Psychological Review
monograph, iv.

3 YERKES, R. M.: This journal, 1902, vi, pp. 434-439. /d7d., vii, pp. 181-198.

4 YERKES, R. M.: This journal, 1902, vi, pp. 445-447.

279

280 Robert M. Verkes.

In the present work attempts are made (i) to note and analyze
the various forms of reaction to photic stimuli exhibited, (2) to dis-
cover their relations to the life-history, habits, and instincts of the
organism, and (3) to study the time relations of the various reactions
in the hope of discovering something concerning the réle of the
nervous system. Although this paper is chiefly a discussion of the
influences of certain environmental factors on reaction time, the central
problems of the whole of the investigation are those of the functional
importance of the nervous system.

REACTIONS to LIGHT AND GRAVITY.

In its natural habitat, attached to marine plants or resting on the
bottom at a depth of several feet from the surface of the water, Gon-
ionema is exposed to intensities of light which are usually less than
that of daylight. When it is in a dense mass of eel grass, as is
usually the case at Woods Hole, the light is very weak, but when it
happens to come to the surface during the day it is exposed to light
which may be of the intensity of direct sunlight. When the weeds
to which Gonionemata are attached are disturbed the animals are
either torn from their attachments, or are stimulated by the disturb-
ance of the water to release their holds, and immediately swim
upward. If the distance to the surface of the water is not more than
three to four feet, they usually continue swimming until the apex of
the bell reaches the surface; they then turn over quickly. As soon
as the medusa has turned the apex of the bell downward, it ceases
contracting and passively sinks in the water; this continues until it
comes in contact with something to which it can attach itself by the
tentacles or manubrium. Animals which start upward at a great
distance from the surface become fatigued before they appear at the
surface, and sink back; but usually they do not turn over as do those
which reach the surface. It is noticeable that at low tide many more
small individuals come to the surface after the water has been stirred
than at high tide. Since the number of large animals does not vary
markedly with the tide, it would appear that the small ones are not
able to swim to the surface when the distance is relatively great.

Why do the animals swim surfaceward instead of in the opposite
direction when they leave their attachment? Evidently there are at
least three factors which might determine the direction of movement
either separately or working together: gravity, pressure, and light.

Reactions of Medusa Gontonema Murbachiz. 281

When a Gonionema swims upward it is swimming against gravity, —
that is, it is negatively geotactic. The medusz show the same kind
of reaction when they swim against a stream of water. While react-
ing positively to the tendency of gravity the animal is moving into
regions of lower pressure, inasmuch as this force varies directly with
the depth of water, and it is also coming into more intense light.
Which, if any, of these agencies determines the direction of the
animal’s movement ?

That light is not the all important cause can be shown by placing
the medusz in a vessel which is illuminated from below alone. Under
such conditions they still swim upward when disturbed. That pres-
sure is not of prime importance is proved by putting the animals into
solutions of different densities, as well as by the use of sea-water with
layers of other fluids above it. The animals uniformly move upward.
It is probable, then, that gravity is chiefly responsible for the re-
action. Ifasmall piece of cardboard or wood is placed on the sur-
face of the water, animals coming against it continue swimming just
as they do when free in the water. Often they continue until ex-
hausted, when they sink, not by turning over, but as do individuals
which have failed to reach the surface. Although negative geotaxis
seems to be the most important feature of this reaction of Gonionema,
it is not probable that light is without influence; it is simply not the
prime and necessary factor in the determination of the reaction.

In the reactions thus far described, we cannot say that Gonionema
is positively phototactic, for although it does move toward the source
of light, and into greater intensities, when swimming surfaceward,
light is not the cause of the movement. On the other hand, experi-
ments in the laboratory have shown that the animals are positively
phototactic to certain intensities of light.!

REACTIONS TO SUNLIGHT.

In this connection reactions to direct sunlight are to be considered.
When Gonionemata in a glazed white earthenware dish are placed in
sunlight, they at once become exceedingly active. At first they swim
surfaceward, often forcing the apex of the bell several millimetres out
of the water as they reach the surface; later many cease coming to
the surface, while others begin to swim downward, bumping against
the bottom of the dish; after a still longer interval cf exposure, some

1 YERKES, R. M.: This journal, 1902, vi, p. 446.

282 Robert M. Yerkes.

are found to move upward with the contractile phase of each beat
and downward with the expansile phase. If now the sunlight is cut
off from one end of the dish, thus leaving one portion bright and the
other shaded, in a few minutes most of the animals will be in the

shaded portion.
Careful observation reveals the following reaction as an explanation
of this fact. When an individual in swimming about chances to cross
from the sunlit region

SUNLIGHT. TRANSITION LINE. SHADOW. rs 5
SURFACE OF into the shadow, it very
WATER. 2 : ‘ Z
Ey sae quickly ceases swim-
i ming and sinks to the
bottom. If, later, in
. swimming about it
srw OF ‘ae chances to cross from
WATER. ey - (a

the shaded region into
the sunlight, it in most
cases immediately
ceases swimming, turns
over, and sinks passively to the bottom. But, in this case, when it
again becomes active, it does not move indifferently in any direction,
as it does when in the shadow; instead, it usually turns in such a
way as to move back into the shaded region. Because of this
reaction, the animals rapidly gather in the shadow. Both ¢he sudden
decrease in light intensity, as the animal passes from sunlight to
shadow, and the sudden increase, as it passes in the opposite direc-
tion, Zemporarily inhibit activity.

To indicate the usual number of animals in the sunlight and in the
shadow, and also the number of times that animals react to the sun-
light by sinking and then swimming back into the shadow, the follow-
ing series of observations is given. During a period of forty-five
‘minutes (see Table 1), fifty-four reactions to the transition line, as
the region of change of light intensity may be termed, were observed ;
of this number forty were of the type already described, 7.¢., the
animal ceased swimming as soon as it entered the sunlight, turned
over, sank to the bottom, then, after a short interval, swam back into
the shadow. The remaining fourteen cases were those of continued
swimming in the sunlight. In almost three cases in four the medusz
swam back into the shadow instead of continuing their activity in the
sunlit portion of the dish, as they would evidently have done had the
intensity of illumination been the same throughout the dish.

Diagram indicating the various positions
assumed by a Gonionema in the sunlight-
shadow reaction.

Reactions of Medusa Gontonema Murbachiz. 283

TABLE I.

SHOWING THE NUMBER OF GONIONEMATA IN SHADOW AND IN SUNLIGHT.

No. in shadow | No. in sunlight
(4 of dish). (3 of dish).

Time.

a.m.
10.00
10.05
10.10
10.15
10.20
10.25
10.30
10.35
10.40
10.45

Total

The reaction just described is to all appearances a perfectly defi-
nite response to sunlight. But, one may ask, is it not a response to
heat instead of to light? Yo obtain evidence on this point a two-inch
screen of alum solution was placed between the sun and the vessel in
which the medusz were swimming. This screen cut off the heat,
whilst it permitted the light to pass with diminished intensity. Un-
der such conditions the animals reacted in essentially the same way.
We may therefore conclude that sunlight, by virtue of its intensity,
causes a negatively phototactic reaction in Gonionema.

During experiments with sunlight it was discovered that the me-
dusze after being exposed to the light for a short time begin to try to
avoid it. When first exposed, they swim upward as in their natural
habitat, whereas after an interval, dependent upon the intensity of
the light and the condition of the animals, they swim downward and
persistently bump against the bottom or edges of the vessel. If a
number of Gonionemata differing in size, sex, and sexual condition,
are exposed to the sunlight as described, those with mature sexual
products (heavy gonads) are the first to try to avoid the light, and
later the others react in similar fashion. Clearly the animals are

284 Robert M. Verkes.

attuned, so to speak, to a certain range of light intensity, and are
negative in their reactions to higher intensities. The “ sexually
ripe”’ individuals are more sensitive to sunlight, and react more
violently to it than immature individuals. The animals discharge
their sexual products when in very weak light or darkness.

THE DIRECTIVE INFLUENCE OF LIGHT.

Previously statements have been made concerning the directive
influence of stimuli! Weak light stimulates the animals to activity,
but apparently does not influence the direction of their movements
to any marked extent; light of medium intensity seems to cause
movement toward the source of light at first, but it is later avoided;
finally, strong light, although it may attract at first, soon repels the
animals. The greater the intensity, the greater the directive influence
of light. We must now inquire, how does light direct the movements
of Gonionema ?

This question can be answered satisfactorily by certain simple
experiments. In the first place we have seen that Gonionema when
it sinks to the bottom after coming into sunlight usually turns
back toward the dark. This apparently is accomplished by the more
forceful and earlier contraction of that side of the bell furthest away
from the shadow. That this inequality of contraction is due to differ-
ence in stimulation seems probable, since the side away from the
shadow is likely to be exposed to slightly more intense light than the
remainder of the bell.

Clearer and more satisfactory evidence of the directive value of
stimuli in support of this statement is furnished by experiments with
tactual and electrical stimuli. When a Gonionema, resting on the
bottom of a dish with the apex of the bell up, is touched at some
point on the margin, z¢ reacts by swimming upward and away from
the side stimulated (Fig. 1). In not more than ten per cent of the
cases does an animal thus stimulated swim toward the region of stimu-
lation. If the animal chances to be resting bell down, the same direc-
tion of movement is taken, but the position of the animal necessitates
such a turning as Fig. 2 represents. Fig. 1 shows the direction of
movement when the stimulus is applied at S to an individual resting
bell up. This reaction is much more satisfactory and definite when
electrical stimulation is employed, since it does not involve the inter-

' YERKES; This journal, 1902, vi, p. 447.

Reactions of Medusa Gontonema Murbachiz. 285

ference with the organism’s movement which touching it is likely to

cause.
turns is determined by inequality
of contraction. The region stimu-

lated contracts first and most

strongly, and the water thus being
driven out of that side of the bell
tends to force the bell forward on
the stimulated side, therefore turn-
ing the region stimulated upward

movement when
animal is stimu-
lated in “bell-up”
position. Smarks
the region stimu-

There can be no doubt that the direction in which the bell

ipa)

Fic.1.—Directionof Fic. 2. — Direction

of movement when
animal is stimu-
lated in “bell-
down” position.
1 marks the posi-

if the animal is bell up, and down-
ward if it is bell down. Any one
who tries the experiment of ap-
proaching electrodes to a point at
the margin of Gonionema will get
a beautifully definite demonstration of the directive influence of a
localized stimulus.

In sunlight the medusz at first come so forcibly to the surface that
half the bell may appear above the water. Usually a bell thus forced
out of the water inclines slightly to one side, and thus tends to turn
the animal over, so that the observer can never be quite sure whether
the turning is due altogether to the action of gravity, or whether it is
in part due to some peculiarity in the final contraction of the bell.
After having been in sunlight for a few minutes, the animals come
to the surface with noticeably less impetus, and soon they fail to
reach the surface at all. But even in the latter case they turn over,
just as they do in ordinary light after they reach the surface. Under
these conditions the turning due to stimulation by light evidently
is accomplished by the bell contraction. Observation indicates that
the side of the organism which is exposed to the most intense light
contracts first and most strongly, thus forcing the bell over. As
soon as the bell is downward, movement ceases and the animal sinks
passively.

Thus far we have learned that Gonionema is either positively or
negatively phototactic, according to the intensity of the stimulus, that
its activity, wzder certain conditions, is inhibited by strong light, that
light and not the heat accompanying it is responsible for the reactions
described, and that the direction of its movements is definitely de-
termined by light as well as by other localized or unequally applied
stimuli.

lated. tion when stimu-
lus was applied;
2 indicates the di-
rection of move-
ment.

286 Robert M. Verkes.

EFFECTS OF INCREASE (LIGHT) AND DECREASE (DARKNESS)
OF PHOTIC STIMULATION.

We now come to the problem of inhibition of activity which has
appeared in various forms during the experiments. Berger?! thus
summarizes the results of a study of Charybdea made by him and
Conant: ‘We have seen that it is very sensitive to light, strong
light as also darkness inhibiting pulsations, while moderate light
stimulated it to activity. Also, a sudden change from weaker to
stronger light, or vice versa, may inhibit or stimulate to activity
respectively. This behavior of Charybdea seems to be correlated with
its habit of life on the bottom. We have no reason to doubt but that
the eyes of the sensory clubs are the seat of light sensation.” And
Romanes,? who has studied the reactions of many kinds of jelly-
fishes, remarks that Sarsia is sensitive to light, and usually responds
to a flash by one or more contractions of the bell. He, moreover,
believes that the change from light to darkness is inhibitory of
activity.?

In studying the reactions of Gonionema to photic stimuli or the
‘“ photoskioptischen Sinn” (light-darkness sense), as Nagel* has
named this kind of sensitiveness, I have found that Berger’s descrip-
tions of the reactions of Charybdea do not hold true for Gonionema,
and further, that his statement that Gonionema’s activity is inhibited
by strong light is misleading.

Gonionemata when taken from the sea and placed in ordinary
daylight are active for a long time. Whenever the intensity of the
light is considerably decreased, they come to rest on the bottom of
the vessel. Any sudden increase in the light causes them to become
active again. In general, we may say that increasing the intensity
of light causes Gonionema when at rest to react by contracting the
bell; decreasing the intensity causes no reaction uniformly, although
in exceptional cases an animal may react regularly to decrease in

1 BERGER, F. W.: Memoirs of the Biological Laboratory of Johns Hopkins
University, 1900, iv, p. 22.

2 ROMANES, G. J.: Jelly-Fish, Star-Fish, and Sea-Urchins ; being a research on
Primitive Nervous Systems. New York, 1885, p. 39 e¢ seg.

8 For discussion of the stimulating effect of decrease in illumination, see
RaAwitz, b.: Jenaische Zeitschrift, 1892, xxvii, pp. 1-232; Drost, K.: Morphol-
ogisches Jahrbuch, 1887, xii, pp. 163-202; NAGEL, W. A.: Der Lichtsinn augen-
loser Thiere Jena, 1896.

* NAGEL, W. A.: Biologisches Centralblatt, 1894, xiv, pp. 385-390.

Reactions of Medusa Gontonema Murbachie. 287

intensity, and fail to react to increase. If, however, a swimming
Gonionema is stimulated by an increase in the intensity of light
it expands and settles to the bottom. So, likewise, an individual
resting quietly in strong light may very frequently be made to react
by cutting off the light. Increase in light intensity uniformly causes
a visible reaction in from one to thirty seconds; but the nature of
the reaction depends upon the condition of the animal at the instant
of stimulation. If it is in a state of expansion (inactive), the light
causes it to contract, hence the locomotor reaction; if it is ina
state of contraction (active or inactive), the light causes it to
expand, hence the absence of a locomotor reaction in the presence
of a visible motor reaction. Decrease in light intensity almost uni-
formly causes a contracted animal to expand and come to rest, but
it does not, except in certain unusual cases, cause an individual
which is resting ina state of expansion to contract and become active.
-Excised margins quite commonly react to decrease as well as to in-
crease of light intensity by contracting, but seldom are normal
individuals found which show this uniformity.

Berger’s statement would be strictly correct if he had said that the
influence of strong light on an active Gonionema is to inhibit its
activity. In view of the above facts, it is clear that increase in the in-
tensity of light has in part the same effect upon an active Gonionema
as has decrease in intensity; both tend to inhibit activity. On the
other hand, a quiescent animal is stimulated to activity by increase in
the light intensity, while it gives no visible reaction uniformly to
decrease in intensity. Obviously we cannot say, then, that strong
light and darkness always inhibit activity. There can be no doubt
that in the majority of cases light, 7. ¢., increase in the intensity of
illumination to which an animal is exposed, causes activity, whereas
darkness tends to keep the animal at rest. In the study of reaction
time to photic stimuli to be considered later, the quieting influence
of darkness and the stimulating influence of light were made use of
as experimenting conditions. The animals when covered, in a dish,
soon come to rest, and when uncovered and thus exposed to more
intense light react quickly by contracting the bell.

The reaction of Gonionema to sunlight which has been described ~
is now seen to be in agreement with the results obtained with day-
light! In the experiments with sunlight the animals emerged from
a weaker into a stronger intensity of light; hence, their activity was

1 YERKES: This journal, 1902, vi, p. 446.

288 Robert M. Verkes.

temporarily inhibited, and they sank to the bottom. Ina few minutes,
however, the strong light forced a motor reaction which for reasons
already pointed out resulted in a movement toward the shadow.

What may be conceived as the relation to the habits of the medusa
of this temporary inhibitory effect of increase in light intensity is of
interest. Gonionema when it reaches the surface of the water, after
being disturbed in its habitat, turns over, and by reason of the in-
fluence of the light, or if not because of it, at least in the presence of
this factor, passively sinks downward with fully expanded bell and
tentacles. As it slowly sinks in this condition, it has a far better
chance of coming in contact with food than it would if it swam
down rapidly with contracted tentacles. A similar line of reasoning
is applicable when we consider the position which is assumed by the
sinking animal. It generally sinks with the bell pointed downward
and the manubrium swinging free in the water above the bell. Now,
as the animal sinks, any prey which becomes ensnared in the ten-
tacles can readily be carried to the lips by the movement of the
tentacles upward with the current, whereas, since the animal is
sinking, it would be difficult for the organs to carry food downward
against the water pressure and enable the lips to seize it. It may
be maintained, therefore, that the reaction is especially suitable for
the obtaining of food.

Light, having been one of the most important stimuli for the
initiating of the surface reaction, — 7. ¢., turning, inhibition of activity,
expansion of organs, — may finally have come to be the cause of a
very similar reaction in the presence of other conditions, so that
increase in light intensity now uniformly causes an active Gonionema
to expand fully for a few seconds and permit itself to sink passively.

THE REACTION TIME OF GONIONEMA TO PHOTIC STIMULI.

Method. — Since a resting Gonionema almost invariably responds
to an increase in light intensity with a sharp locomotor contraction
of the bell, and since the time of this reaction is of such length
as to be measurable with considerable accuracy by means of a
stop watch, the light reaction of this organism offers an excellent
opportunity for the study of the influence of various environmental
conditions on the time of reaction.

The remainder of this paper is a description of experiments made
to determine: the influence of intensity of light on reaction time;

Reactions of Medusa Gonionema Murbachiz. 289

the influence of temperature on the reaction time to light; and the
relation of sex, sexual condition, size of organism, and pigmentation
to reaction time.

The following simple method served in the study of reaction time
to light. Into a white earthenware dish containing about two inches
of sea-water, placed before a window, a single Gonionema was put.
So long as the animal was exposed to the ordinary daylight, it was
active much of the time; but by placing a piece of card-board over

TABLE II.

REPRESENTATIVE SERIES OF REACTION TIMES TO DAYLIGHT.

Animal No. 3. | Animal No. 4.
14 mm. in diameter. | 21] mm. in diameter.

Experiment.

Average

Mean variation

the dish to weaken the light considerably, the animal could be
brought to rest in a short time. Trial showed that from one to
five minutes was sufficient to allow the animal to settle down. If
after a certain period the dish was quickly uncovered the medusa
reacted to the increase in light by sharply contracting the bell. The
time of this locomotor reaction was determined thus. The instant
the experimenter removed the cover from the dish he started a stop
watch, and the instant the animal reacted he stopped the watch and
noted the time of reaction. Then, after permitting the animal to

290 Robert M. Verkes.

swim about actively in the daylight for a period equal to that during
which the dish had been covered, he again covered the dish. In the
following experiments, unless otherwise indicated, the time of ex-
posure to both intensities of light was five minutes, that is, the dish
was covered for that period, and then uncovered for the same period.

TABLE III.

REACTION TIMES TO THREE INTENSITIES OF LIGHT.

Weak light. Medium light. Strong light.

Animal.
Mean
variation.

Mean Mean

cae ’ Sais Mean.
variation. variation.

1.62 i 1.58 1.09

4.67 : ; 1.16
9.16 : 9! 12.92
20:34 : 7.16

SL é 3.08
1.86 ; : 2.03
3.67 : 1.93

General av.

Generally ten reactions of each animal were measured. Table II
gives two representative series of reaction times. The time of re
action to daylight is ordinarily about seven seconds, and as can be
seen from the table, the variability is about two seconds. In no case
was a slow contraction of the bell which did not introduce a loco-
motor reaction counted as a response to light.

The influence of the intensity of light on the time of reaction. —
Several animals were tried in each of three intensities of light. For
the lowest intensity weak daylight, obtained by drawing the shade of
the window in front of the dish, was used, for the medium intensity
ordinary daylight, and for the strong intensity direct sunlight.
Tables III, IV, and V contain the results of the experiments made.

In each table the figures indicate the average for ten reactions, and
their average variability.

Reactions of Medusa Gontonema Murbachit. 291

In Table III, animal No. t shows a regular decrease in the time
of reaction and likewise of variability with the increase in light
intensity. For it the statement, the stronger the stimulus the shorter
and less variable the reaction time, is true. Animal No. 2, however,
reacted to the medium light little quicker than to weak light. The
variability, moreover, is extremely great, 7.76 seconds. The meaning
of these facts is easily discoverable, for when we refer to the record
of individual reaction times, we find that three very slow reactions
are included in the series; hence, the unexpectedly long average
reaction time and the high variability. Clearly this result, because
of the great difference in variability, is not directly comparable with
that obtained with animal No. tr. Careful examination of thousands
of reaction times, and of the conditions determining them, in so far

TABLE IV.

REACTION TIMES TO THREE INTENSITIES OF LIGHT.

Weak daylight. Daylight. Sunlight.

Mean Mean Mean

Mean. oleae oie nee
variation. variation. variation.

1.62 : 3:05 : 2.52

SV/All De 1.63 : 0.55
3.67 5 1.89 : 1.92
2.63 =) : 1.09
4.67 : : i ROS

3.16 : ’ : 1.60

as observable, shows that approximate equality in variability is the
only safe basis for direct comparison of reaction times. By equality
is meant proportionate, not absolute equality. If for one animal
the variability of an average reaction time of ten seconds is four
seconds, for another individual, the conditions remaining constant,
the variability of a reaction time of five seconds should be about
two seconds in order to make direct comparison valuable. The
general averages of Table III are curiously misleading. They in-
dicate a decrease in the time of reaction to medium light as compared
with weak light, and also as compared with strong light. This fact

292 Robert M. Yerkes.

in itself is not surprising in view of what we have already learned
concerning the inhibitory effects of sunlight. But all the results
which show anything approaching proportionate equality in variabil-
ity! indicate a shorter time of reaction for sunlight than for either
of the other intensities. In Table IV, for instance, the results of the
reactions of five animals is given for the three intensities of light.
The results are not for the same individuals for each intensity, as in
Table III, but for animals taken at random.

Comparison of the averages of this table show: (1) decrease in
reaction time with increase in intensity of illumination, and (2)
approximate proportionate equality in variability. In no case is the

TABOR, Vi.

REACTION TIME TO DAYLIGHT AND TO SUNLIGHT.

Daylight. Sunlight.

Animai.
Mean Mean

Mean. mans eins
variation. variation.

3.35 : 202:
1.63 : 0.55
3.44 : 1.92

General av. : 2.81 : 1.70

latter much more than one-third of the reaction time. A variability
in excess of one-third of the reaction time is indicative of irregular-
ities in the series of reactions due to lack of uniformity of conditions.

The averages of Table III are misleading because of the exceptional
reactions of some of the animals. That weak light usually causes a
slower reaction than ordinary daylight is proved by the results of
Tables III] and IV. That sunlight does not at first cause a slower
reaction than daylight is proved by the results of Tables IV and V.
In the latter table the times stand 5.5 to 6.8. The variability of the
sunlight reaction time is 1.70, that is, thirty-one per cent of the re-
action time. The variability of the daylight reaction time is 2.81,
that is, forty-one per cent of the reaction time. This increase in

1 The ratio of variability to reaction time being about the same for different
individuals.

Reactions of Medusa Gontonema Murbachiz. 293

variability with decrease in strength of the stimulus is precisely what
all previous reaction time work leads us to anticipate.

The conclusions to be drawn from the results of Tables II to V are:
(1) Gonionema’s reaction time to light becomes shorter as the stimu-
lus increases in strength; (2) absolute variability of reaction time also
decreases as the stimulus increases; (3) sunlight usually causes
quicker reaction than daylight, and (4) a variability of more than
thirty to forty per cent of the reaction time is indicative of incon-
stancy of conditions.

Relation of reaction time to light to size. — Small Gonionemata
are more active than large ones. Counting the number of contrac-
tions of the bell per minute for a period of five minutes in case of
animals of different sizes yielded the following results :

Animals of 20 mm. in diameter gave an average of about 15 beats per minute.

“c “c 15 te “<c “c “< “c “¢ 73 “c Dy “ “ce “

“ce “e 10 “cc “ce “ce “ “ “ “ “ 2 “ “cc “
30

“ce “ 5 “c “ee “ oe “é “e “ec “ce 51 “ “e “

There are of course noteworthy exceptions to the rule, but un-
doubtedly activeness varies inversely as the size.

Further evidence of the relation of size to activeness and to the
time of reaction was furnished by observations made on Gonionemata
which were left in covered dishes over night, and whose reaction
time to light was noted when they were uncovered in the morning.
The average time of nine small animals under such conditions was
thirteen seconds, whereas for eight large animals it was 21.8 seconds.
A large number of observations gave averages of about fifteen
seconds for the small individuals (5-12 mm.), and about twenty
seconds for the large ones (12-20 mm.).

We have now to ask whether this relation of size to time of re-
action to light appears in the results of the reaction time experiments
already described. Ten reactions for each of twelve Gonionemata,
ranging in size from 5 to 10 mm. in diameter, gave an average re-
action time of 7.3 seconds. And a like number of reactions for each
of eighteen individuals, which ranged from 11 to 16 mm., gave an
average reaction time of eleven seconds. The smaller individuals
are much more active and usually respond to stimulation by light
more quickly.

Relation of reaction time to light to pigmentation. — Gonionemata
usually have granular pigment in the gonads, tentacles, margin and

294 Robert M. Verkes.

radial canals, but the quantity of this pigment varies greatly among
individuals; some are so light as to be almost transparent when in
water, others are yellowish, greenish brown, or even dark brown.
The question may therefore be asked, are heavily pigmented individ-
uals detectably different from those which lack pigment in their
reactions to light?

The average reaction time for thirteen very lightly pigmented
individuals (ten reactions for each) was 7.9 seconds, and for the same
number of heavily pigmented animals 7.2 seconds. This difference
although not great is doubtless indicative of the physiological impor-
tance of pigmentation. Moreover, comparison of all the results shows
that the dark individuals are quicker, size, sex, temperature, etc.,
being considered, than are light ones. This fact came out most
strikingly in the general tests of light sensitiveness made by placing
ten Gonionemata in a large dish, leaving them covered over night,
and then noting their positions and time of reaction when they
were exposed to light in the morning. By this method a group of
ten light animals was found to react in 22.5 seconds under the
conditions which gave for a similar group of dark individuals a re-
action time of fifteen seconds. As in case of large and small animals,
an individual is often found which gives contradictory results, so in
the case of light and dark individuals one occasionally meets with
marked exceptions to the rule that dark animals react more quickly
than light ones.

Relation of reaction time to light to sexual condition. — There
can be no doubt that the reaction of Gonionema to light is largely
dependent upon its sexual condition. Individuals with ripe gonads
if brought into the laboratory can be made to discharge their sexual
products in from thirty minutes to two hours by putting them in the
dark.! In earlier experiments attention was called to the fact that
animals with heavy gonads when exposed to direct sunlight become
negative to it, and swim toward the bottom of the dish sooner than
do immature individuals.. The greater sensitiveness of sexually ripe
individuals is thereby -indicated.

Reaction-time results of this paper show no pronounced differences
between males and females.

The influence of changes in temperature on the time of reaction. —
A sudden increase in temperature shortens the reaction time, a sudden

1 See PERKINS on the Development of Gonionema murbachii, Proceedings of
the Academy of Natural Science, Philadelphia, 1902, liv, pp. 755-756.

Reactions of Medusa Gontonema Murbachie. 295

decrease lengthens it, or causes complete inhibition of reaction. The
purpose of the following experiments was to ascertain the effects
of a gradual change in temperature upon the photic reactions of
Gonionema.

A single animal, for the experiments, was placed in a small earthen-
ware dish containing water, and this dish was placed in the water of a
larger vessel. Then, by the application of heat to the outer vessel,
the temperature of the water in which the medusa swam was raised
regularly at the rate of one degree in ten minutes. By placing ice in
the water of the outer vessel, the temperature was lowered at the
same rate.

For each animal used the reaction time to daylight at the tempera-
ture of the sea-water was first determined. Then as the temperature
was gradually changed one reaction per minute was taken, thus,
allowing for irregularities of reaction and time to read and record

TABLE VI.

REACTION TIME TO DAYLIGHT AT DIFFERENT TEMPERATURES.

MENS (Cc NE OSI:

Mean Mean
variation. variation.

3:22 ab 1.60

3.74 : 0.89

2.04 ; 0.64
3.30 : 59

3.00 . 1.18

results and take the temperature readings, it was possible to get at
least five reaction time readings while the temperature changed one
degree. As in the previous experiments, the reactions were given to
increase in light intensity caused by uncovering the dish at regular
intervals, These experiments were performed during August and
September, and as the temperature of the water from which the
Gonionemata used were gotten fell four or five degrees during that
period, an opportunity was given for noting whether the reaction time
of the animals changes with this environmental change. In Table VI,

296 Robert M. Yerkes.

the reaction time to daylight is given for four animals taken from
water about 17° C., and for the same number of animals taken
from water of 19° C. temperature. At the higher temperature the

TABLE ‘VIL

THE EFrecr OF GRADUAL INCREASE IN TEMPERATURE ON REACTION TIME
TO LIGHT.

Reaction Mean

Temperature. ; 5 pe
time. variation.

secs.

19° (normal) 9.1 1.40
7.8
6.5

6.2

6.3
BEY)

Saif
D2,
4.6
3.8
3.7
Sh
2.7
2.6
4.8
est

failure to respond.

reaction time is 7.2 seconds, the variability only 1.18; at the lower
temperature the time is 11.2 seconds, the variability 3. It is sig-
nificant that such a great increase in variability accompanies the
lower temperature. This variability, it may be remarked, is greater
in proportion to the time of reaction as well as absolutely. For the
19” reactions, the ratio of variability to reaction time is I : 6.1 (vari-
ability of 16+ per cent), for the 17° reactions the ratio is I : 3.7
(variability of 27— per cent). This shows that temperature is an

Reactions of Medusa Gontonema Murbachiz. 297

important factor in the determination of the variability of reaction
time. That strength of stimulus is one of the most important deter-
minants is already well known,! but for the interpretation of reaction
time results it is of importance to discover precisely what other con-
ditions need to be considered. By such studies as this I hope to be
able to discover the relations of the chiefly significant environmental
conditions to the time and variability of an animal’s reactions.

Table VII presents the results of series of experiments with three
Gonionemata. Each animal was placed in a dish of water of normal
(sea-water) temperature, the reaction time to daylight determined,
and the temperature then raised at the rate of one degree in ten

TABLE VIII.

THe EFFECT OF GRADUAL DECREASE IN TEMPERATURE ON REACTION TIME
To LIGHT.

Reaction Mean

Temperature. : sie
time. varlation.

19° (normal)

18°

15.80

most individuals irresponsive.

minutes until reactions to light ceased. Each result (¢.g. 20°, 7.8
seconds) in the table is the average time for fifteen reactions, five
for each of three animals, which were given while the temperature
was increased one degree. This table indicates a gradual shortening
of the time of reaction until 32° is reached. Between 32° and 33°
the reactions are quickest; but if an animal be exposed to this tem-
perature for a few minutes, the reactions rapidly become slower and
soon cease. At 33° an increase in reaction time begins, and before
34° is reached reactions usually fail entirely, and the animals soon

1 YERKES: Harvard psychological studies, 1903, i, pp. 579-638.

298 Robert M. Yerkes.

die unless removed to a lower temperature. In a few cases the vari-
ability has been given in Table VII in order to show the decrease in
variability which is correlated with the quickening of reaction time
due to increase in temperature. The shortest reaction time, that at
32°—33°, is also the least variable. Gradual increase in temperature
up to a certain point shortens the reaction time of Gonionema to
light; what is true of decrease in temperature ?

The apparatus employed in the study of the effects of increase
in temperature was here used again, with ice in the outer vessel.
The averages of Table VIII are based on the reactions of four
animals. The reaction time usually lengthens as the temperature
decreases, but in certain cases a shortening of the time at first is
followed by a rapid lengthening. In these exceptional cases cold
seems to act as a stimulus, at first, in much the same way as does
heat.

Decrease in temperature, as is shown by the results of Table VIII,
gradually increases the time of reaction until reactions finally cease
entirely at a temperature of from 12° to 14°. An animal which has
been exposed to low temperature for hours often reacts to daylight
when at a temperature of 9°-10°.

Increase of temperature is accompanied by greater activeness of
Gonionema until 33° is reached. At that point the animals begin to
weaken, contraction becomes irregular and partial, and the margin
takes the form of a square instead of preserving its usual circular
form. Apparently heat causes the contraction of the portions be-
tween the radial canals. The manubrium usually expands fully at
from 30° to 33°, and becomes irresponsive to mechanical or photic
stimuli.

The effects of decrease of temperature on Gonionema in contrast
with those of increase are a general lessening of activeness and a
tendency toward quiescence in a state of expansion. At from 8° to
12° most individuals become entirely irresponsive to ordinary stimuli.
They do not appear to be weakened or in any way injured by the
low temperature, as in case of high temperatures, for after removal
from ice water to water of ordinary temperature they soon recover
their usual activenessand respond normally. Heat causes contraction ;
cold causes expansion.

keeactions of Medusa Gonionema Murbachit. 299

LOCALIZATION OF SENSITIVENESS OF GONIONEMA TO PHOTIC
STIMULI.

Early in the course of this experimental study it became evident
that Gonionema’s reaction to light partly depends upon its position
with reference to the source of light. Usually the animals after a
period of activity come to rest with the apex of the bell on the bot-
tom of the experiment dish; this position we may know as the “ bell-
down” position. In this position the under surface of the margin,

-TABIER, BXe
SERIES OF REACTION TIMES SHOWING THE INFLUENCE OF POSITION.

+ indicates reactions when the bell was up. — indicates reaction when the bell was
down. | indicates reaction when the bell was on edge.

No. 1, 14 mm. No. 2,11 mm. No. 3, 18 mm.
Dark color. Light color. Light color.
Temprl/Ag: Mempo Wie: Hemp ZLe?

Sih 5.3— Toa
10.4— @a= 10.8—
13.0— 17.0+ Tete
Oo= 6.8—

the inner surface of the bell, and the manubrium are exposed to the
light. Sometimes, however, an animal settles down with the apex of
the bell uppermost, this, the “ bell-up” position, does not expose
so much of the margin to the action of light as does the “ bell-down”

300 Leobert M. Yerkes.

position. Gonionemata which have been left undisturbed in weak
light for a few hours are usually found resting “ bell-up.”

Three series of reaction times, given by individuals differing in
size and color, are presented as evidence of the relation of reaction
time to the animal’s position with reference to the light. Reactions
when the animal was resting “bell-up” are followed by a + sign;
reactions when the animal was resting “ bell-down,” by a — sign.
The results of Table IX indicate that the “ bell-up ” position is seldom
assumed, and that when in this position the medusz react much
more slowly to light than when resting “ bell-down.” An examina-
tion of all records obtained shows that Gonionema is much more
likely to come to rest “ bell-up,” in the presence of some disturbing
condition, such as high or low temperature, than under usual condi-
tions, while in almost all cases the reaction time is much shorter for
the “ bell-down ” position.

This evident dependence of reaction time to light on the position
of the animal necessitated the rejection of all ‘“ bell-up” reactions
from averages, except where the results were used for the special pur-
pose of showing the influence of position. The necessity is evident
when one examines such series as those of Table IX. There, in case
of No. 1, three exceptional (“ bell-up” +) reaction times, if included
in the series, increase the average greatly, and cause the result to
be quite misleading.

Having noted the importance of position, we have now to ask why
the reaction time for the ‘“ bell-down” position is so much the
shorter? That the difference is caused by greater ease in movement
is unlikely; that the ‘bell-down” position is that in which the
animals most frequently react to stimulation by light, and, therefore,
the more favorable is possible ; yet, one first of all is inclined to
believe that a difference of sensitiveness is at the bottom of the
matter. If the “bell-up” position causes certain organs or parts of
Gonionema which are especially sensitive to light to be covered,
whereas the opposite position leaves them freely exposed to the
stimulus, a sufficient cause for the difference in reaction time
appears.

With a view to determining what portions of the medusa are
sensitive to light, and at the same time solving the problem of the
difference in reaction time above noted, certain experiments now to
be described were tried.

First, the effect upon photic reaction time of excising all the

Reactions of Medusa Gonionema Murbachi. 301

tentacles of an individual about 2 mm. from the margin was noted.
This experiment seemed worth while because of the evident impor-
tance of the tentacles as sensory, and as orientation organs. The
question is, are they sensitive to light in any marked degree, and
does the medusa when deprived of them react otherwise to light than
the normal individual? In many cases the operation of tentacle ex-
cision causes a shock from which the animal never fully recovers,
but in others the reactions to light continue uninterruptedly.

The reactions of a 16 mm. individual are presented in Table X.
For this individual the normal reaction time before the operation was

ANIC ULO: DS
REACTION TO LIGHT AFTER REMOVAL OF TENTACLES.

Reaction time of normal animal, 4.7 seconds. Reactions taken ten minutes after re-
moval of tentacles. Missed indicates failure to react within one minute.

Missed + 20.8+-
6.2— 3:4—=
15.6— 1-4—
98— : Missed +-
8.7— : 4.8+
Zz : 2.2—
1.2— Missed +
1.6— : IT —
1.8— 2 Voy,

Missed 4- ; 44.]4+

Variability.

General average of all reactions) (26). 2 3 2 = : 5.74
Average of all ‘“bell-up” reactions (+) (5) . . . . 12.06

Average of all “bell-down” reactions (—) (21) .. . : Peele

4.7 seconds. After the operation the “bell-up” reactions were ex-
ceedingly long, the “bell-down” reactions unusually short. This
result, which was gotten also with other individuals, seems to indicate
that when a normal animal is resting “ bell-up,” the tentacles are
stimulated by light sooner than are the organs of the under side of

302 Robert M. Yerkes.

the margin; hence, the animal reacts more quickly with the tentacles
than without them. When the tentacles are removed, the “ bell-
down” reaction time is slightly quicker, because of the increased
irritability of the margin. The results show: (1) that the tentacles
are sensitive to light, and (2) that they are not responsible for the
difference in the reaction times to light for “ bell-up” and ‘ bell-
down” positions.

Experiments with the marginless bell prove conclusively that the
bell itself is irresponsive to light. In these tests, light stronger than
direct sunlight and lamplight was not used. The marginless bell
reacts to strong mechanical, chemical, and thermal stimuli, but never,
so far as my observation goes, to photic stimuli. It thus appears
that the general surface of the bell is insensitive to light.

It may here be mentioned that animals whose tentacles were re-
moved by pulling them off in such a way as to destroy the marginal
bodies never reacted to light. The manubrium of such individuals
seems to be sensitive, but it does not initiate bell contractions.

We are now confined in our search for the special sense organs for
photic stimuli to the margin of the bell. Margins when excised
at a distance of four or five millimetres from the edge in some cases
contract closely and fail to respond with any regularity te daylight ;
others, however, remain expanded, except when stimulated, and
react uniformly. The facts brought out by experiments on excised
margins are: (1) that the margin, containing the marginal bodies,
is exceedingly sensitive to light, if quickness of reaction is to be
taken as evidence of sensitiveness, (2) that the margin, unlike the
normal animal, responds regularly to decrease in light intensity as
well as to increase, (3) that the upper surface of the margin is less
sensitive than the lower surface, (4) there is some evidence that the
‘“bell-up”” margin reacts more quickly to decrease in light intensity
than to increase, whereas the reverse is sometimes true for the
“ bell-down ” margin.

The ‘“bell-up” margin reacts to increase in light intensity very
seldom, but to decrease almost uniformly. When the time of reaction
(13.6 seconds, Table XI) for increase in intensity when the margin
is in the “bell-up” (+) position, is compared with the reaction to
increased intensity when the margin is in the “bell-down” (—)
position, account should be taken of the frequency of reaction as in-
dicated at the bottom of Table XI, as well as of the time. In these
experiments, increase of intensity was obtained by uncovering the

Reactions of Medusa Gontonema Murbachir. 303

experiment dish and exposing the animal to daylight, decrease by
covering the dish and thus cutting off a part of the light.

The excised margins always tend to maintain the “ bell-up” posi-
tion, and have to be turned over repeatedly during experiments which
require the ‘‘bell-down” position. In this we have additional evi-
dence of the greater sensitiveness of the under surface of the
margin.

TABLE XE.
REACTION TIME OF EXCISED MARGINS To INCREASE AND DECREASE OF LIGHT
INTENSITY WHEN IN “ BELL-UP” AND “ BELL-DOWN ” POSITIONS.

When in “ bell-up ” (+) position. When in “ bell-down” (—) position.

Increase Decrease Increase Decrease
in intensity in intensity in intensity in intensity
of light. of light. of light. of light.

Animal.

Reaction | Varia- | Reaction | Varia- | Reaction | Varia- | Reaction | Varia-
time. bility. time. bility. time. bility. time. bility.

20.07 1.1 0.29 y | 16.67
6.50 2.8

1.9 : 11.3 10.98

Reaction to about Reaction Reaction to about | Reaction almost
30% of stimuli. uniformly. 90% of stimuli. uniformly.

Certain observations concerning the reactions of normal Gonione-
mata to light bear on the same point. One animal, a 26 mm. female
with ripe gonads, reacted to increase in light intensity uniformly
when “ bell-up” (+) in from eight to fifty seconds, and when “ bell-
down” (—) in from five to ten seconds. To decrease in intensity
this individual reacted in about 25 per cent of the experiments,
and of this 25 per cent almost all were “ bell-up” reactions.
These results are entirely in harmony with those obtained with ex-
cised margins. The above individual when subjected to strong lamp-
light reacted uniformly to increase in intensity whether “ bell-up”’
or “bell-down.” It never came to rest in the light otherwise than
‘bell-up,” and as long as it was turned over by the experimenter in

304 Robert M. Verkes.

2)

such fashion as to prevent it from taking that position, it continued
swimming. Furthermore, the reactions in the “ bell-up”’ position
were very much slower than those in the “ bell-down” position. The
latter could be obtained only by turning the animal over, and then
cutting off the light immediately.

As evidence of the greater sensitiveness of the lower as compared
with the upper surface of the margin, as well as of the commonness
of reaction to increase but not to decrease of light intensity, the fol-
lowing results obtained with an exceedingly dark individual 17 mm. in
diameter are also presented. This animal always came to rest “ bell-
up” when in daylight, and just as regularly changed to the “ bell-
down”’ position during the time that the dish was darkened. When
in a shaded portion of the dish it usually assumed the “bell-down ”
position. To increase in light it often reacted so quickly that it was
impossible satisfactorily to measure the time with a stop watch. In
many cases the time was evidently less than one second. To de-
crease in light intensity, it reacted once in 4.3 seconds when “ bell-
up,” but thereafter twenty repetitions of the stimulus at regular
intervals of a minute when the animal was sometimes “ bell-up,”
sometimes “ bell-down,” failed to call forth a reaction. These indi-
vidual reactions bring out with unusual clearness what has been
found to be true in general for Gonionema.

In view of the above results we may conclude that the margin of
the bell contains organs which are especially sensitive to photic
stimuli; and since destruction of the marginal bodies destroys re-
sponses to light the probability is strong that they are the special
organs of photic stimulation! They are heavily pigmented organs
so placed that light can affect them directly only when the animal is
“ bell-down,” hence the slowness of reaction when an animal is in the
“bell-up position.

SUMMARY AND CONCLUSIONS.

1. In its natural habitat Gonionema is negative to intensities of
light greater than ordinary daylight. When disturbed, it leaves its
attachment to submerged sea-weeds and swims surfaceward. If it
reaches the surface while swimming, it turns over as soon as the
apex of the bell emerges from the water, ceases contracting the bell,

' For a description of the Nervous System of Gonionema, see the preliminary
paper of Miss I. H. Hype: Biological bulletin, 1902, iv, pp. 40-45.

Reactions of Medusa Gonionema Murbachiz. 305

and sinks passively with fully expanded bell and tentacles. Although
in moving surfaceward the medusa swims toward the source of light,
the direction of movement is not determined by light. The turning
reaction, inhibition of activity, and expansion which mark the appear-
ance of an animal at the surface of the water although not entirely
due to light are, at least in nature, uniformly associated with an in-
crease in the intensity of illumination.

2. Under experimental conditions, Gonionema moves toward the
source of light, z.¢., is positively phototactic. It comes to rest in the
darkest portion of the vessel, and is therefore negatively photopathic
to ordinary intensities of light.

3. Direct sunlight strongly stimulates the animals. When first
exposed, they swim surfaceward toward the light, but after a period
of exposure, dependent upon the intensity of the light and the size,
pigmentation, and sexual condition, they become negative to the
light and swim downward, and into the shaded regions of the
vessel.

4. When exposed to sunlight in a dish one end of which is shaded,
they soon gather in the shaded portion, and at least eighty per cent
of them are always found there. Although when an animal enters
the shaded part of the vessel its activeness decreases, and it fre-
quently comes to rest, this does not suffice to explain the gathering
of the animals in that region. Observation shows that in most cases
as soon as an animal passes from the shade into the sunlight it ceases
swimming, turns over, and sinks with expanded organs, just as it does
in its natural environment on reaching the surface of the water. Ap-
parently increase of light in this case has an inhibitory influence.
As soon as the medusa reaches the bottom of the vessel, it again
contracts, but this contraction is of such a nature as to turn the
animal toward the shaded region, so that it again swims back into
that portion of the vessel.

5. Light, like other stimuli, determines the direction of movement
of Gonionema through the unequal contraction of the bell due either
to localized or unequal stimulation. That portion of the bell most
strongly stimulated reacts by a more forceful contraction than the
remainder of the bell, and thus tends to turn the animal away from
the region of stimulation.

6. Electric stimulation most clearly demonstrates the “ directive
influence” of stimuli because it can be applied without interfering
with the animal’s movements by direct contact. A Gonionema rest-

306 Robert M. Verkes.

ing with the apex of the bell uppermost (‘“ bell-up” position), when
stimulated at a certain point on the margin by the approach of elec-
trodes, reacts by a contraction of the bell, which, because it is more
forceful at the point of stimulation, forces the side of the bell stimu-
lated upward most rapidly, thus tending to remove the animal from
the region of the stimulus, instead of taking it toward it. A “bell-
down ” individual reacts to the same kind of stimulus by a contraction
which, because of its inequality, forces the side stimulated downward
most rapidly, thus causing the bell to turn away from the stimulus.

7. That Gonionema is able to direct its movements by means of
such unequal contractions as are here mentioned, is proved by the
fact that animals which at first swim to the surface in direct sunlight
later turn over before reaching the surface.

8. Increase in light intensity uniformly causes a motor reaction
in quiescent individuals, and the inhibition of movement in active
individuals.

9. Decrease in light intensity usually causes the inhibition of
movement in active animals, but rarely does it act as a stimulus to
activity in case of resting animals.

10. Strong light is injurious to Gonionema, and a few hours’ expos-
ure is commonly fatal.

11. To the increase in light intensity resulting from the uncover-
ing of the vessel containing it, Gonionema uniformly gives a loco-
motor contraction of the bell. The time of this reaction in daylight
is from five to ten seconds.

12. The reaction time to light decreases with increase in the
intensity of the light. For weak daylight it is 9.4 seconds, and for
ordinary daylight 7.0 seconds.

13. The time of reaction to sunlight is 5.5 seconds as compared
with 7.0 seconds for daylight.

14. The variability of the reactions to light is usually about one-
third of the reaction time. A greater variability than this in ordi-
nary ‘reactions is indicative of some exceptional condition pikes
should be taken into account.

15. The activeness of Gonionema varies inversely as the size of the
animal,

16, Small animals, other things being equal, react more quickly to
light than do large ones.

17- Heavily pigmented individuals react more quickly than do
those which are lightly pigmented.

Reactions of Medusa Gontonema Murbachit. 307

18. Sexually mature individuals appear to be more sensitive to
light than immature or senile animals,

19. Increase in temperature gradually shortens the reaction time
to light. At 33° the reaction time is about 2.5 seconds, with a varia-
bility of only 0.4 of a second. Beyond this temperature the reaction
time rapidly lengthens, and at 34° the animal soon perishes.

20. Decrease in temperature gradually lengthens the reaction time
to light, until reaction fails entirely at 10°-12°.

21. Gonionemata, when “bell-up” in the water, react to light fall-
ing upon them from above much less quickly than they do when
“bell-down” in position. This, together with certain results obtained
with excised margins, proves that the under surface of the margin is
the portion of the medusa especially sensitive to light.

22. Since destruction of the marginal bodies renders the margin
irresponsive to light, we conclude that these bodies are the sense
organs for photic stimuli.

23. The marginless bell is entirely irresponsive to light intensities
not greater than that of direct sunlight.

24. Excised margins in many cases react to either increase or
decrease of light intensity in much less time than do normal animals.
They differ markedly from the latter in the uniformity of their reac-
tions to decrease in light.

EXPERIMENTS IN ARTIFICIAL PARTHENOGENESIS:

By’ E.(P. LYON.
[From the Hull Physiological Laboratory of the University of Chicago.]

Be epee of American investigators have successfully repeated

Loeb’s experiments on artificial parthenogenesis. In Europe,
Delage has taken up the work with marked success. Most of those
who have worked on the problem at Naples have, on the contrary,
reported negative or indifferent results. During the past autumn I
spent some weeks at the Naples Zodlogical Station; and at the sug-
gestion of Professor Loeb, who thought that one familiar with the
material and conditions at Woods Hole might be better able to cope
with the difficulties at Naples, I made experiments in parthenogenesis
with a variety of forms and methods.

PRECAUTIONS,

In all my experiments the dishes and instruments were sterilized
by boiling. All sea-water used was heated in glass to 60°—70° C.,
and filtered. After cooling, enough distilled water was added to
make up the loss due to evaporation. The practically sterile sea-
water thus prepared was thoroughly shaken with air. The animals
used were thoroughly washed under a violent stream of tap-water.
The hands of the operator were washed in very hot fresh water.

EXPERIMENTS ON ECHINODERMS.

I. Arbacia.— Arbacia pustulata found at Naples resembles, at least
superficially, Arbacia punctulata of the American coast. Physiologi-
cally it differs from our species in a greater natural tendency of the
eggs to divide parthenogenetically without any treatment whatsoever.
As a rule, the eggs did not begin to change for from twenty to
twenty-four hours. Then they began to segment rapidly, and it was
not unusual to find, a few hours later, that 80 per cent or go per cent
of the eggs were in various conditions, from two cells to forms that
appeared to be irregular and pathological morule. I say “appeared
to be” because I made no tests to show that the loose masses were

308

Experiments in Artificial Parthenogenesis. 309

really composed of cells, nor did I watch their production from the
start. These forms were very numerous, and were similar to those
pictured by Ariola.! It will be remembered that he inadvertently
used solutions which Loeb had employed for another purpose, but
not for artificial parthenogenesis. Ariola’s failure was probably due
to this error. And the forms he figures are apparently not different
from those found in my experiments among untreated eggs, as well
as among those which had been subjected to treatment, but not to
the exact conditions for successful parthenogenetic development.
These forms easily fell apart, so that often little was left of a culture
of eggs but a mass of debris.

In some experiments in the untreated controls, a few cleavage
forms were found which appeared more nearly normal; and I often
examined my controls with extreme care to see whether swimming
larvee were produced. On two occasions two or three weak ciliated
fragments were found; never any normal larve. That normal par-
thenogenesis may sometimes occur in Arbacia pustulata seems to
me only remotely possible. Indeed, the possibility that the ciliated
fragments were not derived from the eggs at all should not be lost
sight of, although it seems to me a remote conjecture, considering the
manner of preparation of the water and eggs used. At any rate, I
am sure that accidental impregnation with sperm is not to be con-
sidered in these cases. The controls were most carefully examined
at times when fertilized eggs would be in early cleavage stages and
easily detected, and not one was found.

In spite of the above-mentioned natural tendency toward develop-
ment, artificial parthenogenesis was more difficult to produce in
Arbacia pustulata than in the Woods Hole species. There was
greater variability in the material furnished from day to day. This
may be due to the perennial breeding habit of Arbacia pustulata as
contrasted with the limited period of Arbacia punctulata. It is
harder to find females with eggs in exactly the right stage. Under
the proper conditions, however, unequivocal results were obtained.

An easy method of getting eggs unmixed with body liquids or
intestinal contents was found in Arbacia pustulata. If the animals
were held under the fresh-water tap for a minute or two (as was done
in washing them) and then left in the air for about a minute, they
shed their eggs in great quantities. With a pipette the eggs were
washed into sterile sea-water. The males, also, when similarly treated,

1 ARIOLA: Atti Societa ligure, Genoa, Igo, xii, p. 12.

310 LSPS Lyon

shed their sperm, but a somewhat longer time of exposure to the
water and air was required. The eggs thus obtained fertilized nor-
mally with the sperm, and excellent cultures of normal plutei were
obtained. This method of obtaining the eggs was used in almost all
the experiments on Arbacia. No experiments were made to ascer-
tain the nature of the stimulus which caused the eggs to be shed.
The temperature of the Naples hydrant water was noticeably low.
Cold, mechanical stimulation and the action of fresh water are pos-
sible explanations.

Hydrochloric acid has been found by Loeb to be an efficient re-
agent for causing artificial parthenogenesis in star-fish. He found
it did not succeed in Arbacia punctulata. But, strangely enough, it
is one of the best reagents I found for Arbacia pustulata. Usually
2,3, 4, 5, 6, and 7 c.c. of a solution of 7% hydrochloric acid in sea-
water were added respectively to dishes containing 100 c.c. of sea-
water. Eggs immersed in these solutions were taken out at intervals
of from two to fifteen minutes. It was found advantageous to wash
away the excess of acid by dropping the eggs from a pipette into
test-tubes filled with sea-water. The eggs sank slowly through the
water, which was then drawn off. More sea-water was added, and
the eggs poured into dishes containing about 100 c.c. Some of the
best results were obtained from 2 c.c. acid in 100 c.c. of sea-water,
ten to fifteen minutes’ exposure; 3 c.c. acid, seven to twelve minutes’
exposure; 4 C.c. acid, nine minutes’ exposure; 7 c.c. acid, five min-
utes’ exposure. In general, a short exposure to stronger acid might
be said to give equal results with longer exposure to weaker solu-
tions; but the exact duration for strong solutions was harder to
ascertain. In the best experiments perhaps ro per cent of the eggs
developed to swimming larve. Many of these swam up to the top
of the liquid, just like larvae from fertilized eggs. They formed fully
developed plutei, which lived as long as individuals produced from
fertilized eggs and kept under the same conditions.

For extracting water from eggs according to Loeb’s first successful
method for producing artificial parthenogenesis, I used potassium
chloride and sodium chloride. These two salts have been found by
Loeb! to be best adapted to this purpose, although, of course, sugar
and other non-electrolytes may be used. Potassium chloride, in my
hands, gave the best results. Solutions containing 10 to 16 c.c. of
2} potassium chloride in 100 c.c. of sea-water and exposures of

1 J. Logs: Archiv fiir Entwickelungsmechanik der Organismen, 1902, xiii, p. 482.

Experiments in Artificial Parthenogenesis. 311

one to two hours were used. These are the strengths of solutions
and approximate times used by Loeb. The Naples material is not
different markedly from that at Woods Hole, so far as parthenogene-
sis by this method is concerned. Normal plutei were produced.

While I was working at Naples, Delage! published his first paper
on parthenogenesis produced by carbon dioxide. He succeeded with
star-fish, and calls carbon dioxide the ideal reagent, but failed with
sea-urchins. I tried his method on the Naples forms. Siphons of
sterile sea-water were charged with the gas by the local mineral
water company. This water was drawn off, and after the active
effervescence ceased, solutions of one half sea-water and one half
carbon dioxide water, one third sea-water, and two thirds carbon
dioxide water, etc., were prepared. The eggs were exposed to these
solutions for various times from three minutes to several hours. Six
to twenty minutes seemed best.

The undiluted carbon dioxide solution acted like too strong acid.
The pigment of the eggs was dissolved out, and the eggs became
sticky and agglutinated. Weaker solutions caused very marked cell
division. Within three or four hours as many as 90 per cent would
sometimes begin development. But no active larvee were produced
from Arbacia eggs.

Shaking failed to produce larve from Arbacia eggs, though eggs
that had been mechanically agitated showed cell division earlier and
in larger amount than the unshaken control.

During the last weeks of my work (November) it became more
difficult to obtain artificial parthenogenesis in Arbacia; and perhaps
the material is still more difficult to work with in winter, which is
the season most workers have used it. At the same time difficulty
in obtaining natural fertilization was experienced. In some cases I
was able to cause artificial parthenogenesis in eggs which would not
fertilize with sperm. Usually, however, eggs that failed to develop
by one method also failed by the other.

2. Strongylocentrotus. — The species used at Naples was Strongy-
locentrotus lividus. It did not give good results with hydrochloric
acid, though cell division was accelerated. On the other hand, potas-
sium chloride gave better results than with Arbacia. Thousands of
fine plutei were produced in a dozen or more experiments, and some

1 DELAGE: Comptes rendus de l’académie des sciences, 1902, Cxxxv, p. 570
and p. 605. See also, Archives de zoologie expérimentale et générale, 1902, 3me
Setie, X, p. 213.

31S Lied MN eyon.

were alive when I left Naples twenty-two days after the experiment
began. The best results were obtained by adding from 10 to 14 c.c.

of 2} m potassium chloride to 100 c.c. of sea-water. The eggs were
left in these solutions from one to two hours, one and one half hours
being, perhaps, more usually satisfactory than any other time.

Sodium chloride did not give so good results as potassium chloride,
although plutei were obtained by its use.

In order to keep these or any other parthenogenetic larvze alive for
any length of time, it was necessary to pick them out from the dishes.
containing disintegrative fragments of eggs which failed to develop,
and place them in clean sea-water. But larve produced by natural
fertilization require the same care if undeveloped eggs or dead larvee
are present in any quantity.

Carbon dioxide gave better results with Strongylocentrotus than
with Arbacia, but it is far from the ideal reagent for either species.
It stimulates cell division very markedly. A few larve were pro-
duced from eggs of Strongylocentrotus exposed from two to eight
minutes to a solution consisting of one half to five sixths carbon
dioxide water and one half to one sixth sea-water. Some of these
became plutei.

Delage believes that carbon dioxide is a specific reagent for star-
fish eggs, that it does not act by its acid property, and fails alto-
gether with sea-urchins, including Strongylocentrotus. It is worthy
of notice, perhaps, that in my experiments, while carbon dioxide
solutions appeared to affect the eggs in general like an acid, yet they
produced larvee in Strongylocentrotus and not in Arbacia; while the
reverse was true of hydrochloric acid. I consider the experiments too
few, however, to be sure that this difference would always be found.

For reasons given later in the paper, I was led to try potassium
cyanide dissolved in sea-water. In four experiments with Strongy-
locentrotus parthenogenetic larvee were obtained, and some of them
became plutei. Solutions and exposures which were successful were
Experiment A, ;7,, twenty-four hours and <75p forty-eight hours;

Experiment B, 545 and 7% PWanEIO Es hours; Experiment Gye ge

400)
forty-two hours; Experiment D, #4, forty-two hours. Not many
larvee were obtained, but a very great tendency to segmentation was
noted. If eggs were left in the potassium cyanide solutions, nuclear
and cell divisions began sooner or later, even sometimes in solutions

as strong as ;%). This is remarkable when compared with my ! re-

* Lyon; This journal, 1902, vii, p. 56.

Experiments in Artificial Parthenogenesis. an

sults on Arbacia at Woods Hole. In one experiment I find a record
of many fine plutei produced in g{4p and 99600 potassium cyanide,
the eggs having remained in the solutions continuously. Unfor-
tunately no record of the control is found among my notes in this
case.

3. Spherechinus granularis. — This large species was brought in
once only, and but one ripe female was found. Eggs treated with
hydrochloric acid and potassium chloride in the same manner as
Arbacia showed a great deal of segmentation, and a few good larve
were found. The best result was from an exposure of two hours to
a solution of 10 c.c. of 24. potassium chloride in 100 c.c. of sea-
water. Shaking gave no larve, although a large amount of segmen-
tation was noted.

4. Holothurians. — Neither Holothuria tubulosa nor Holothuria
stellata was ripe. The eggs would not fertilize, although the sperm
used appeared very active. Eggs of the former were treated with
hydrochloric acid, and a few pathological swimming larve were
found. The successful solution contained 3 c.c. #4 hydrochloric
acid in 100 c.c. of sea-water, and it acted on the eggs five minutes.
This solution and others caused a considerable amount of cell divi-
sion, while the control was unchanged.

EXPERIMENTS ON ASCIDIANS.

Ciona intestinalis. — Although these animals are dioecious, Castle
has shown that the eggs are practically always cross-fertilized. I
.confirmed this observation. Moreover, if one carefully cleaned the
body wall with alcohol, and then punctured the oviduct, the eggs
thus obtained from several individuals could be mixed without get-
ting any development. As a rule, however, I kept the eggs from
different individuals separate. Hydrochloric acid, potassium chlo-
ride, and carbon dioxide, in all strengths and exposures, gave no
results. No tendency to cleavage, even, was noticed, although I
made over forty experiments, and used several different lots of
material. Mechanical agitation also failed.

RELATION OF OXYGEN TO ARTIFICIAL PARTHENOGENESIS.

The great amount of cleavage that eventually took place in dishes
of unfertilized but untreated eggs of Arbacia pustulata attracted
attention early in my experiments. In Strongylocentrotus the ten-

314 TP LYON.

dency to division is hardly less marked. By studying the conditions
under which this phenomenon took place, it seemed possible to gain
some insight into the causes of artificial parthenogenesis. Thinking
that the falling apart of the blastomeres which was so noticeable
might be due to lack of oxygen, like the disintegration observed in
my potassium cyanide experiments at Woods Hole, I tried the effect
of pure oxygen. In two sterile flat-bottomed flasks of 250 c.c. capac-
ity were placed equal quantities of sea-water and unfertilized eggs
of Arbacia. The layers of liquid were less than one centimetre deep.
Through one of the flasks a strong stream of well-washed and purified
oxygen was passed for some minutes, and thereafter a slow stream.
The gas did not pass through nor mechanically agitate the liquid.
The other flask was covered with a small inverted beaker, and there-
fore exposed freely to the air. At the end of twenty-four hours,
and at intervals thereafter, eggs were removed from each flask:and
examined.

The result was the reverse from what I had expected: division in-
variably began much sooner and went on to a much greater extent in
the air-filled flask than in the oxygen-filled flask. The eggs in the
oxygen remained in good condition and capable of fertilization after
those in air had gone to pieces. I did not attempt to find limits;
but in one experiment, for example, eggs kept in oxygen were fer-
tilized after ninety hours and produced plutei, whereas those kept
in air had all fallen into formless debris in less than sixty hours.
While Arbacia unfertilized eggs in air showed a large percentage of
cleavage in from twenty to twenty-four hours, those in pure oxygen
often failed to show division in one egg in a thousand at the end -
of three days. These experiments succeeded equally well with
Strongylocentrotus.

It would be impossible, or at least very difficult, to get the eggs of
sea-urchins (especially Strongylocentrotus) wholly free from bacterial
infection. Sooner or later all my cultures became putrid. This
always happened earlier in the air-filled than in the oxygen-filled
flask. One might, therefore, be inclined to attribute the earlier
death of the eggs in air to bacteria. The truth is, rather, that the
earlier presence of large numbers of bacteria in the air-filled flask
was due to the earlier death of the eggs. Not until some time after
the eggs had begun to fall apart in large numbers could bacteria be
detected. The dead eggs furnished a good culture medium for the
few germs present in the original preparation.

Experiments in Artificial Parthenogenests. 315

It appeared, therefore, that lack of oxygen tended to induce division
and development in sea-urchin eggs. It was this idea that led me
to attempt to produce parthenogenesis by the use of potassium
cyanide, as described earlier in this paper. The reasonable degree
of success which attended these experiments pointed in the same
direction. But whether a wider application of the method is possi-
ble, and whether all the means of producing artificial parthenogenesis
really affect. the respiration of the cells and produce a condition of
lack of oxygen are matters of merely interesting conjecture at the
present time.

Loeb! noticed that fertilized eggs of Fundulus, Ctenolabrus, and
Arbacia in a current of hydrogen might segment a little before the
control in air. I? found a slight hastening in the development of
fertilized Arbacia eggs in very weak potassium cyanide solutions.
Mathews? was able to start the development of Arbacia eggs by
alternate exposures to hydrogen and air. These observations, to-
gether with those of Delage with carbon dioxide, and the action of
anzesthetics observed by others, may eventually find a single expla-
nation. Delage, however, expresses the belief that carbon dioxide
(as also other agents which cause artificial parthenogenesis) does
not act by causing asphyxia, but by arresting protoplasmic activity
in some other way.

RELATION OF TEMPERATURE TO PARTHENOGENESIS.

Loeb and others have noticed that temperature is an important
factor in artificial parthenogenesis. A variation of a few degrees
seems to make a great difference in the result. Few exact experi-
ments, however, have been made. It is imaginable (1) that the
temperature of the reagent to which the eggs are exposed is the
important factor; or (2) that the temperature of the sea-water to
which they are returned and left to develop is the deciding element.
The few experiments which I made seemed to indicate that the tem-
perature of the solution is of paramount importance. The temper-
ature of the sea-water in which development takes place is also
important, but chiefly so (between 16° and 22°) in affecting the
rate of development. The experiments were too few, however, to

1 J. Loes: Archiv fiir die gesammte Physiologie, 1893, lv, p. 530.
2 Lyon: This journal, 1902, vii, p. 56.
3 MATHEWS: This journal, 1900, iv, p. 343.

316 EP. Lyons

permit of more than tentative conclusions. But it seems to me the
experiments speak against Delage’s idea that the action of the solu-
tions is inhibitory; that the essential factor in producing artificial
parthenogenesis is a temporary suspension of activity. Some change
is going on in the eggs during the time the reagent acts. The
change is of such a nature that cell division can go on easier and
faster afterwards. The change is slowed by lowering the tempera-
ture. That an active process is concerned is also indicated by such
experiments as those I have described in Arbacia, where the time
varied inversely as the strength of reagent used; e. g., hydrochloric
acid. Other investigators have announced similar observations.
I will describe some experiments on temperature : —

Experiment 119.— Three dishes were prepared, each containing roo c.c. of
sea-water + 14 c.c. of 24 m KCl solution. ‘These dishes were kept re-
spectively at 17°, 20°, and 24°C. The mixed eggs of three specimens of
Strongylocentrotus were distributed to these dishes. Eggs were removed
after exposures of 3, 1, and 2} hours, and placed in sea-water at 20°.
The half-hour exposure gave no results. ‘The one hour exposure at 24°
gave a very few larve; at 17° and 20°, no results. The 2} hour ex-
posure at 17° and 20° gave numerous plutei; at 24°, one fragmentary
blastula only.

Experiment 119 a.— Same as preceding, except that eggs after exposure to
the solution were kept in sea-water at 17°. The exposures at 24° gave
no results. Exposures at 17° and 20° for 2} hours gave numerous larve,
which developed more slowly than those similarly treated in Experiment
119. The plutei formed were, perhaps, less uniformly normal. But
some of them lived 18 days.

Experiment 123.— Two dishes for each of the following concentrations: —
100 ¢.c. sea-water + respectively 12 C.c., 14 C.c., 16 GG. Of 25 meer
One dish of each concentration at 19°; the other at 23°. Eggs of
Strongylocentrotus exposed 1, 14, 14, 13, 2 hours, then placed in sea-
water at room temperature, about 19°. Exposures of 1% and 1# hours
to the 14 c.c. and 16 c.c. solutions at both temperatures gave abundant
larvee. Rut 23° was much more favorable. Some of these formed plutei

and lived 22 days. The 12 c.c. solutions gave no larve.

Experiment 131. Three dishes for each concentration: 100 ¢.c. sea-water +
respectively 14 and 16 c.c. 2} m KCl. One dish of each concentration
kept at following temperatures, respectively, 15°, 19°, 23°. Eggs of
Strongylocentrotus exposed 14, 13, 13, 2, and 4 hours. Removed to

Experiments in Artificial Parthenogenests. aby

sea-water at room temperature, approximately 19°. ‘Two hours after
removal from the solutions one third to one half of those eggs which had
been at 23° in either concentration for 1} to 2 hours were in cleavage.
Often as many as ten or twelve cells had been formed. In many eggs
several nuclei were visible, although division of the cytoplasm had not
taken place. Some of those exposed at 23° for four hours divided within
forty minutes into ten or twelve cells. But none of the four hour speci-
mens formed larve. The eggs treated at 19° and 15° were far behind
the 23° eggs in division. By far the best and most numerous larvee were
from those eggs exposed at 23° for 14 or 1} hours to the solution of
16 c.c. 24 m KCl in too c.c. sea-water.

The temperatures used by Greeley! and myself were so different
that comparison of our results is of little value. My experiments do
not uniformly bear out his conclusion that a long exposure at a low
temperature gives results equal to, or better than, a shorter exposure
at a higher temperature. But the temperatures used by me were not
far apart, and the optimum time element introduces a factor of un-
certainty. Unless the eggs are given, at each temperature, the length
of exposure best adapted to it, comparative results are of no value.
And the observer cannot be sure he has found the optimum time
unless eggs were taken out at very frequent intervals. I did not
always do this; and some of my experiments entirely support
Greeley’s view. I am therefore undecided as regards Strongylo-
centrotus, although before I saw Greeley’s paper I was inclined to
think 23° more favorable than 19° or 16°, irrespective of time.

SUMMARY.

1. The eggs of the sea-urchins at Naples may be caused to develop
(during September, October, November, at least) by approximately
the same means used by Loeb at Woods Hole.

2. In a few instances larve of Strongylocentrotus were obtained
by the use of carbon dioxide.

3. In four experiments larvae of Strongylocentrotus were obtained
by the use of potassium cyanide.

4. The pronounced tendency of Arbacia pustulata and Strongy-
locentrotus lividus unfertilized eggs to segment after about twenty
to twenty-four hours can be overcome by exposing the eggs to pure
oxygen.

1 GREELEY: Biological bulletin, 1903, iv, p. 129.

S15 a) E. P. Lyon.

5. The temperature of the solution to which the eggs are exposed
to induce parthenogenesis is very important. The temperature of
the sea-water to which the eggs are returned for development is
probably less vital, although the rate is slower at low temperatures.

6. All efforts to cause the parthenogenetic development of Ciona
intestinalis failed.

THE RELATIONSHIP BETWEEN THE FREEZING POINT
DEPRESSION AND SPECIFIC GRAVITY OF URINE,
UNDER VARYING CONDITIONS OF METABOLISM,
AND ITS CLINICAL VALUE IN THE ESTIMATION
OF SUGAR AND ALBUMIN.

BY eNGy El, A SCLOW ES ii: I):
[Gratwick Research Laboratory, University of Buffalo.]

CONTENTS.

Page
Introduction. . . : Air bel ak tee. uch x OS
‘Specific gravity and de nreacion of fecane Som of rion (AVES 5c 5 5 yall
Specific gravity and depression of freezing point of the urine of normal srichovid Wels
under abnormal circumstances, and of pathological urines other than those
containing sugar and albumin . . ; sees nd og | Se
The average molecular or ion weight of the sibeeniees deceived 3 murine « <4 4 1326
Influence exerted by individual ions 328
Therapeutic treatment with chlorides . Oe
Analytical methods based upon the fact that thera average ve Wel Bit af the i ion in urine
is slightly less than 60 . 330
Estimation of sugar in diabetic urines : swe 330
Estimation of sugar in artificially Eas sugar urines 335
Estimation of albumin 338
Analysis of urines containing both sugar a aioe 341
Summary . 342

INTRODUCTION.

Is the course of his work on the theory of dilute solutions Van’t
Hoff developed the relationship existing between their chemical
constitution, osmotic pressure, and freezing point. Raoult subsequently
made use of the freezing point of solutions as a direct means of deter-
mining the molecular weight of the dissolved substance, and it is to
this investigator and Beckmann that we are indebted for the simple
apparatus now universally employed for this purpose. For many
years Beckmann’s apparatus has been in constant use in most physi-
ological chemical laboratories for determining the freezing point of

1 This paper was read before the Buffalo and Niagara Falls Chemical Society
in March, 1903.

319

320 GTi A. Clowes.

blood, urine, etc. Unfortunately these determinations have usually
been made with a purely practical object in view, the comparison of
osmotic pressures, very little attention being paid either to Van’t Hoff’s
theoretical generalization of Avogadro’s law upon which the method
itself is really based, or to the valuable deductions which may be
drawn by this means regarding the nature and proportions of the in-
dividual substances producing the osmotic effect.

Van’t Hoff’s and Raoult’s conclusions may be expressed as follows :

1. The osmotic press:ire exerted by any substance in solution is
independent of the nature of the molecules of that substance and
simply proportional to their number. And this pressure corresponds
with that which would be exerted by an equal number of molecules
of any ideal gas under the same physical conditions.

2. The lowering of the freezing point of any solution below that of
the solvent employed is directly proportional to the osmotic pressure
and consequently to the number of molecules dissolved, regardless of
their nature. Or, expressed in another form, solutions of different
substances in the same solvent, dissolved in such proportions as to
have the same freezing point, are isosmotic and equimolecular.

It follows from these laws that the average size of the molecule or
dissociated ion present in a given physiological solution may be esti-
mated as directly proportional to the weight of substance dissolved in
a given quantity of fluid, and inversely proportional to the lowering
of the freezing point (or osmotic pressure) of the solution. This fact
has been expressed by Raoult and Beckmann in the form of a simple

equation ;

ee ee
A

where JZ is molecular weight of substance in question, to be deter-
mined; / is a constant having the value of 18.5, when water is em-
ployed as the solvent; sz is the known weight of substance dissolved
in 100 c.c. of solvent, and A is the observed depression of the freez-
ing point in degrees C.! )

Hoping by this method to throw some light upon the nature of the
disturbances in metabolism taking place throughout the body in the

* When, as in the case of urine, the exact weight of dry solids cannot be accu-
rately determined without considerable loss of time, an approximate value of 7
may very readily be obtained by multiplying the specific gravity of the solution,
taken at 15° C., by the factor 2.33; thus for #z we may substitute specific gravity
multiplied by 2.33.

Experiments on Urine. 321

course of such diseases as cancer, diabetes, etc., we have made a prac-
tice of determining the freezing point of stomach contents, blood,
transudates and urines, both as a whole, and also in separate fractions,
in those cases in which allowance could be made for the influence
exerted by the fractionating agents employed. The specific gravity
as well as the proportions of the various chemical constituents of the
fluid in question were determined, as a matter of routine, in the course
of our experiments on metabolism; and the relationship obtaining
between these quantities and the freezing point was calculated. This
paper will be especially devoted to the consideration of certain rela-
tionships existing between the freezing points and specific gravities
of normal urine; and further, to deviations from the normal observed
in diabetic urines, which afford a simple means of estimating the
sugar which they contain.

SPECIFIC GRAVITY AND DEPRESSION OF FREEZING POINT OF
NoRMAL URINES.

The lowering of the freezing point in the normal urine varies within
very wide limits, being directly influenced by the consumption of
water in the same manner as is the specific gravity. In fact, on com-
paring the freezing point directly with the specific gravity, a constant
relationship is found to obtain which only varies within such narrow
limits as almost to fall within the range of experimental error.!

This will be seen from an examination of Table I, representing a
series of experiments made from day to day on the writer’s urine, and
Table II, a series of determinations on the urines of normal individuals,
the only regulations observed being a full normal diet and restriction
within reasonable limits in the consumption of various salts, medica-
ments, and alcoholic stimulants.

1 This same observation has been made by Fucus, and the result published in
the Zeitschrift fiir angewandte Chemie, in October, 1902, an abstract of his article
appearing in the Biochemisches Centralblatt, January, 1903. Were it not that
his theoretical conclusions are open to criticism, and that those phases of his work
bearing directly upon the estimation of sugar in diabetic urines leave much to be
desired, the publication of these results at this early date would be unnecessary.
Realizing that the empirical factors published by Fuchs needed revision, lest they
bring an excellent method of analysis into disrepute, we felt that the immediate
publication of this portion of the work (apart from the complete studies in metab-
olism referred to above) would be justified.

322 G. H. A. Clowes.

In these tables the first column represents the specific gravity of
urine determined at 15° C.; the second column gives the depression
of the freezing point of the urine in question below that of water, and
is indicated by A; the third column, represented by A’, gives the cal-
culated freezing point derived from the decimal portion of the specific
gravity by means of the factor 75, the source of which will be explained
later; the fourth column shows the relationship existing between S,
the decimal portion of the specific gravity determined at 15° C., and
A, the observed freezing point depression. An inspection of these

IAB CES a

| F. P.
Spec. grav. at |
SAG:

——————SSSSa
Observed Calculated

1.87
1.02
1.485
167
1.38

o7he,

Average

>
figures makes it appear evident that Vike constant, about 0.01337.

In both cases the average of a large number of determinations is
found to be about the same, and the variations are, in no case, more
than 4 per cent from the mean. Expressed in another form, each
0.01 of specific gravity is found to be approximately equivalent to

a lowering of the freezing point of ae =0.75 C. It will nawebe
3
seen from what source the factor 75, made use of in the third column
and referred to above, is derived. Each 0.01 of the decimal portion
of the specific gravity of a normal urine is equivalent to 0.75° C.
lowering of the freezing point, and the theoretical freezing point of
any urine may be determined by multiplying the decimal portion
of the specific gravity by the factor 75. For example, in the first

Experiments on Urine. 323.

ease on Table I, the specific gravity = 1.0252; then 0.0252 x 75 =
1.89, a figure which compares very closely with the observed freezing
point 1.87. In a second case, taken from Table II, No. 9, specific
gravity is 1.0202, the observed freezing point 1.51°; 0.0202 X 75 =
1.51, which exactly coincides with the observed freezing point. These

TABLE II. NORMAL INDIVIDUALS.

lig 126

7 CERT Ee a a
Sa a Ey soanad Calculated

2
=

1.0126 | 0.0128
1.0114 | 0.0129
1.0287 1 0.0131
1.0255 0.0130
1.0248 | 0.0132
ole | 0.0132
1.0132 0.0135
1.0242 0.0133

1.0202 5 0.0134
1.026 | 0.0134
1.0259 ‘| 0.0135

1.0255 | 0.0137
1.0225 | 0.0138
1.0225 0.0138
1.0219 : 0.0140

INCRE. oe a vee os ho 0 01337

~

same factors are found to hold good, although within slightly wider
limits, in the case of normal individuals under abnormal conditions,
and in the urine of patients suffering from various diseases and under-
going varying treatments, provided sugar and albumin are absent.
This phase of the subject will be dealt with in the next section.

324 G. H. A. Clowes.

SPECIFIC GRAVITY AND DEPRESSION OF FREEZING POINT OF THE
URINE oF NORMAL INDIVIDUALS UNDER ABNORMAL CIRCUM-
STANCES, AND OF PATHOLOGICAL. URINES OTHER THAN THOSE
CONTAINING SUGAR AND ALBUMIN.

It is not proposed at this stage to enter into a minute study of the
chemical constitution of various urines as compared with their freez-
ing point and its deviation from the theoretical. That phase of the
subject will be discussed ina later section. It is, however, interesting
to note that the increase in the proportion of chlorides and other salts
having smaller ions than the average in urine, produces less effect
upon the freezing point than would be anticipated. It is even more
remarkable that pathological urines, as, for example, those of cancer
patients in the last stages, in which the proportion of normal constit-
uents is seriously interfered with, and the chlorides have practically
disappeared, should retain their normal freezing point within very
narrow limits, as will be seen from an inspection of Tables III and
IV, cancer cases.

The same close agreement was obtained in the urine of a typhoid
fever case which for several days was almost free from chlorides and
contained as much as 3 per cent of urea.

DABBLE UI CANCER CAS naiae

Fake

Spec. gray. at ——————————
LSC. Observed Calculated
A. f

Table IV gives results obtained during the last three weeks of his
life, from the urine of a patient suffering from cancer of the stomach.
This urine contained not more than one-tenth to one-twentieth of the
normal proportion of chlorides, and yet, as will be seen, the average

S
value of x does not vary to any considerable extent from the normal.

Lixperiments on Urine. 325

TABLE IV. CANCER CASE W.

BoP.

Spec. grav. at ee
15oee Observed Calculated
A. 4

1.020

1.0198

1.019 1.41
1.0125 0.92
1.0162 1.19
1.0135 - 1.00
1.0152 1.07
1.0121 0.91

1
2
3
4
5
6
i
8
9

1.0117 0.92
1.0124 0.87
1.0141 1.04
1.0103 0.79

Mean

The urine of patients before and after epileptic seizure has also been
examined in a considerable number of cases. The freezing point has
been found to remain fairly constant within reasonable limits. We
have also included the urine of animals in our series of experiments.
Table V gives the results obtained in two cases from a series of four
guinea-pigs, from which it will be seen that a fair uniformity exists
amongst the individual animals, although the difficulties associated
with the accurate determination of the specific gravity lead to a much
larger experimental error in the case of animals’ urine, where the nor-
mal specific gravity is from one-third to one-half that of human beings.

It will be seen from a consideration of the results enumerated in
this section, that a fairly constant relationship is found to obtain be-
tween the specific gravity and freezing point, even under the most
abnormal conditions, and in urines of widely varying chemical compo-
sition. On the other hand, no constant proportion can be traced
between the chlorides, or any other chemical constituent either of

326 G. H. A. Clowes.

normal or pathological urines, and the freezing point. Merely the
sum total of effects is a constant, the individual components varying

TABLE V. URINE OF EXPERIMENTAL GUINEA-PIGS.

F. P.

Spec. grav. ———————
(approx.) Observed Calculated Difference.
atalscic:

1.009
1.012

1.010
1.011

1.010
1.011
1.011

1.0105

within wide limits. The actual influence exerted by various constitu-
ents of the urine, and the way in which the proportions of these sub-
stances are adjusted, will be considered briefly at a later stage.

THE AVERAGE MOLECULAR OR JON WEIGHT OF THE SUBSTANCES
DISSOLVED IN URINE.

From the observation referred to above, that in normal urine the
depression of the freezing point is about 0.75° for each o.o1 in the
decimal of the specific gravity, it is comparatively simple to calculate
the average molecular weight of the individual ions present in urine
from the formula referred to above;

M= 18.5 —

if we assume that each 0.01 in the specific gravity determined at 15°
is equivalent to 2.33 per cent of normal solids in the urine! Then,

‘ A large number of determinations of the dry solids in urine have been made
by the writer from time to time, which have proved very conclusively that the
factor 2.3 to 2.4, as commonly accepted, does not always represent the correct
figure, but merely an average. Since, however, work of the nature described above

Lixperiments on Urine. ae

2.33
7 — 5.5 See ye5:

We have several times in the course of our work compared the
average molecular weight as obtained from the specific gravity, with
that obtained from the total weight of solids found on evaporating the
urine. The average of a large number of determinations is practically
the same, but the figures obtained, making use of the specific gravity,
are much more uniform than those obtained from the weight of solids

‘direct. In fact, it may be said that the factor a= = is much more

total solids

EP:
sible explanations for this: the first is the difficulty of determining total
solids correctly, and the second is the fact that urea and salt (NaCl),
the two most important constituents of the urine (so far as specific
gravity, total solids, and freezing point are concerned), whilst possess-
ing ions differing widely in their relative average weights of 60 and
30, also exert a very different effect upon the specific gravity, but in
the inverse direction, such that aqueous solutions of urea or sodium
chloride possessing the same specific gravity contain more nearly equal
quantities of ions, than do solutions of these substances of equal per-
centage concentration. The most important factors in determining
approximately the specific gravity and freezing point of urine, are:
(1) the percentage of urea, chlorides, and other readily dissociated salts
producing a very marked effect upon the freezing point, and (2) the
percentage of those residual organic constituents of the urine, such
as uric acid, creatinine, pigments, and other organic substances which
may be looked upon as practically inert so far as effect upon the
osmotic pressure or freezing point is concerned.

Since such a constant relationship appears to obtain between the
specific gravity and freezing point, it must be assumed that a con-
stant equilibrium is maintained between these two types of urinary
constituents. Since the average molecular weight is slightly less

nearly constant than is the factor There are two pos-

is not absolute, but merely relative; and since the principal endeavor is to reach
an average, and that within a fairly wide range of experimental error, it has been
decided to use the factor 2.33 when dealing with normal urines. As will be shown
later, it is necessary to employ 2.8 as the factor for sugar solutions, and 3.8 to 4
for those of albumin, so far as these individual constituents are concerned. It
will also be seen that in the estimation of sugar by this method referred to later, it
is a matter of indifference whether we employ the factor 2.33, 2-2, or even 2.0,

328 Gs Tae al CLO,

than 60, and the average of the electrolytes lies considerably below
that figure, it must be assumed that some direct relationship exists
between the proportion of these bodies and of the non-conducting,
large molecular complexes present in the urine. This question will
be dealt with in the next section.

Disregarding the total weight of solids and specific gravity, we have
in several cases calculated the probable average size of the ion in a
series of normal urines, allowing for the effect exerted by each indi-
vidual constituent at the concentration in question, and have invariably
obtained figures lying between 50 and 60, usually about 55. It will’
be impossible to present the data bearing on this question at present,
as the material involved would carry us too far from the main object
of this paper, the analysis of sugar urines, etc., dealt with in a later
section.

THE INFLUENCE EXERTED BY INDIVIDUAL IONS.

As stated in the last section, the first important factors to deter-
mine are: (1), What proportion of the total nitrogen of the urine is
present in the form of urea; (2) The variation in the proportion of
chlorides from the normal, and (3) The amount of inert or non-dis-
sociated material. The variations in the amount of chlorides in urine
is very marked. Incertain normal cases, the total excretion of sodium
chloride may reach a figure equal to that of urea. In the last stages
of cancer, in typhoid fever, etc., when the proportion of chlorides in
the diet is considerably restricted, and in addition the body shows a
tendency to retain those salts, the excretion of chlorides may fall to
one-twentieth part of the normal or even entirely disappear. Such
variations in the proportion of those bodies having the smallest ionic
weight should, if other constituents remain normal, lead to a very
marked variation of the relationship between specific gravity and
freezing point. We must, therefore, assume that some condition of
equilibrium actually exists within definite limits between the propor-
tion of salts possessed of small ions and organic bodies possessed of
a large molecular complex. It is true that in such pathological con-
ditions the amount of ammonia in proportion to the total nitrogen is
increased two and three times the normal, but the total quantity is,
in any case, too small to account for more than a small portion of the
effect produced by a diminution in chlorides; and, in such a case as
that dealt with in Table IV, the ammonia would be practically

Experiments on Urine. 329

counterbalanced by the uric acid increase. The phosphates and
sulphates, whilst varying within fairly wide limits, possess, on the
average, larger ions than the chlorides, and do not, under normal
circumstances, constitute nearly so large a proportion of the total.
The influence which they exert, after allowing for their organic com-
binations, does not vary to a very considerable extent.

It is not our intention to deal with this question from a theoretical
standpoint in this paper, as any ideas advanced in this direction must
necessarily be purely speculative; but we consider it highly probable
that there is a tendency toward the maintenance of a definite state of
equilibrium between the various organic excretory products and the
simple salts, such as chlorides, sulphates, etc., which may account for
the non-removal from the body, in the course of certain; diseases, of the
products of sub-oxidation and impaired metabolism when an insuffi-
cient supply of salts is excreted with the urine. These bodies accu-
mulate in the system, and may, in their turn, retain, associated with
themselves, those quantities of salts which are introduced with the
diet. This phase of the subject, whilst of considerable importance,
cannot be dealt with in this paper.

THERAPEUTIC TREATMENT WITH CHLORIDES.

Considerations of this nature may, however, be looked upon as a
further argument in favor of the addition of sodium chloride to the
milk diet of patients suffering from fevers, cancer, etc., a procedure
frequently resorted to by clinicians, as the result of practical experi-
ence. This course is advocated by Hatcher and Sollman,! in a paper
dealing very completely with the effect of diminished sodium chloride
excretion in typhoid fever. Unfortunately, these authors have not
recorded the specific gravity of the urine in their tables, and since
they themselves consider the figures referring to total solids as in-

accurate, the factor, , derived from their work

sp. gr. total solids
is very far from a constant. There is, however, an indication that the
large excess of sodium chloride which they administered brought about
a disturbance in equilibrium of the system in a direction opposite to
that exhibited in the total absence of chlorides, and it appears prob-
able that the best state of equilibrium was maintained when not more

1 HATCHER and SOLLMAN: This journal, 1902, vol. viii, page 139.

330 G. H. A. Clowes.

than five or six grams of sodium chloride were excreted in the space
of twenty-four hours. We must, however, leave this subject for a later
paper dealing with disturbances in chloride equilibrium in general.

ANALYTICAL METHODS BASED UPON THE FACT THAT THE AVERAGE
WEIGHT OF THE ION IN URINE IS SLIGHTLY LESS THAN 60.

In diabetic urine, for example, we have to deal with a solution con-
taining an admixture of normal urinary constituents having an average
molecular weight of about 60, and sugar molecules having an average
weight of 180. The estimation of sugar in diabetic urine, based upon
this relationship, will be dealt with at a later stage. The estimation
of albumin present in solution in urine, dependent upon the fact that
the molecule of albumin is so large as to produce a practically negligible
effect upon the freezing point, will be treated in Section 11.

Estimation of sugar in diabetic urines. — Since the average mo-
lecular weight of the normal substances present in urine is 57.5,
and that of sugar is 180, it is obvious from a consideration of Van’t
Hoff’s generalization referred to at the commencement of this paper,
that the effect produced upon the osmotic pressure of any solution
by a given weight of sugar would be slightly less than one-third (that

is to say, in the proportion of 75) of that exerted by an equal

weight of normal urinary solids. Now, if the effect exerted upon the
specific gravity of solutions due to the introduction of a given weight
of sugar be the same as that due to the solution of the same
weight of normal urine constituents, the problem involved would be
comparatively simple; but this is not the case.
_ As the result of a series of careful determinations made in the
Laboratory, we find that the proportions are as six to five; that is
to say, six grams of sugar dissolved in one hundred of water would
produce a solution having about the same specific gravity as five
grams of ordinary urine solids in one hundred of water. Conse-
quently the factor 2.33 cannot be applied to sugar-containing urines
in order to determine the weight of substance dissolved. We have
to make use of another factor for the sugar, which is about 2.81.
(See note, page 326.)

Before making the first freezing point determination on diabetic
urine, the following calculation was carried out. If the freezing point

Experiments on Urine. 200

of the diabetic urine were first determined, and then the theoretical
freezing point of the same urine were deduced from the specific grav-
ity, making use of the factor 75, as though it were normal urine, the
difference between these two freezing points would be due to the
relatively larger size of the sugar molecules as compared to the nor-
mal urine molecules. Whilst exerting slightly less effect upon the
specific gravity than would an equal weight of urine solids, the sugar
should merely exert one-third of the effect upon the freezing point
that would normally be expected from urine constituents.

If the specific gravity were influenced to the same extent by sugar
and normal urine solids, since the molecule of the latter is about one-
third of the sugar molecule, each sugar molecule would be represented
in the freezing point on the normal urine basis, only to the extent of
one-third of its weight, the remaining two-thirds being accounted for
by the difference between the observed and theoretical freezing point,
and one-half of that figure would consequently be equivalent to the
number of sugar molecules present in the solution; from which it
will be seen that it would only be necessary to multiply the difference
between the theoretical and observed freezing points of such urine by
the factor 5, in order to determine the percentage of sugar in the
solution.

The effect exerted by sugar and normal urine solids upon the
specific gravity is, however, not the same. As stated above, five
grams of the normal solids produce about the same result as six
grams of sugar, so that if the normal factors made use of in calculating
the solids of urine be employed, the molecule of sugar would appear
as merely 2 of 180=150. Since this 150 functions in the same
manner as 574 of the normal urine constituents, the difference, 150 —
573 = 924, is unaccounted for and is represented by the difference
between the theoretical and observed freezing point; and in order to
determine the percentage of sugar we should multiply the difference

573

CB

in question by the factor which is about 0.6. A simple mathe-

matical derivation of the factor 6 employed in our work is included in
the form of a footnote.!

1 0.75° of lowering of the F. P. is equivalent to 0.01 of specific gravity of urine
when the latter is induced by normal solid constituents of urine without the admix-
ture of sugar or albumin. It may be calculated and also shown by experiment,
that 0.75° is equivalent to 0.026 of specific gravity due to sugar in any dextrose-
containing solution; consequently if we represent the decimal portion of the

332 G. Hf. A. Clowes.

We have since made a large number of determinations of the sugar
in urine by this method, multiplying the difference between the theo-
retical and observed freezing points by the factor 6, and have com-
pared results thus obtained with the analyses by means of polarization
of light, fermentation of sugar, loss of specific gravity due to fermen-
tation, reduction of Pavy’s solution, etc. This method has shown
itself extremely accurate in almost all cases and no special precautions
have been taken at any time. It has been our endeavor, so far as
possible, to carry on analyses under those conditions which obtain in
the ordinary clinical laboratory, so that it may not be said that this
method is one which can only be made use of where special physico-
chemical corrections for errors may be introduced.

The following tables, VI, VII, and VIII, refer to the urine of patients
suffering from diabetes; the first column giving the specific gravity,
the second the observed, and the third the calculated freezing point ;
the fourth gives the difference between these latter, and the fifth the
percentage of sugar calculated from the above by means of the factor
6; the sixth column gives the determination of the sugar by polariz-
ation of light, and the seventh any check determination that may have
been made by some other dependable method; the eighth gives the

é

S : E :
value of x’ which varies considerably from the normal (0.0133), in

these cases.

specific gravity of the urine as S and the fraction of that specific gravity due to
the action of sugar as 1, then,

S ° é
1. 0.01 0:75° = theoretical F. P:
A S~ > 0.75° = theoretical F. P. of luti
, ers 750 — a
0.026 7s neoretical F. P. of a pure sugar solution.
2 Ser ee SNe) Vee nell
-: (sax 1 0.01 ) 0.75° = the observed F. P.

Subtracting equation 3 from equation 1,

eu Zo) ((Ditioin bea)
So ge 1200 ‘

and since the weight of sugar in 100 c.c. = the portion of the specific gravity due
to sugar, multiplied by 280,

: imal Die Toy 1, 1 sen
Weight of sugar % = dua a Les = 6 times Diff. in F. P.

Experiments on Urine. G22

TABLE VI. DIABETES CASE A.

1 ea Be
—$<_—__——. Sugar
Ob- Cal- Sharer Seale oe | Tosetie
served. culated. : See | spe fr.

per cent. per cent. per cent.

1.50 2.62 é 6.7 6.8

—

a2 2.50 6.5 6.3

1.565 2.07 7.2 Ve
1.325 2AT 6.9
1.475 2.82 8.1

1.86 2.91 6.3
1.94 2.81 Se
1.90 2.63 3 4.4
1.835 We 2) 308,
1.865 2.45 L 39
1.845 2.475 3.8

2
3
4
5
6
7
8
9

SS = =
no K SO

1.72 2.49 4.6
1.70 Aad 4.7
1.90 2.83 5.6

—
w

—
Tuy aS

1.89 2.56 4.0
1.70 Bai ass 6.2
1.82 2.82 ‘ 6.0

—
On

_
~I

334 G. H. A. Clowes.

TABLE VIL “DIABETES CASE: B.

z A : Sugar
c a eT ea
Day) | Peee4 On. "Cae a eae by ae loss in

5° C.
bee served. culated. Sp. gr.
per cent. per cent. per cent.

1.0376 .605 2.82 ue 7.3 7.4

1.0315 : 2.36 5.8 5.9
1.0265 : 199 Sell Sh
1.0271 : 2.03 3.5 33
1.0221 toile 1.66 Neff 2.0
1.0285 6. 2.14 3.0 2.9

1.0363 . Dh : Ded
1.0372 : 2.79 Shi)

1.0376 : 2.82 : 6.5

TABLE VIII. DIABETES CASE, CG:

ft, Eh
ee : Sugar
eee et at Ob- Cal- Difference. ie ee oe loss in
: served, culated. : hae a Sp. gr.

per cent. per cent. per cent.

1.0275 1.29 2.06 4.6 5.0
1.0247 1.40 ee 2h

1.018 1.247 aos 0.6 0.6
1.0225 1.355 2.0 2.0

1.0235 W225 Se 3.0
1.0165 1.089 0.9 0.6
1.0138 0.963 0.5

1.0126 0.965 Sace nil
1.0192 1.49 5050 nil

1.0204 1.205 23 1.9

1.0199 NEAL : 2.0

1.0168 1.26
1.0093 0.748

1.0166 1.055

1.019 1.055

Experiments on Urine. 335

As an example of the method of calculation, taking the first deter-
mination in Table VI, the urine having a specific gravity of 1.0349
had a freezing point 1.50° below that of water.

The decimal portion of the specific gravity,

0044906 75. 2.62-~

The difference between theoretical and observed freezing points,
3.62, —~ 150.1, 3,
1.12 X 6 = 6.72 per cent sugar.

The result obtained by polarization was 6.8 per cent. A second
example, the first on Table VII, showed a specific gravity of 1.0376
and an observed freezing point of 1.605”.

The decimal portion of the specific gravity,

D.0370" 75 = = 2752.
Oz — 1.005 —= 1.225,
22h <6 = 7.35 per cent of sugar.

In this case the polarization method gave 7.3 per cent, and the loss
of specific gravity by fermentation 7.4 per cent.

Estimation of sugar in artificially prepared sugar urines. — In order
to confirm these results, I prepared some artificial sugar urines, by

TABLE IX.

Bee

Sugar Spec. grav. at Ol eee , Sugar Sugar by
added. eG: a a | diff. X 6. | rotation.

served. culated. |

per cent. per cent. per cent.
0 1.0239 V9 : i nil nil

Diff., 0.0073 Diff., 0.215

2.03 1.0312 : ms Mas 2.01 2.0

Diff., 0.0073 iff

4.06 1.0385 : f ; 4.08 4.0

Sugar
Spec. grav. at ae Sugar | Sugar by ae
1c: Ob- Cal Seats Cte OalkOtacion: by fer-

served. culated. mentat’n.

per cent. per cent. per cent.
nil nil nil

3.0 Syl 3.1
EY, 59 5.8

336 G. H, A. Clowes.

dissolving varying proportions of sugar in normal urines of known
specific gravity and freezing point. In Tables IX, X, and XI, three
such series of experiments are given.

TABLE XI.

Sugar |Spec. grav. at Ob- Cal- : Sugar Sugar by

added. L5oNE: Space: oalareal diff. X 6. | rotation.

per cent. per cent. per cent.
0 AAS : nil nil

2.65 2.05 2.73
4.90 2.295 4.83

In order to facilitate the work and save unnecessary correction,
those urines were chosen which possessed an exactly normal freezing
point. The effect upon the specific gravity and freezing point on
adding two and four grams respectively in the first series, and three
and six grams in the second series, was carefully observed. It was
found that the effect exerted upon the specific gravity by each 1 per
cent of sugar added was about equal to 0.00355 to 0.0036, whilst the
influence exerted upon the freezing point varied between 0.100 and
0.107, the theoretical being 0.103. Thus, the average effect upon the
freezing point was so nearly equal to the theoretical that the molec-
ular weight of sugar might be calculated by means of the formula

M= aT RO) 180. (See Table IX.)
A 103

The difference between the observed and calculated freezing point
for 2.03 grams of sugar added, equal 0.335°, which, multiplied by

= 2.01 per cent of sugar, the rotation showing 2 per cent. On
addition of 4.06 grams, the effect upon the freezing point was equal to
0.68°, which, multiplied by 6, gives 4.08 per cent of sugar, with a
rotation equal to 4 per cent. Thus, there can be absolutely no doubt
that the factor which was first calculated on the theoretical basis and
subsequently demonstrated to be correct, making use of diabetic
urine, and normal urine to which sugar had been added, must be
accepted ; and the figures obtained by Fuchs bearing upon the subject
of diabetic urine must be looked upon as erroneous. He gives 10 as
the factor by which the difference between the two freezing points

Experiments on Urine. 337

must be multiplied, which would, as may readily be seen, lead to
serious error in each case presented in the tables accompanying this
article; and further, this factor is not in agreement with the theory
and calculations based upon the relative size of the molecules.

Practical method of working. — The method which we have adopted
in the laboratory is as follows: The specific gravity of the urine is
determined at 15° approximately, and any variation in temperature
allowed for by adding 0.0001 for each 1° C. above 15°. The freezing
point of the urine is determined, and at the same time the urine is
polarized. The theoretical freezing point is calculated from the
specific gravity by multiplying the portion of the specific gravity after
the decimal point by 75, the observed freezing point is subtracted
from this figure, and the difference in degrees Centigrade is multi-
plied by 6, which gives very approximately the percentage of sugar in
the urine. The results agree very closely with those found by other
methods, such as polarization, provided there is nothing present in
the urine which interferes with the other methods. So far as I am
able to say, this method is less influenced by external conditions than
any other. It is also possible to make a determination in a very short
space of time, as many as ten or twelve such determinations having
been made ina single hour. It is very surprising that radical changes
in diet do not affect the result very materially, although I have on
occasion observed slight error due to the treatment of diabetic cases
on a restricted diet, with large quantities of bicarbonate, which, pos-
sessing as it does much smaller ions than the average of those present
in urine, should diminish the average ionic weight of the urine.

It should be borne in mind that this method is more especially
applicable to urine containing a.large percentage of sugar; for, the
larger the proportion of sugar in a urine, the smaller is the proportion
of ordinary urine solids. In fact, a point is often reached at which
the normal constituents are so distributed in a large bulk of urine that,
were the sugar removed, the specific gravity due to their influence
would not be more than 1.002 or 1.003. In urines of very slight
concentration, less than half a per cent, it is hardly safe to depend
upon this method for absolute results, although very satisfactory for
a series of comparative results from day to day. Above 2 per cent
up to 7 and 8 per cent, the method was found to be accurate with
a maximum experimental error of 0.3 to 0.4 per cent, the average
variation not being as much as 0.1 per cent.

As stated above, we have, in carrying out this work, refrained from

338 G.NIT..A.. Clowes.

introducing corrections for sub-cooling, atmospheric conditions, etc.,
whereby the experimental error might be reduced in a physicochemi-
cal laboratory to an almost negligible quantity. The method may be
employed, as stated above, with fairly accurate results, in any physi-
ological chemical laboratory, or even in the clinical laboratory of hospi-
tals. In studying and treating severe cases of diabetes, it is of the
greatest value to obtain accurate information as early as possible, and
in that respect we have obtained the most satisfactory results by com-
bining this method and that of polarization, confirming our results
from time to time by one of the other methods.

One point of especial importance might be mentioned at this stage.
Betaoxybutyric acid, when present in urine, interferes very perceptibly
with rotation of the plane of polarized light, causing the dextrose to
show a lower figure than would be anticipated from the actual per-
centage of sugar present in the solution. When large quantities of
this acid are present, the freezing point results deviate from the rota-
tion figure. In fact, the agreement of the results obtained by these
two methods for a period of several days justifies the assumption that
the acid in question is absent. This fact alone is of considerable
importance, as the fermentation method usually employed, and sub-
sequent examination for betaoxybutyric acid, involves a considerable
expenditure of time. We have, in presenting cases, purposely chosen
those which showed large quantities of sugar with only traces of the
oxyacids, in order not to complicate the problem to too great an
extent. We have, however, had under observation cases in which
considerable quantities of the oxyacids were present, sufficient to
exert a more than negligible effect upon the rotation and freezing
point. We have not, however, ventured to make use of this differ-
ence as a means of estimating the quantity of oxybutyric acid, owing
to the fact that the sources of experimental error are so numerous in
such a case as to make its use out of the question, except when large
quantities of the acid are present, which is seldom the case.

Estimation of albumin. — The quantitative estimation of albumin
in urine is extremely difficult to carry out accurately; and since it is
possible, in the majority of cases, to form a fair idea of the actual
amount of coagulable albumin present in urine, by our method, we
propose to give it in outline, although it has not been as thoroughly
investigated or proved as serviceable as has the method of estimating
sugar.

Whilst the sugar molecule is 180, as compared with the 57 to 60 of

Experiments on Urine. 339

urine solids, the molecular weight of albumin is so large as to be
looked upon as infinitely great when compared with the 60 of urine
solids referred to above. That is to say: the effect produced upon
the freezing point by albumin should be so small as to be negligible,
especially in view of the fact that it is seldom present in urine in larger
quantities than 1 per cent. This does not, however, apply to the spe-
cific gravity, which is influenced by albumin, when actually in solution,
just as it is by other solids. Bearing this in mind, we tried the effect
of adding acouple of drops of acetic acid to the filtered urine, deter-
mining the specific gravity and freezing point, then boiling the urine
in order to precipitate the albumin, which would be thrown out of
solution by means of dilute acetic acid, and once more determining the
freezing point and specific gravity in the clear centrifuged portion.

Boiling the urine in order to precipitate the albumin, and centrifug-
ing in order to remove the precipitated solids, leads to an increase in
its concentration sufficient to interfere materially with any calculation,
based upon the loss of specific gravity effected by the removal of the
albumin, unless we have some means at our disposal of reducing the
second specific gravity to the same basis of ion concentration as
the first. The slight change in specific gravity would not be sufficient
to interfere materially with the degree of dissociation of the salts, and
although the removal of albumin should lead to an increased activity
on the part of certain ions, others have probably been completely re-
moved in combination with the proteid. These factors, however,
appear from the results obtained to counterbalance one another, and
may, therefore, be neglected. Our procedure is as follows:

The decimal portion of specific gravity determined after the addition
of a drop or two of acetic, but before boiling, is designated S,. The
freezing point determined in the same urine at the same time is repre-
sented by A,. After boiling and centrifuging, the decimal portion
of the specific gravity and the freezing point, once more determined,
are indicated as S, and A,. Then, in order to reduce S, to the same
ionic concentration as S,, the former is multiplied by A, and divided
by A,. This quantity is then subtracted from S,, which gives that
portion of the specific gravity figure which may be attributed to the
influence of the bodies precipitated by boiling (albumin). It has
been found that each 0.01 of specific gravity is equivalent to 3.8 or 4
per cent of albumin. The difference obtained above is therefore mul-
tiplied by 400 in order to obtain the percentage of albumin. Table
XII gives a series of results obtained in this way, in which a very

340 GoTh A. Clowes.

fair agreement may be observed between the method in question and
other accepted methods which require a much larger expenditure of
time.

As this method is somewhat complicated, it will be simpler to illus-
trate it by means of an example. In Case 1, the specific gravity,
determined at 15° C., by means of a very accurate Westphal balance,

TABLE XII.
ESTIMATION OF ALBUMIN (IN URINES, ETC.).

Albumin
by other
methods.

Ss : 2 ; Albumin
1: 2 2: fi "| Diff. X 400.

per cent. per cent.

0.0261 | 0.0262 | 1.93! 0252 0.0009 0.36 0:35

0.0161 | 0.0173 | 1.265 0.0007 0.28 0.30
0.0153 | 0.0164 285 | 0.01487 0.00043 0.17 0.20
0.0075 | 0.0052 0.0045 0.0030 1.20 1.00
0.0211 | 0.0225 0.0203 0.0008 0.32 0.35
0.0135 | 0.0139 ; 0.01314 0.00036 0.14 0.13

0.0223 | 0.0223 : 0.0207 0.0016 0.64

6
7
8 0.0205 | 0.0212 | 1. 0.0205 nil nil
9

0.0259 | 0.0283 0.0259 nil

EXUDATE CANCER CASE K.
0.0117 | 0.0082 | 0.525 .625 | 0.0070 0.0047

EGG ALBUMIN SOLUTION.
| 0.0197 | 0.0030 | 0.245 , 0.0046 0.0151
|

was 1.0261 before, and 1.0262 after boiling. The freezing points were
respectively 1.935 and 2.010 before and after boiling. The factor

Ae
eae consequently equal to 0.0252. 0.0252 subtracted from
0.0261, the decimal portion of the original specific gravity, gives
0.0009, which, multiplied by 400 = 0.36 per cent.

It must be mentioned at this stage, that whilst a large proportion
of determinations give accurate results, the possibilities of error are
considerable, and in certain cases we have been unable to explain devi-
ations from the correct figure, such as that exhibited in Case 8, in the
table. If accurately and carefully handled, such a method should be
of considerable use, especially from a theoretical standpoint, and in

Experiments on Urine. 341

those cases in which we have to deal with a large percentage of albu-
min, the experimental error should be practically negligible, if the
precaution is taken of making both specific gravity determinations at
the same time, and at exactly the same temperature, and using the
same thermometer for the freezing point determinations in two
successive experiments.

A normal urine free from albumin has been included in the table
as Case 9, in order to show that there is nothing in normal urine to
interfere with satisfactory results by this method. We have also in-
cluded two cases of albumin solutions of high concentration, one an_
exudate in a cancer case containing 1.75 per cent of albumin, and the
other a solution of egg white in water, containing six parts of albumin.
These solutions were handled in exactly the same way as the urines,
the second specific gravity being reduced to the same ionic concen-
tration as the first. There are, of course, very disturbing factors in
this case, owing to the removal of certain salts by the precipitated
proteids and to the neglect of the influence exerted upon the freezing
point by the proteids themselves. Consequently, the close agreement
in the results obtained from the egg-white solution can only be looked
upon as accidental.

This method is merely offered in the form of a suggestion, as hav-
ing possible application for purposes of confirmation and also in
experimental work. We should not recommend it for daily use in
the laboratory, as has been done in the case of the method employed
for the estimation of sugar.

Analysis of urines containing both sugar and albumin. — It will be
seen from the above that albumin and sugar may both be estimated
in the same urine by means of the freezing point method. A few
drops of acetic acid are added, S, and A, determined, the urine is
boiled, and S, and A, obtained. The albumin is estimated, as de-
scribed in the last section, by combination of these factors, and the
sugar may be determined, making use of S, and A,, calculating the
theoretical freezing point from S,, deducting A, from the result, and
multiplying the figure obtained by 6. The few drops of acetic acid
added exert no appreciable effect upon the estimation of sugar by this
method. In fact, it may be said that albumin itself, when present in
small quantities, exerts so slight an effect as not materially to interfere
with the direct estimation of sugar in urine.

342 G. Hf. A. Clowes:

SUMMARY.

1. In normal urine the depression of the freezing point is directly
proportional to the specific gravity, and may be determined from the
latter by multiplying the figures after the decimal point by 75. (As
an example, if the specific gravity is 1.022, 0.022 x 75 =1.65, which
is found by experiment to be the depression of the freezing point of
such a solution in degrees C.) (See Tables [and II.) This portion
of the work is in agreement with the findings of Fuchs.

2. In pathological urines other than those containing sugar and
albumin, the same tendency to maintain a constant proportion between
the freezing point and specific gravity may be observed, although
within rather wider limits than in the case of normal urines. (See
Tables III and IV.)

3. The average weight of the molecule or ion in urine may be de-
termined from these figures as approximately 56 to 60, making use of
Beckmann’s formula,

where J7 = molecular weight of substance dissolved; / =a constant
having the value of 18.5 when water is employed as the solvent; mz =
the quantity of substance dissolved, assumed in this case to be 2.33
grams of substance for each o.o1 of specific gravity; and A = the
observed depression of the freezing point (0.75° for each o.o1 of spe-
cific gravity ).

4. In cancer, typhoid fever, and other diseases in which the excre-
tion of chlorides is extremely small, the effect exerted upon the freezing
point depression is not so great as might be expected from the removal
of such a large proportion of the smaller ions. This tendency to
maintain a constant state of equilibrium in the average size of the
ions and molecules present in the urine may account for the high re-
tention of the products of sub-oxidation and abnormal metabolism,
known to exist in these cases. This view of the case may be advanced
as a further argument in favor of the therapeutic administration of
chlorides to patients on a milk diet, in order that molecular equilib-
rium of the urine may be maintained, and a more complete elimination
of the products of metabolism effected.

5. In diabetic urines, the quantity of sugar may be very readily
estimated by determining the lowering of the freezing point of the

Experiments on Urine. 343

urine in question, and calculating the theoretical lowering of the
freezing point from the specific gravity. The difference between
these two quantities in degrees C., when multiplied by 6, gives a very
close approximation of the actual percentage of sugar present in the
solution. This factor, 6, calculated on a theoretical basis (see footnote,
page 331), and subsequently confirmed in practice, is dependent on
the fact that the molecular weight of sugar is 180, whilst the average
of urine solids is below 60. (See Tables VI, VII, VIII, IX, X, and
XI.) These results are not all in accord with the observations made
by Fuchs, who arrived at his factor, 10, by purely empirical means.

6. The quantity of albumin may be estimated by determining the
specific gravity and freezing point in a urine to which a couple of
drops of dilute acetic acid has been added, then boiling in order to pre-
cipitate the albumin, filtering and once more determining the specific
gravity and freezing point. The second specific gravity is reduced to
the same ionic concentration as the first, by multiplying its decimal
portion by the freezing point before boiling, and dividing by the
freezing point obtained after boiling. This calculated specific gravity
is then subtracted from the original specific gravity determined before
boiling, and the difference multiplied by the factor 400, which gives
the percentage of albumin. (See Table XII.)

7. Ina urine containing both albumin and sugar, the former is first
estimated by addition of acetic acid, as above, and the freezing point
and specific gravity obtained after boiling are employed in the estima-
tion of sugar.

NEW EXPERIMENTS ON THE PHYSIOLOGICAL ACTION
OF -THE. BPROTEOSES.

By FRANK P. UNDERHILL.

(Research Scholar of the Rockefeller Institute.)

[From the Sheffield Laboratory of Physiological Chemistry, Vale University. ]

CONTENTS.

Page
Introductory. . . Re ees ee ke i en ete eee ea lees OD
Technique of the Ce euaments eek: ies, ae Ae eT reek Vel ehth Oy OO
Intravenous injections of typical native s proteits eee 354

Intravenous injections of proteoses prepared by the digestion of aroteids i animal
origin with vegetable enzymes . . . 5 eh

Are proteoses prepared from proteids and Ses both oF veperable § origin,
MON-LOXICGH we) Fs ue, : Sl

Intravenous injections of proteoses ie pared feat repeiable mroteids by hydrolysis
with acids or with superheated water. . . Sa able ys op iene) ik RODS
The physiological action of native proteoses . . : 359

Are products prepared from edestin with acid, accordine to the methods at Pick
and Spiro, truly non-toxic? . . : : 362

Are the toxic effects of proteoses destroyed By Ave Pick anid Spiro methotie ai
Pere e eS aE Sep sae ames eo ote ak ees Ore See eee et OG
Conclusions . . . ee 2 ee eden a Soa wie cp ere Nay ae ies aly Or eee SL
SPINGEL DKESSUECIELACINPS <i... ek, 6 le ee ee! el me

INTRODUCTORY.

HE attempt to follow the fate of ingested proteids in their pas-
sage from the alimentary canal was the occasion for the original
studies on the behavior of proteolytic digestion products introduced
directly into the circulation. Since the early experiments in Ludwig’s
laboratory numerous investigators have devoted their attention to
this subject.!. The physiological reactions called forth by intravenous
injections of such materials — the “ peptone” of earlier writers — are
diverse, and include the rendering of the blood incoagulable together
with changes in its reaction and composition; an acceleration of
lymph-flow; a fall in arterial pressure; anuria; deep narcosis; and
other toxic symptoms, as well as a certain degree of immunity toward
subsequent injections.

1 The earlier literature is reviewed in the paper by CHITTENDEN, MENDEL,
and HENDERSON: This journal, 1898-99, ii, p. 142; see also Studies in Physio-
logical Chemistry, Yale University, 1901, p. 279.

345

346 frank P. Underhill.

The reactions just described are not common to all species of
animals; or at any rate they are not obtained with equal readiness
and intensity. The organism of the dog is particularly susceptible to
the effects of intravenous injections of the products of proteolysis.
In the cat the characteristic symptoms are evoked somewhat less
readily, larger doses being necessary to produce comparable results.
The rabbit, on the other hand, is extremely resistant, and, except in
rare cases, fails to respond in so far as the phenomena involving
the blood are concerned. Quite recently Persano! has found that
although commercial ‘ propeptone” preparations are fatal to guinea
pigs, they fail to render the blood non-coagulable except when
injected rapidly into fasting animals. Persano also secured positive
results with the toad (Bufo vulgaris and B. viridis) and the tortoise
(Emys europea). In Elasmobranchs Bottazzi? noted an effect on
coagulation with 0.5-0.8 gm. “‘ peptone” per kilo; and in those marine
invertebrates, the blood of which does not ordinarily clot, peptone
injections were found to bring about abnormal changes of another
order.

It is well, perhaps, to emphasize these variations in the behavior
of different species, since they have at times been overlooked in com-
paring the experimental results of different investigators. Thus
Fiquet ® noted that various digestive products which were purified by
methods of reprecipitation suggested by Gautier were apparently
harmless when injected intravenously in relatively large quantities
into rabbits. Nencki* has taken occasion to call attention to the
importance of Fiquet’s work in reference to future studies on the non-
coagulation of “ peptone-plasma ;” and Pick and Spiro® have attached
great importance to the negative results noted. All of these inves-
tigators seem to have overlooked the natural immunity of rabbits to
injections of proteoses. Moreover it is said to be possible to produce
a noticeable retardation of blood-coagulation even in rabbits when
sufficiently large doses are rapidly injected.6 All investigators are
agreed with reference to one experimental factor, namely, the

' PERSANO: Archives italiennes de biologie, 1902, xxxvii, p. 409.

2 BotrTazzi: Jbzd., 1902, p. 63.

8 FiQuET: Comptes rendus de la société de biologie, 1897, xlix, p. 459.

* NENCKI: Jahresbericht fiir Thierchemie, 1897, xxvii, p. 196 (foot-note).

® Pick and Spiro: Zeitschrift fiir physiologische Chemie, 1900-01, xxxi,
Pp. 237:

° Cf Noir: Bulletin de Académie royale Belgique (Classe des sciences),
1902, No. 11, p. 866, foot-note 1.

On the Phystological Action of the Proteoses. 347

necessity of making the injections rapidly.1. Only under such con-
ditions are the typical phenomena provoked in even the most
susceptible animals. ‘It seems reasonable to assume that the
intensity of vaso-motor effects is not so much a function directly
of the absolute quantity injected as of the quantity in the circulation
per unit of time, or, in other words, of the concentration of the
substance in the blood.”? Finally the dosage is not without con-
siderable importance in determining the intensity and duration of the
effects obtainable. In general the proteoses provoke a marked
response in doses of 0.3-0.5 gm. per kilo of body-weight in dogs.
According to Thompson ® doses of Witte’s peptone as low as 0.02 gm.
per kilo may retard blood-coagulation. Smaller doses produce con-
trary effects, coagulation being accelerated; while less than 0.0075
gm. per kilo is entirely without influence. Rapidly injected doses
of 0.010-0.015 gm. per kilo never fail, according to Thompson, to
produce a fall in blood-pressure.

It was to the phenomena provoked that the earlier investigators
turned their attention ; and little consideration was given to the more
careful isolation or chemical differentiation of the materials injected.
The latter, briefly designated ‘‘ peptone,” were essentially a mixture
of the neutralized products obtained by the gastric (or sometimes
pancreatic) digestion of crude fibrin or egg-white. With the progress
in the study of digestive proteolysis came the endeavor to ascertain
the physiological influence of better characterized products. The
literature of this subject has been reviewed in detail in an earlier
paper from this laboratory ;* it will suffice here to refer to the con-
tributions of Pollitzer,® Grosjean,® Arthus and Huber,’ Thompson,*
Ledoux, Spiro and Ellinger,!? Chittenden and Mendel associated with
Henderson" and McDermott.” To these may be added the very
Cf. NotF: Loe. cit., p. 871 and 883.

CHITTENDEN, MENDEL, and HENDERSON: This journal, 1898-99, ii, p. 150.
THOMPSON: Journal of physiology, 1896, xx, p. 455.

Cf. CHITTENDEN, MENDEL, and HENDERSON: Loe. cét¢., p. 142.
POLLITZER: Journal of physiology, 1886, vii, p. 283.

GRrosJEAN: Archives de biologie, 1892, xii, p. I.

ARTHUs and Huser: Archives de physiologie, 1896, viii, p. 857.
THOMPSON: Journal of physiology, 1896, xx, p. 455; also 1899, xxiv, p. 374.
Lepoux: Archives de biologie, 1896, xiv, p. 63.

10 Sprro and ELLinGER: Zeitschrift fiir physiologische Chemie, 1897, xxiii,
p- 12!.

11 CHITTENDEN, MENDEL, and HENDERSON: Loc. cit, p. 142.

122 CHITTENDEN, MENDEL, and McDermott: This journal, 1898, ii, p. 255.

mp CO NS

o oo «1 oo co

348 Frank P. Underhill.

recent experiments of Nolf! which have been published since the
completion of our work. Without repeating the particulars of the
various investigations they may be said to indicate that pronounced
differences exist between the physiological action of the group of
proteid substances known collectively as proteoses (propeptones)
and that of the peptones— the latter being substances not pre-
cipitable by ammonium sulphate. Compared with the proteoses the
peptones are relatively inert in the directions specially indicated
in this paper. Furthermore the proteoses differ in respect to their
efficiency in producing toxic symptoms. Thus according to several
of the investigators heteroproteose is more active physiologically
than some of the other fractions obtainable by the modern methods
of proteid differentiation (precipitation by salts, dialysis) ; while
other products like gelatoses and caseoses seem to act only in large
doses. Earlier studies in this laboratory? have shown that typical
results can be obtained not only with products of peptic proteolysis,
but also with proteoses prepared with the vegetable enzyme papain |
or even by the action of acids alone. Finally Nolf? has undertaken
a large number of experiments with proteoses (propeptone) prepared
from fibrin by peptic digestion, by pancreatic digestion, and by
autolysis. In the latter case the proteolysis is attributed to the
presence of an enzyme derived from the leucocytes entangled in
the fibrin. His experimental study has demonstrated that under
the conditions noted (namely, rapid injections of larger doses in
the dog) all these proteoses retard the coagulation of the blood and
lower arterial pressure. The peptic products, however, are relatively
more toxic in this respect. Such variations are doubtless attributable
to differences in the relative proportions of different specific proteoses.
That the latter should differ in their physiological action, even though
they are derived from the same mother substance, need not surprise
us in view of the apparent chemical dissimilarities which recent inves-
tigation has revealed in regard to composition of the proteoses.4 Indeed
it is only to be expected that with the refinement of chemical methods
for the separation of proteoses, the well-defined individuals of this

group of proteid compounds will be found to react in specific ways
in the organism.

1 NoL¥F: Loe. cits

? CHITTENDEN and MENDEL, with HENDERSON and McDermott: Loe. cét.
®. NOLE: Locnet

* Cf. PIcK: Beitrage zur chemischen Physiologie, 1902, ii, p. 509.

On the Physiological Action of the Proteoses. 349

It should be pointed out in this connection that the characteristic
physiological reactions noted after proteose injections can also be
obtained with a variety of other substances of known or partly
known composition. This statement applies, for example, to the
nucleic acids, as has been demonstrated in a recent investigation
by Mendel, Underhill, and White in this laboratory! The doses
required are far smaller than in the case of the proteoses, 0.05 gm.
per kilo of body-weight sufficing to bring about a noticeable fall in
arterial pressure, lymphagogic effects, and retardation of coagulation
in the case of the nucleic acid of the wheat embryo (tritico-nucleic
acid). Similar results were previously noted by Bang,? who injected
pancreas nucleoproteid. In addition to this may be recalled the
older observations on lowered arterial pressure and retardation of
blood-coagulation induced by injections of various enzymes,’ bac-
terial and tissue products, snake venoms,’ protamins,® etc. Many
of these products also act in extremely small doses. But this fact,
by itself, will not justify the assumption that in view of the larger
doses of proteose necessary to produce comparable symptoms, the
physiological action of the proteoses is due to some highly toxic
contaminating substance like those referred to. Otherwise one
might, for example, as well attribute the tetanic action of large doses
of brucin preparations to the presence of small quantities of adherent
strychnine which acts in far smaller doses than its companion
alkaloid.

Although the toxic properties of the proteoses have been generally
accepted as a specific characteristic of their behavior after direct
introduction into the circulation, this view has lately been subjected
to severe criticism in Hofmeister’s laboratory by Pick and Spiro.’
Those investigators demonstrated, in the first place, that enzymes

1 MENDEL, UNDERHILL, and WHITE: This journal, 1903, viii, p. 377- The
literature is cited in this paper.

2 BANG: Zeitschrift fiir physiologische Chemie, 1901, xxxil, p. 201.

3 Cf. ALBERTONT: Centralblatt fiir die medicinische Wissenchaften, 1878, xvi,
p- 641; Satviot1: Jézd., 1879, xvii, p. 596; DASTRE and FLoREscO: Comptes
rendus de la société de biologie, 1897, xlix, p. 847.

4 CHITTENDEN: Digestive proteolysis, 1895, p go.

5 WOLFENDEN: Journal of physiology, 1886, vii, pp. 327,357,305; KANTHACK:
Ibid., 1892, xiii, p. 272; MARTIN: /67d., 1893, xv, p. 380; also Proceedings of the
Royal Society of New South Wales, 1892.

6 THompPson : Zeitschrift fiir physiologische Chemie, 1900, xxix, p. I.

7 Pick and Spiro: Loc. cit.

350 Frank P, Underhilt.

can be dispensed with in the production of physiologically active
mixtures from proteid substances. This was, however, quite to be
expected after Chittenden, Mendel, and Henderson had shown that
products obtained by the hydrolysis of egg-albumin with sulphuric
acid induce a pronounced physiological reaction, as well as those
formed from the same proteid by pepsin digestion! With the
products of pancreatic digestion of fibrin Pick and Spiro failed to
obtain any alteration in the coagulability of the blood, and they call
attention to the similar failure of Fano? in earlier experiments. With
reference to the actual proteose content of the “trypton” used by
Fano nothing is known. Pick and Spiro state that the unpurified
digestion mixture used by them contained “an abundance of albu-
moses precipitable by half-saturation with ammonium sulphate.”
The actual dosage of proteose in the single experiment recorded
(Versuch III) was not ascertained nor is any statement made regard-
ing the rate of injection. Chittenden, Mendel, and Henderson
obtained pronounced reactions with isolated proteoses prepared by
the action of pancreatic juice on ‘“‘antialbumid ;”? and in accord with
this are the recent experiments of Nolf* with isolated ‘‘ propeptone ”
obtained (by saturation with ammonium sulphate) from the products
of the digestion of fibrin with pancreatic extracts. The products of
autolysis of fibrin also failed to give positive results in the hands
of Pick and Spiro. But here again no attempt was made to study
the isolated proteoses, the authors contenting themselves with the
statement in one experiment (VI) that the material injected “still
contained small amounts of the albumoses precipitable by half-
saturation with ammonium sulphate.” In direct contradiction to
these observations are the positive results obtained by Nolf with
“ propeptone”’ isolated from comparable material.? Nolf attempts to
explain the negative results of the German investigators on the basis
of his own experiments by assuming that they failed to make the
intravenous injections with sufficient rapidity. From our experience
we think it not unlikely that the actual content of proteose in Pick and
Spiro’s injection mixtures was smaller than they intended it to
be. Furthermore, we have as yet no means of knowing to what

1 CHITTENDEN, MENDEL, and HENDERSON: Loc. cit.
2 FANO: Archiv fiir Physiologie, 1881, p. 277.

3 CHITTENDEN, MENDEL, and HENDERSON: Loc. cit.
2 INOLE: oc: ect.

© NOE =) Loc: cz.

On the Physiological Action of the Proteoses. 35%

extent the non-proteose constituents of mixtures such as they used
may influence the reactions attributed to the proteoses. In fact, it
will be shown in the experimental part of this paper that the property
of producing the typical physiological effects is lost when purified
proteoses are heated with dilute acid until the biuret reaction can no
longer be obtained.

In their discussion of the question as to whether the clot-retarding
agent in proteolysis mixtures is a proteose, Pick and Spiro conclude:
‘‘ Albumose mixtures obtained by the action of trypsin, autolysis, or
alkalies on crude fibrin have no action on blood-coagulation, although
they can be demonstrated to contain albumoses which cannot, at the
present time, be distinguished from those obtainable by the action
of pepsin-hydrochloric acid or acid alone.” + But as indicated in the
preceding review, their evidence is as yet inadequate to maintain this
assumption. In addition to the contradictory experience of Nolf and
other criticism noted above, it must again be emphasized that Pick
and Spiro made no attempt to inject the proteoses as such, prepared
by the various methods referred to. This seems to us a matter of
primary importance, however; and the criticism applies equally to
the subsequent chapter in Pick and Spiro’s paper in which they have
discussed their failure to obtain anti-clotting substances by the
hydrolysis of pure proteids (casein and edestin) with acids alone.
Here again we must refer to the earlier experiences in this laboratory
with the proteoses prepared from egg-albumin by the action of acid
alone. The significance of these positive experiments does not seem
to have impressed Pick and Spiro, although they quote the negative
results obtained by Chittenden, Mendel, and Henderson in a trial of
anti-albumid — not ordinary proteose — made from edestin. Pick
and Spiro’s method consisted in digesting purified casein, or edestin,
a crystalline proteid from hemp-seed, with 0.8 per cent hydrochloric
acid for several days in athermostat. The solutions were neutralized,
the acid proteid removed, and the filtrates dried on a water bath.
Solutions of these resulting mixtures were then injected into dogs
and the effects on the coagulability of the blood noted. The details
reported show in one case that the injection was made in four min-
utes; in the other no data are given. The presence of proteoses was
tested for by fractional saturation trials with ammonium sulphate. In
the case of the edestin products the largest fraction was that obtained
by one-half saturation. The isolated proteoses were not injected.

1 Pick and Spiro: Loc. cit., p. 250.

352 Frank P. Underhill.

In repeating these experiments under similar conditions we have
been unable to obtain large quantities of proteoses in so short a time.
And in this connection it may be well to point out the difficulty of
separating out the acid proteid completely from such mixtures. How
difficult it is to remove edestin itself from solutions by coagulation
alone was pointed out long ago by Chittenden and Mendel; and the
precautions necessary for the complete separation of acid proteids
have lately been indicated by Hawk and Gies.2, Experiments which
we have made according to the directions of Pick and Spiro have
given us similar negative results. But, as will be seen later, the
isolated proteoses prepared by the acid hydrolysis of pure proteid,
(edestin, etc.) were found to exert the characteristic retarding action
on blood-coagulation; and in those trials in which the hydrolysis
was extended over a longer time than that suggested by Pick and
Spiro and in which the content of proteoses was actually determined,
positive results were likewise noted.

Pick and Spiro have attempted to demonstrate the non-toxicity of
the proteoses in still other ways which involve, according to them, a
purification of the proteid. They have found: (1) that the anti-
clotting substance is destroyed by alcohol in weakly alkaline, but not
in acid solutions; (2) that the anti-clotting substance is presumably
not preformed in crude fibrin, but is liberated or manufactured by
the action of acids. Fibrin would, then, contain a fore-runner, the
mother-substance of the toxic substance; and the so-called phys-
iological albumose-reactions are accordingly to be attributed to a
contamination of the proteid — of the fibrin used, in the case of com-
mercial “ peptones.” In further support of this view the authors find
that anti-clotting extracts (free from proteoses) can be prepared from
various tissues of the body, viz., the gastric and intestinal mucosa
and the pancreas; while negative results are obtained with albumose-
containing extracts of testicles, thymus, suprarenals, submaxillary.
and lymphatic glands, and cesophageal membrane. In formulating
their explanation it has apparently not occurred to Pick and Spiro
that there may be (and probably are) different anti-clotting sub-
stances in various tissues of the body; nor have they made mention
of the possible presence of compounds (like some nucleoproteids)
which facilitate coagulation and may act antagonistically to the
proteoses. The presence of non-proteid anti-clotting compounds in

' CHITTENDEN and MENDEL: Journal of physiology, 1894-95, xvii, p. 48.
* HAWK and Gigs: This journal, 1902, vii, p. 460.

On the Physiological Action of the Proteoses. 353

the tissues can surely not be taken fer se as evidence of the inactivity
of proteoses prepared from tissue proteids. In the case of the pan-
creas, for example, the anti-clotting power of its constituent nucleic
acid compounds has already been mentioned. For the anti-clotting
substance (or substances) derived from the tissues, the name “ pep-
tozyme” has been proposed, its antecedent being designated as
‘““peptozymogen.” Peptozyme is, then, a substance very resistant
to the action of dilute mineral acids and to heating in neutral or
faintly alkaline solutions; and since its antecedents are present in
many of the animal tissues, proteoses derived from the latter by
digestion must be contaminated with it. Thus preparations made
with pepsin may contain active peptozyme derived from the gastric
mucosa; fibrin products like ‘‘Witte peptone” are doubly contam-
inated. In attempting to account for the positive results which
Chittenden, Mendel, and McDermott! obtained with proteoses pre-
pared by digestion of coagulated egg-white with papain in adkaline
media, Pick and Spiro have made the purely hypothetical assumption
that the peptozymes (derived from the egg?) are resistant to papain.
Hypothetical as much of this is, it is by no means impossible or even
improbable. But misguided by their inadequate experiments with
pure proteids, etc., discussed above, the authors have gone a step
further and defend the theme: “Es gibt ‘ Peptone’ ohne ‘ Pepton-
wirkung’ und ‘ Peptonwirkung’ ohne ‘ Peptone’” (p. 272).

The latter part of this proposition no one will deny. We propose
to examine somewhat more closely the statement that the proteoses
per se are physiologically inert in so far as their action on the coagu-
lation of the blood and on arterial pressure is concerned. Pick and
Spiro have summarized their position as follows: ‘It is possible to
obtain by proteolysis (e.g. by trypsin, autolysis, alkali; with casein
and edestin by acid also) typical albumoses and peptones which fail
to show any influence upon blood-pressure when they are injected
into the blood. But even the typical active products prepared by
acid or pepsin-acid methods, lose this activity by a relatively harm-
less purification (treatment with alcohol) without undergoing a change
in chemical character” (p. 271-272).

In the present research proteoses have been prepared from a
number of purified proteids of different origin by various methods
of hydrolysis, the presence of hypothetical anti-clotting compounds
being excluded as far as our present knowledge and methods would

1 CHITTENDEN, MENDEL, and McDeERmotTT: Loc. c7t.

354 Frank P. Underhill.

permit. In addition to this, many of the experiments of Pick and
Spiro have been repeated in the hope of arriving at an explanation
of the discrepancies which have arisen.

TECHNIQUE OF THE EXPERIMENTS.

The experiments were carried out upon full-grown dogs which
were always made to fast for a day previous to the experiment.
Narcosis was induced by a subcutaneous injection of a mixture of
morphine and atropine sulphates,’ and when necessary, as during the
operative procedures, complete anasthesia was maintained by the
administration of A.C.E. mixture. Arterial pressure was recorded
from the left carotid artery with a mercurial manometer. The injec-
tions were made at a temperature of 38° C. into the right femoral vein
from a burette. The blood samples were drawn from the left femoral
artery, a clean, dry cannula being used for each sample. A small
portion of blood was always allowed to flow through the cannula
before the samples (2-3 c.c.) were collected in dry, graduated cylin-
ders of uniform calibre. The rate of coagulation was taken to be
the time which elapsed before the cylinder could be inverted without
the loss of adrop. The substances injected were dissolved either in
0.7 per cent sodium chloride solution or in distilled water to which
enough sodium carbonate was added to give a weak alkaline reaction,
the volume used being equivalent, where possible, to 4 c.c. per kilo
of animal. The injections were made rapidly —usually within a
minute.

Unless otherwise stated, the proteoses used were obtained by satu-
ration with ammonium sulphate after neutralization of the digestive
mixtures and removal of undigested residues and the neutralization
precipitate. The precipitates of proteoses were re-dissolved, dialyzed
until extirely free from ammonium sulphate, concentrated, precipi-
tated with alcohol, washed with hot alcohol, and finally dried with
ether.

INTRAVENOUS INJECTIONS OF TYPICAL NATIVE PROTEIDS.

Heidenhain? in his experiments on lymph-formation, has, in one
case in which a native proteid was injected, noted that it had a

! The usual dose employed was ten milligrams of morphine sulphate and one
milligram of atropine sulphate per kilo of animal.
* HEIDENHAIN: Archiv fiir die gesammte Physiologie, 1891, xlix, p. 209.

On the Physiological Action of the Proteoses. 355

lymphagogic effect and yielded a non-coagulable lymph. Spiro!
showed, however, that the injection of crystallized serum albumin
is without influence. In the present investigation intravenous in-
jections have been made with crystallized egg-albumin, ovomucoid,
and the vegetable proteid excelsin obtained from the brazil-nut,
Bertholletia. A typical protocol follows.

Experiment 18. — A bitch received an injection (82 c.c.), containing 4 gms.
(0.5 gm. per kilo) of crystallized egg-a/bumin.? ‘The reaction of the solu-
tion was slightly acid to litmus. The injection lasted 50 seconds. No
influence was noted upon blood-coagulation ; nor was there any fall in
blood-pressure.

That the animal was not “immune”? follows from the fact that a
subsequent injection (32 c.c.) of 4 gms. (0.5 gm. per kilo) of so-called
hemialbumose* (made by the action of acid upon egg-albumin), lasting
75 seconds, completely suspended coagulation and lowered the pressure
from 155 mm. to 20 mm. of mercury. At the end of an hour it had
risen to 35 mm. Deep narcosis followed, together with marked intestinal
peristalsis.

Similar experiments carried out with ovomucoid and with excelsin
(in doses of 0.5 gm. per kilo body-weight) led to identical results.

It follows from these experiments that the typical native proteids
may be injected rapidly without influencing the coagulation and
arterial pressure of the blood, even when they are introduced in
considerable quantity.

INTRAVENOUS INJECTIONS OF PROTEOSES PREPARED BY THE
DIGESTION OF PROTEIDS OF ANIMAL ORIGIN WITH
VEGETABLE ENZYMES.

The only experiments that have been carried 0 t to determine the
influence of injections of proteoses prepared by the digestion of animal

1 Spiro: Archiv fiir experimentelle Pathologie und Pharmakologie, 1898, xli,
p- 148.

2 Prepared according to HOPKINS and PINKus: Journal of physiology, 1898-99,
mii. p: P30:

3 By “immune” is meant the natural peculiarity of an animal, in that substances
which generally have certain effects upon a species are without influence upon
certain individuals. ‘Temporary immunity may be established towards proteoses ;
for example, an injection of proteoses usually renders a subsequent injection
ineffective, provided the time which has elapsed between the two injections is not
too long.

4 CHITTENDEN, MENDEL, and HENDERSON: This journal, 1898-99, ii, p. 173.

356 Frank P. Underhill.

proteids with vegetable enzymes are those of Chittenden, Mendel,
and McDermott.! These investigators observed that proteoses and
peptones prepared from egg-albumin by the action of papain behaved
like products formed by pepsin. Since Pick and Spiro assume that
the toxic effects noted after injection of proteoses are due to a con-
taminating substance derived from the animal tissues, and that these
effects are not obtained when casein-acid products are employed, it
was of interest to determine whether caseoses prepared by the agency
of plant enzymes are non-toxic. Such injections have been made
with bromelin and “ Papoid”’ caseoses. These products were pre-
pared by the action of bromelin —the proteolytic enzyme of the
pineapple —and ‘“ Papoid’’— a commercial papain preparation, on
casein in alkaline media.

Experiment 5.— A bitch of 5 kilos received an injection (20 c.c.) of 5 gms.
(1.0 gm. per kilo) of caseoses prepared with bromelin.? The reaction of
the fluid was faintly acid. The injection lasted 40 seconds. ‘The animal
remained quiet. The following figures from protocols are given merely
to indicate the method of observation employed in the numerous experi-
ments to be reported. The average ® clotting time of the blood before
the injection was 8 minutes.

Sample collected after} Time of clotting. Blood-pressure in mm. mercury.

1 min. oe Before injection .. . . 140
3 min. sans At end of injection . . . 55

8 min. ater 2 min. after mjection asa mod

23 min. plist 3 Min.atter injechiony mim

4 hrs. 22 min. 2 hrs. 27 min. 5 min. after injection .
7 min. after injection .

4 hrs. 22 min. after injection

1 Had not clotted in twenty hours.

It may be concluded, therefore, that injections of caseoses formed
by the agency of vegetable enzymes provoke the effects typical for
other proteoses.

1 CHITTENDEN, MENDEL, and McDErRmMotT: Loc. cit.
* Prepared according to CHITTENDEN: Journal of physiology, 1894, xv, p- 459.
8 The average of three or four samples.

On the Physiological Action of the Proteoses. 357

ARE PROTEOSES PREPARED FROM PROTEIDS AND ENZYMES,
BOTH OF VEGETABLE ORIGIN, NON-TOXIC ?

If the hypothetical “ peptozyme” of Pick and Spiro is truly a con-
tamination derived from the animal organism and is not present in
the vegetable kingdom, as their results would seem to indicate, pro-
teoses prepared from the vegetable proteids with an enzyme also of
plant origin might be expected to give negative results when injected
into the blood stream. In the absence of experiments on this subject
recorded in the literature, trials were made with products prepared
from crystallized edestin (of hemp-seed) by the action of bromelin
and papain.

Experiment 10.— A dog received an injection (22 c.c.) of 2.75 gms. (0.5
gm. per kilo) of edestin proteoses prepared in the usual manner from a
digestion of crystallized edestin by Merck’s Japa in an alkaline medium.
The reaction of the fluid was faintly acid, and the injection lasted 30
seconds. ‘There was evidence of great excitation (cries, struggles, etc.),
together with labored breathing. Defeecation also occurred. Later deep
narcosis appeared.

The average clotting time of the blood before the injection was 10
minutes.

The samples of blood collected at short intervals for 2 hours following
the injection did not clot within 24 hours. A sample collected 2 hours
4 minutes after the injection clotted in 22 minutes. The blood-pressure
showed the typical fall followed by a gradual rise.

A subsequent injection of “ Witte’s peptone” (0.3 gm. per kilo)
caused a fall in pressure comparable with that of the first injection, except
that it returned to the normal more speedily.

Proteoses prepared from edestin with bromelin and injected in
doses of 1.3 gms. per kilo provoked effects similar to those observed
for proteoses prepared from edestin with papain.

Since the proteoses employed in these experiments were entirely
free from contaminating substances of animal origin, and further,
since according to Pick and Spiro the toxic body, “ peptozyme,”’ is
not developed in an alkaline medium (as in pancreatic digestion), we
attribute the results to the proteoses themselves.

358 Frank P. Underhill.

INTRAVENOUS INJECTIONS OF PROTEOSES PREPARED FROM
VEGETABLE PROTEIDS BY HYDROLYSIS WITH ACIDS
OR WITH SUPERHEATED WATER.

Chittenden, Mendel, and Henderson! have noted that albumoses
prepared by hydrolysis with acids exert a very toxic influence when
injected. Pick and Spiro later observed similar effects after injection
of products obtained from fibrin with acids. With their product
prepared from edestin no such results were noted, the coagulability
of the blood and the arterial pressure remaining unchanged. No
experiments were carried out with the zso/ated proteoses, however,
the products employed being mixtures of substances remaining after
removal of the neutralization precipitate.

As has been noted above, excel/szx, a simpie proteid obtained
from the brazil-nut, is without influence upon the coagulability of
the blood or arterial pressure. If this proteid be hydrolized with
dilute hydrochloric acid and the resulting proteoses separated, the
physiological effects produced by them resemble those called forth by
other proteoses. Furthermore proteoses prepared by the action of
sulphuric acid on edestin and zein? provoke similar effects.

Experiment 46.— he animal received an injection (34 c.c.) of 4.25 gms.
(o.5 gm. per kilo) of profeoses, prepared by the action of 0.8 per cent
hydrochloric acid on exce/sin for 20 hours in a sterilizer, and separated in
the manner usually employed in this investigation. The reaction of the
solution was faintly alkaline, and the injection lasted 45 seconds. ‘There
was evidence of great excitation (cries, struggles, etc.), followed by deep
narcosis. Increased intestinal peristalsis was also noted.

The average clotting time of the blood before the injection was 14
minutes. The samples of blood collected at frequent intervals for 2
hours after the injection showed no tendency to clot 24 hours later. The
usual fall in pressure was observed.

Injections of proteoses formed from zein® and edestin (by the
action of 2 per cent sulphuric acid in a sterilizer) in doses of 0.2-0.9

1 CHITTENDEN, MENDEL, and HENDERSON: Loc. cit.

* SZUMOWSKI: Zeitschrift fiir physiologische Chemie, 1902, xxxvi, p. 198.
This investigator has shown that injections of zein itself provoke a retardation
of blood-coagulability and a fall in arterial blood-pressure. The effects are, how-
ever, not so marked as with the injection of proteoses from zein.

® The zein was prepared as follows: Corn meal from which the starch had
been removed was extracted with 95 per cent alcohol. The extract was filtered

On the Physiological Action of the Prateases. 359

em. per kilo either retarded or entirely inhibited the blood-coagu-
lation. The typical phenomena connected with the blood-pressure
were always present.

The following experiment with proteoses, prepared from edestin
by the action of superheated water in the autoclave, and separated
in the usual manner, was also carried out.

Experiment 14. — A bitch of 7.5 kilos received an injection (50 c.c.) of 2.75
gms. (0.36 gm. per kilo), of edestin-proteoses in a faintly alkaline solu-
tion. The injection lasted 67 seconds. ‘The animal remained quiet.

The average clotting time of the blood before the injection was 9%
minutes.

Blood samples collected 1 min., 6 min., 32 min., and 1 hour 12 min.
after the injection, clotted in 3 hours, 2 hours 37 min., 27 min., and 12
min. respectively. The usual fall in the pressure of the blood was also
observed.

‘These experiments suggest the conclusion that proteoses resulting
from the hydrolysis of vegetable proteids may exert a toxic influence
in the manner described.

THE PuysioLoGicAL ACTION OF NATIVE PROTEOSES.

In view of the observations recorded above, showing that typical
native proteids exert no influence upon either blood-coagulability or
arterial blood-pressure, and since proteoses prepared from these native
proteids, either by means of enzymes or hydrolysis with acids, are
toxic, it remained to ascertain whether the proteoses naturaily oc-
curring in the vegetable kingdom have any toxic influence when
injected.

Several experiments of this order have been carried out with native
proteoses obtained from the brazil-nut, hemp-seed, and the wheat
embryo.

Brazil-nut proteose was prepared as follows: The comminuted nuts were
extracted with benzine to remove fat, and then with warm distilled water.
The aqueous solution was allowed to stand in the cold for several days,
when the excelsin separated. ‘lhe filtrate was‘concentrated to a small
volume, filtered free from resulting coagulum, and the proteose present
precipitated from the filtrate with alcohol.

and poured into a large volume of water containing a little sodium chloride in
which the zein was insoluble. The hard precipitate was thoroughly washed with
distilled water.

360 Frank P. Underhill.

Hemp-seed proteose was prepared by extracting hemp-seed at 60° C. with 3
per cent sodium chloride solution from which the edestin separated on
cooling. ‘The filtrate, when concentrated by heat and freed from coagu-
lated proteid, was saturated with sodium chloride with the addition of
acetic acid. The resulting precipitate of proteoses was dissolved in water
and dialyzed free from salt.

Wheat-embryo proteose was prepared by extracting wheat germ meal with 3
per cent sodium chloride solution heated to 70° C., dialyzing the extract
in water, coagulating the coagulable proteids by heat, and precipitating
the proteoses by dialysis in alcohol. This precipitate was dissolved in
water, saturated with ammonium sulphate, the resulting precipitate redis-
solved in water, the solution dialyzed free from sulphate, and then dia-
lyzed into alcohol.

Typical protocols follow.

Experiment 13. — Bitch of 5.1 kilos. The animal received an injection
(20.4 c.c.) of 1.36 gms. (0.26 gm. per kilo) of ative proteoses, obtained
from the drazzd-nut,’ in a faintly acid solution. ‘The injection lasted
40 seconds, during and succeeding which for a short period the animal
gave evidences of marked excitation (cries, etc.).

The average clotting time of the blood before the injection was 9
minutes.

Blood samples collected during 30 minutes following the injection had
not clotted 12 hours later. Within 20 hours a soft clot had formed.
Samples collected 49 min., and 2 hours after the injection, clotted in
4 hours 25 min., and 1 hour 58 min., respectively. The usual fall in
blood-pressure was observed.

Chittenden, Mendel, and Henderson? have shown that proteoses
prepared from typical proteids by enzymes or by hydrolysis with
acids may also exercise a lymphagogic influence, at the same time
rendering the lymph incoagulable and producing characteristic
changes in its composition. A single experiment is submitted
here, showing the influence of a za#/ve proteose upon the blood and
lymph.

Experiment 81. — A dog of 12 kilos received an injection (48 c.c.) of 6 gms.
(0.5 gm. per kilo) of native proteoses (obtained from hemp-seed), in a
weakly alkaline solution. The injection lasted 75 seconds, during which,

* The preparation was kindly furnished by Dr. T. B. Osborne of the Con-
necticut Agricultural Experiment Station.
2 CHITTENDEN, MENDEL, and HENDERSON: Loc. cit., p. 162.

On the Physiological Action of the Proteoses. 361

and for a short period following which, the animal showed evidences of
great excitation (cries, struggles, etc.). This period of excitation was
followed by deep narcosis, which culminated in death 2} hours later.

TABLE SHOWING INFLUENCE UPON LYMPH-FLOW, ETC.

Lymph-flow in Total solids Ash of
10 minutes. of lymph. lymph.

c.c, per cent. per cent.

10.53-11.03 2.3
11.03-11.13 21 I;

4.00 0.87

Injection (see above)

11.16-11.26 ded,
11.26-11.36 6.7
11.36-11.46 7.0
11.46-12.06 36
12.06-12.16 2.2
12.16-12 26 Seu
12 26-12 36
12.36-12.46
12.46-12.56
12.56-1.06

1.06-1.16

1.16-1.26

1.26-1 36

1.36-1.46

1.46-1.56

1.56-2.06

There was no change in the mode of respiration to account for the second in-
creased lymph-flow. None of the samples after injection showed the least tendency
to clot, and all samples after the injection, with the exception of the first two, were
tinged with red.

The average clotting time of the blood before the injection was 15
minutes.

Blood samples collected at intervals during an hour following the
injection were still liquid 24 hours later. Samples collected 1 hour

362 Frank P. Underhill.

37 min., and 2 hours 36 min., after the injection, clotted in 2 hours
57 min., and 44 min., respectively. The typical fall in blood-pressure
was noted.

The results of these and several similar experiments indicate that
in doses of 0.3-0.5 gm. per kilo native proteoses may provoke the
effects characteristic of other proteoses.

ARE PRODUCTS PREPARED FROM EDESTIN WITH ACID, ACCORDING
TO THE METHODS OF PICK AND SPIRO, TRULY NON-TOXIC?

Pick and Spiro have shown that products formed by the action of
acid upon fibrin exercise a toxic influence when injected. Products
similarly obtained from edestin had no influence either upon the
blood-coagulability or upon the blood-pressure. These experiments
form the basis for their assertion that the typical effects noted after
proteose injection are due to some contamination derived from the
animal tissues and absent in the purified vegetable proteids.

In the present investigation the same method of preparing products
has been followed.

100 gms. of crystallized edestin were allowed to stand in 2 litres 0.8 per cent
hydrochloric acid for one week, at 38° C. The mixture was then
filtered, the filtrate neutralized with sodium carbonate, and the resulting
precipitate of acid-proteid filtered off. The filtrate was evaporated to
dryness on a water bath. ‘The dry residue amounted to 17 gms. of a
light brown powder. This material was injected. In the following ex-
periment it was intended to inject 7.25 gms. (1.25 gms. per kilo), but
2 gms. of the material were insoluble.

Experiment 2.— (Compare Experiment X, Pick and Spiro”). A bitch of
5-8 kilos received an injection (18.8 c.c.) of 5.25 gms. (0.9 gm. per kilo)
of the above product in a faintly alkaline solution. ‘The injection lasted
35 seconds, during which, and for a short period following which, the
animal gave evidence of marked excitation (cries, struggles, etc.). No
influence on blood-coagulability was noted. A noticeable fall in pressure
was observed.

Four minutes after injection the dog suddenly died. Upon examina-
tion it was found that there were clots in the right ventricle and in the
large blood-vessels. The intestines were very pale. Analysis of the
product injected showed that of the total nitrogen, 14.9 per cent was in
the form of compounds precipitable with zinc sulphate.

.G. Pick and SPIRO? Loc. Scir, pean
2 Pick and SPiro's) Loci p- 252.

On the Physiological Action of the Proteoses. 363

A second series of experiments was then carried out as follows:

130 gms. of edestin were dissolved in 44 litres of 0.8 per cent hydrochloric
acid, and the solution divided into two equal parts. One portion (Por-
tion I) was allowed to digest one week at 38° C., and then treated as in
the previous experiment. It yielded 25 gms. of the final product. ‘The
second portion (Portion II) was allowed to digest three weeks, and then
treated as above. It yielded 30 gms. Since in the second portion the
bulk of the proteid present was precipitated as acid-proteid on neutraliza-
tion, this precipitate was further digested with 0.8 per cent hydrochloric
acid at roo° C. for 12 hours, and then treated as the other portions
(= Portion III). The different portions were then injected. Protocols
follow :

Injections of Portion I. LZxperiment 25. — A dog of 7 kilos received an
injection (28 c.c.) of 8.75 gms. (1.25 gms. per kilo) of Portion I in a
faintly alkaline solution. The injection lasted 60 seconds. Evidence
was obtained of excitation (cries, etc.). No influence on blood-coagula-
bility was observed. ‘There was, however, a noticeable fall in arterial
pressure.

Analysis of Portion I showed that of the total nitrogen 87.8 per cent
was in a form precipitable with zinc sulphate. A subsequent injection of
“ Witte’s peptone ”’ failed to have any influence upon blood-coagulability ;
but a distinct fall in blood-pressure was observed.

Experiment 35.—A bitch of 6 kilos received an injection (24 c.c.) of 7.5
gms. (1.25 gms. per kilo) of Portion I in a faintly alkaline solution.. The
injection lasted 30 seconds, and the animal remained quiet.

The average clotting time of the blood before the injection was 14
minutes.

Blood samples collected 1 min., 6 min., 27 min., 50 min., 1 hour 17
min., 1 hour 50 min., after the injection, clotted in 24 min., 42 min.,
23 min., 53 min., 28 min., and 41 min., respectively.

The blood-pressure fell very rapidly and remained stationary.

Injections of Portion II. Lxferiment 30. == oN dog of 6 kilos received an
injection (24 c.c.) of 7.5 gms. (1.25 gms. per kilo) of Portion II in a
faintly alkaline solution. The injection lasted 30 seconds. There was
evidence of great excitation (cries, struggles, etc.), for several minutes,
followed by deep narcosis. ‘Ihe average clotting time of the blood be-
fore the injection was 17 minutes.

Blood samples collected 2 min., 6 min., 24 min., and 4o min., after the
injection, showed no tendency to clot 24 hours later. Samples collected
1 hour to min., and 2 hours 6 min., after the injection, clotted in 3

364 Frank P,. Underhill.

hours, and 54 min., respectively. The typical fall in the blood-pressure
was also obtained.

Experiment 82.— A bitch of g kilos received an injection (36 c.c.) of 4.5
gms. (0.5 gm. per kilo) of Portion II in a faintly alkaline solution. ‘The
injection lasted 43 seconds. The animal remained quiet throughout.

The average clotting time of the blood before the injection was 21
minutes.

Blood samples collected 1 min., 4 min., 6 min., and 56 min., after the
injection, clotted in 45 min., 43 min., 58 min., and 28 min., respectively.
There was a noticeable fall in the blood-pressure.

Analysis of Portion II showed that of the total nitrogen 51.3 per cent
was in a form precipitable with zinc sulphate.

Injections of Portion III. Hxperiment 41.— A dog of 7 kilos received an
injection (35 c.c.) of 8.75 gms. (1.25 gms. per kilo) of Portion III in
a faintly alkaline solution. ‘The injection lasted 35 seconds. ‘There was
evidence of great excitation (cries, struggles, etc.), which was followed by
deep narcosis.

The average clotting time of the blood before the injection was 21
minutes.

Blood samples collected 1 min., 10 min., 25 min., 48 min., 1 hour
20 min., and 1 hour 4o min., after the injection, clotted in 62 min., 43
min., 25 min., 48 min., 1 hour 38 min., and 45 muin., respectively. The
typical fall in blood-pressure was also noted.

Analysis of Portion III showed that of the total nitrogen 20 per cent
was in a form precipitable with zinc sulphate.

From these experiments it is seen, that although products formed
from edestin by a week’s hydrolysis with hydrochloric acid may be
without influence, yet with products formed by three weeks’ hydroly-
sis the effects noted after injection are typical for the isolated pro-
teoses. The failure of Pick and Spiro to obtain similar results is
apparently due to the fact that very little true proteose was contained
in their products. In many attempts of the writer to isolate the pro-
teoses in the usual manner by saturation with ammonium sulphate,
the effort was fruitless. Little or no proteose was present. It is a
matter of extreme difficulty to remove all the acid-albumin-like body
from such digestion mixtures, and this is the substance of which their
product was probably largely composed. These facts account for
the varying proportions of proteid precipitable with zinc sulphate,
recorded above, — proportions which in no way correspond to the
effects noted; for with Portion I, which contained 87.8 per cent of

On the Physiological Action of the Proteoses. 365

nitrogen precipitable with zinc sulphate, the influence of the injection
was very slight, while with Portion II, in which only 51.3 per cent
of the total nitrogen was precipitable with zinc sulphate, the effects
were very marked. Just what is the nature of this body is unknown,
but it is non-coagulable in neutral, and soluble in faintly alkaline,
solutions. It may be assumed, therefore, that 0.8 per cent. hydro-
chloric acid acting on edestin for a period of a week is not always
sufficient to produce proteoses enough to call forth any noticeable
effects after injection; while the products formed by a three weeks’
hydrolysis of edestin with the same strength of acid give all the
reactions typical of proteoses isolated from other sources.

Inasmuch as excelsin, the brazil-nut proteid, had been shown to
be without influence when injected, material was prepared from it by
hydrolysis with 0.8 per cent hydrochloric acid at a temperature
of 100° C. for a period of twenty hours, and was separated according
to the Pick and Spiro method. The product provoked effects of the
most pronounced character. Of the total nitrogen of the material
17.6 per cent was precipitable with zinc sulphate. Here no difficulty
was experienced in causing a complete precipitation of the acid-
proteid on neutralization, and in all probability the nitrogen precipi-
table by zinc sulphate is a true indication of the quantity of proteoses
present.

Injections of products formed from excelsin with hydrochloric acid.
Experiment 45.— A dog of 10.3 kilos received an injection (41.2 C.c.)
of 7.73 gms. (0.75 gm. per kilo) of the product described above, in a faintly
alkaline solution. The injection lasted 45 seconds. The animal re-
mained quiet.

The average clotting time of the blood before the injection was 12
minutes.

Blood samples collected at frequent intervals during 15 hours following
the injection were still liquid 24 hours later. There was a noticeable fall
in blood-pressure followed by a gradual rise.

A second experiment, in which a much smaller dose (0.38 mg. per
kilo) was given, was also carried out. The effects of this injection
were in every way as marked as in the experiment just given.

It has been shown above, that if edestin be treated long enough
with hydrochloric acid, there are formed products capable of calling
forth effects similar to those following proteose injection. Will
proteoses, if treated with acid until there is no longer any body
present giving the biuret reaction, yield any substance or substances

366 Frank P. Underhill.

which will induce analogous physiological reactions? The attempt
to answer this has been made with proteoses, prepared from edestin
by heating with 2 per cent sulphuric acid in a steam sterilizer for
nine hours.

1o gms. of the isolated proteoses ,were boiled with roo c.c. of 3 per cent
hydrochloric acid, for a period of 70 hours, after which no biuret reaction
could be obtained. This solution was neutralized exactly with 74 sodium
hydroxide, and a portion injected.

Experiment 36.— A dog of 8 kilos received an injection of 4o c.c. of the
above solution corresponding to 4 gms. (0.5 gm. per kilo) of proteoses,
in a faintly alkaline solution. The injection lasted 1 minute and 4o
seconds. The animal remained quiet. There was no influence upon
blood-coagulability, and only a slight fall in arterial pressure, which
immediately regained the normal. A second injection (49 c.c.) of
4 gms. (0.5 gm. per kilo) of the undecomposed proteoses to which
4o c.c. of 3 per cent hydrochloric acid were added and the mixture
neutralized with /%; sodium hydroxide, followed. The purpose of the
addition of the acid and its subsequent neutralization was to make the
salt content of the second preparation exactly comparable with that of
the first injection. The final reaction of the fluid was faintly alkaline,
and the injection lasted 48 seconds. The animal remained quiet.

The average clotting time of the blood before the injection was 11
minutes.

Blood samples collected 1 min., 11 min., 28 min., 66 min., and 74
min., after the injection, clotted in 4 hours 41 min., 6 hours 6 min.,
5 hours 50 min., 1 hour 22 min., and 54 min., respectively. The typical
blood-pressure phenomena were observed.

From these experiments it seems evident that products may be
formed according to the method of Pick and Spiro, cited above, which
are capable of inducing effects in the organism analogous to those
‘called forth by ordinary proteose injection. We find no occasion to
attribute these effects to products other than the proteoses present ;
and when the latter are split up into simpler compounds the peculiar
physiological reactions are no longer obtained.

ARE THE Toxic EFFECTS OF PROTEOSES DESTROYED BY THE PICK
AND Sptro MerHops oF ‘“ PURIFICATION ” ?

According to Pick and Spiro, the proteoses lose their toxic effects
when they have been treated with either alcohol alone or alcohol and
alkalies. In the present research some of the experiments of these

On the Phystological Action of the Proteoses. 367

investigators have been repeated, and the results obtained fail to con-

firm their observations. The purification of ‘‘ Witte’s peptone” was
first carried out.

60 gms. of “ Witte’s peptone ” were dissolved in one litre of equal parts of

distilled water and gs per cent alcohol, and boiled on a water bath under
a reflux condenser for 5 hours. After cooling, the insoluble portion was
filtered off and dried. The filtrate was evaporated to dryness on a
water bath. During the boiling of the substance sulphur derivatives
were liberated in considerable quantity. A piece of absorbent cotton
moistened with lead acetate and placed in the top of the condensing tube
was very readily turned black, owing to the formation of lead sulphide.

Influence of the insoluble portion. xferiment 34. — (Compare Pick and
Spiro, Experiment XVIII"). A dog of 10.5 kilos received an injection
(42 c.c.) of 5.25 gms. (0.5 gm. per kilo) of the znsoluble portion of
“ Witte’s peptone ” treated as above, in a faintly alkaline solution. The
injection lasted 30 seconds. The animal remained quiet, but 52 minutes
after injection suddenly died.

The average clotting time of the blood before the injection was 16
minutes.

Blood samples collected at frequent intervals after the injection showed
no tendency to clot 24 hours later. The blood-pressure fell from 105
mm. Hg to 9 mm., where it remained until the animal’s death.

Influence of the soluble portion. Exjeriment 37.— (Compare Pick and
Spiro, Experiment XIX”). A bitch of to kilos received an injection
(40 c.c.) of 5 gms. (0.5 gm. per kilo) of the soluble portion of “ Witte’s
peptone” treated as above, in a faintly alkaline solution. The injection
lasted 30 seconds. There was evidence of great excitation (cries, etc.),
followed by deep narcosis.

The average clotting time of the blood before the injection was 17
minutes.

Blood samples collected 1 min., 7 min., 21 min., and 23 min., after
the injection, had not clotted 24 hours later. The typical blood-pressure
phenomena were also observed.

The next experiment corresponds with Experiment XX? of Pick
and Spiro, in which “ Witte’s peptone” was injected after being
boiled ten hours with alcohol.

A litre of a 6 per cent solution (equal parts distilled water and 95 per cent
alcohol) was boiled on a water bath, under a reflux condenser for 10

PAL OG. CH 3..p-. 250. 2 Loe. Cibey Do 2b Je & Loc. cit, Pa 257-

368 Frank P. Underhill.

hours. At the end of this period the mixture was evaporated to dryness
upon the water bath. In this experiment sulphur compounds were also
liberated in considerable quantity during the boiling.

Experiment 38. —A dog of 9.2 kilos received an injection (36.8 c.c.) of
3.4 gms. (0.38 gm. per kilo) of ‘‘ Witte’s peptone ” treated ro hours with
alcohol, in a faintly alkaline solution. There was evidence of marked
excitation (cries etc.), which was followed by deep narcosis. The in-
jection lasted 30 seconds.

The average clotting time of the blood before the injection was 14
minutes.

Blood samples collected 2 min., 7 min., 32 min., 58 min., 1 hour 27
min., and 1 hour 41 min., after the injection, clotted in 7 hours, 1 hour,
36 min., 30 min., 17 min., and 15 min., respectively. The usual fall in
blood-pressure was observed.

Experiment 39,— A dog of 8 kilos received an injection (32 c.c.) of 4 gms.
(0.5 gm. per kilo) of the above product, in a faintly alkaline solution. The
injection lasted 30 seconds. The animal remained quiet.

The average clotting time of the blood before the injection was 9
minutes.

Blood samples collected at frequent intervals for 14 hours after the
injection showed no tendency to clot 24 hours later.

‘The blood-pressure fell from 140 mm. Hg to 24 mm., where it
remained.

From the preceding it appears that treatment of “ Witte’s peptone ”
with alcohol, according to the methods of Pick and Spiro, has little or
no detrimental influence on the active agent. In the experiment
(Experiment 38), where the toxic action of ‘“ Witte’s peptone”
seemed to be decreased somewhat after treatment with alcohol, may
not the fact that some decomposition of the products had taken place,
as indicated by the liberation of sulphur derivatives, suffice to account
for the lessened toxicity of the substance?

In a further series of experiments, protoproteose was separated
from heteroproteose in the manner described by Pick and Spiro.1
The original product was obtained from a peptic digestion of edestin.
The separation of the two primary proteoses from each other was
effected as follows: 10 grams of the mixture were dissolved in
120 c.c. of water, to which were added 2 volumes of 95 per cent
alcohol, and the whole boiled thirteen hours on a water bath, under

* Loc. cit., p. 255. See also Pick: Zeitschrift fiir physiologische Chemie, 1898,
XXVili, p. 21g.

On the Physiological Action of the Proteoses. 369

a reflux condenser. The mixture was allowed to cool, when a
flocculent precipitate (heteroproteose) separated. This was dried.
The filtrate (protoproteose) was evaporated to dryness on the water
bath. About 5 grams of each were obtained.

Injection of protoproteose. Hxferiment 40. — (Compare Experiment XV,
Pick and Spiro’). A dog of 8.3 kilos received an injection (33.2 c.c.)
of 4.15 gms. (0.5 gm. per kilo) of Arofoproteose, in an alkaline solution.
The injection lasted 30 seconds. The animal remained quiet.

The average clotting time of the blood before the injection was 15
minutes.

Blood samples collected at frequent intervals during the first hour after
the injection had not clotted 24 hours later. Samples collected during
the second hour after the injection clotted in 8 hours. The typical
phenomena connected with the blood-pressure were observed.

Injection of heteroproteose. xferiment 43.— (Compare Experiment XVI,
Pick and Spiro.”) A dog of 9.5 kilos received an injection (38 c.c.) of
4-75 gms. (0.5 gm. per kilo) of Aeteroproteose, in a faintly alkaline solu-
tion. The injection lasted 30 seconds. The animal remained quiet.

The average clotting time of the blood before the injection was 10
minutes.

Blood samples collected 2 min., 4 min., 29 min., 54 min., and 1 hour
27 min., after the injection, clotted in 38 min., 47 min., 36 min., 31 min.,
and 22 min., respectively. The fall in blood-pressure was typical.

A third method ? of purification as given by Pick and Spiro was also
repeated in the present investigation. According to these authors,
proteoses subjected to treatment with an alkaline alcohol solution for
twenty-four hours lose their toxic effects.

The proteoses employed in this experiment were “ hemialbumoses,” already

referred to under Experiment 18, page 355. ‘They were treated as
follows: 20 grams of hemialbumoses were dissolved in 200 c.c. of water,
and 2 volumes of g5 per cent alcohol added. The mixture was made
faintly alkaline with sodium carbonate, and boiled for 24 hours on a
water bath, under a reflux condenser. At the end of this time the solu-
tion was cooled, and a small precipitate which separated was filtered off.
The filtrate was evaporated to dryness on the water bath, and injected.
The precipitate noted above was too small for an injection.

Injection of ‘ purified” hemialbumose. A xferiment 20.— A dog of 6 kilos
received an injection (24 c.c.) of 3 gms. (0.5 gm. per kilo) of “ purified”

EL OGn Clic. Po 255. 2 TOG Clit De 255 S Loc. cit... pe 250:

370 Frank P. Underhill.

hemialbumose, in a faintly alkaline solution. ‘The injection lasted 55
seconds. During the injection and for a short period following there
was evidence of great excitation (cries, etc.), and this period of agitation
was succeeded by deep narcosis. Evidence was also obtained of in-
creased intestinal peristalsis.

The average clotting time of the blood before the injection was 13
minutes.

Blood samples collected 2 min. after the injection clotted in 6 hours
25 min. Samples collected 6 min. and 34 min. after the injection clotted
within 20 hours. Samples collected 2 hours 7 min., and 2 hours 54 min.,
after the injection, clotted in 62 min., and 24 min., respectively.

The blood-pressure fell from 118 mm. Hg to 30 mm., and 2 hours
later had risen to 70 mm., where it remained.

Other proteoses, obtained by peptic digestion of egg-yolk vitellin,
and separated in the manner usual in this investigation, were likewise
subjected to treatment with an alkaline alcohol solution. In this case
considerable sulphide was set free. Practically all of the material was
soluble. A protocol follows:

Injection of “purified” vitelloses. Lxperiment 23.— A dog of g.8 kilos
received an injection (39.2 c.c.) of 4.9 gms. (0.5 gm. per kilo) of the
above product, in a faintly alkaline solution. The injection lasted 45
seconds. There was evidence of great excitation (cries, etc.), which
was followed by deep narcosis. Defeecation occurred.

The average clotting time of the blood before the injection was 10
minutes.

Blood samples collected at frequent intervals during 2 hours following
the injection showed no tendency to clot 24 hours later.

The blood-pressure fell from ro8 mm. Hg to 28 mm., and at the end
of 2 hours had risen to 71 mm.

_ The results of the experiments here recorded justify the conclusion
that the methods of purification adopted by Pick and Spiro are of
questionable value, since the products so treated lose scarcely any of
their toxic properties.

In an article by Spiro and Ellinger’ the statement is made that
““somatose”’ is an albumose made without the use of enzymes, and
when injected is without influence on blood-coagulability.

The proteoses separated from such a preparation (Friedr. Bayer &
Co., Elberfeld, Germany), by precipitation with ammonium sulphate

L GEOC AGILE 5 PseiG On

On the Phystological Action of the Proteoses. 371

and dialyzed free from salt, provoked the effects typical for proteoses.
The proctocol follows :

Injection of somatose. /xferiment 28.— A dog of 7.9 kilos received an
injection (31.6 c.c.) of 3.95 gms. (0.5 gm. per kilo) of Arofeoses separated
from somatose, in a neutral solution. The injection lasted 34 seconds.
The animal remained quiet.

The average clotting time of the blood before the injection was 18
minutes.

Blood samples collected at frequent intervals during 40 min. after the
injection had not clotted in 8 hours, but within 20 hours were found
clotted. Samples collected 53 min. after the injection clotted in 51 min.

The blood-pressure fell from 125 mm. Hg to 70 mm., and within 4 min.
had risen to 100 mm., where it remained.

CONCLUSIONS.

A review of the experimental work presented in this paper leads us
to the conclusion that there is at present no occasion for attributing
the physiological effects following the injection of proteoses into the
circulation to the presence of contaminating substances derived from
animal tissue or elsewhere. Typical purified vegetable proteids
which when injected are themselves inert in this regard, yield on
hydrolysis with acids, or even water alone, proteoses which provoke
the characteristic reactions. The proteoses which are formed by the
action of proteolytic enzymes of vegetable origin (bromelin, papain)
on purified proteids, likewise alter zz vzvo the coagulability of the
blood and call forth the other well-known symptoms of proteose
injections. The proteoses occurring in nature in the vegetable
kingdom are similarly active, as was to be expected if the toxic
properties are a function of these products fer se. No method of
“purification” has been found which will deprive proteoses of this
characteristic physiological behavior in the circulation; when the
chemical make-up of the proteoses is profoundly altered and they
lose their chemical identity, the typical physiological action may
also be lost.

We regard the failure of other investigators to obtain toxic products
by hydrolysis of proteids with acids, etc., as attributable, in part at
least, to the fact that they injected mixtures of unknown composition
and probably containing either insufficient doses of proteoses proper,
or antagonistic compounds simultaneously present. It has, in fact,

372 Frank P, Underhill.

been demonstrated above that it is possible to prepare, by methods
with which Pick and Spiro failed to obtain positive results, products
which will provoke effects analogous to those obtained with isolated
proteoses. Incidentally we have again learned the differences in the
susceptibility of different animal species to the toxic action of injected
proteoses, by comparing the response in the dog with that in the cat.
Recent studies on immunity have shown equally striking natural
differences among animals, and have again emphasized the influence
which the mode of introduction of toxins may exert. Finally, there
is at present no reason for denying the existence of many chemically
unrelated substances which call forth apparently similar responses in
the animal organism.

This investigation has been carried out at the suggestion of
Professor Lafayette B. Mendel, to whom the writer desires to express
his obligation for help and criticism. The expenses of the research
were defrayed by an appropriation from the Rockefeller Institute for
Medical Research.

373

On the Physiological Action of the Proteoses.

TYPICAL BLOOD-PRESSURE TRACINGS SELECTED FROM THE

PRECEDING EXPERIMENTS.

(09¢ a8ed 99g) ‘pass-duiay wo1j asoajoid aaneu jo uonoaluy —‘T¢ Juawnedxy

a
NOILDICNI ‘NIN & Y314dV “NINO 431450 "NIN 6 U3LSV NIN SIMD NIN 274310

Y—

(‘Sg¢ aded aag) ,sasotunq|ermay,, Jo uonoafur yuonbasqns { uwinqe-339 poz[eisks9 Jo uoNoeluy — “eT Juotiiedxg|
NOILIZCNI 4Sui4 WOILIaFN| GNODSS

»—-

ek, Te Re a eS ae ie ae al ae ane

ON RIGOR MORTIS.

BY SOTTO FOLEN:

[From the Chemical Laboratory of the United States Fish Commission, Woods Hole, and
the Chemical Laboratory of the McLean Hospital for the Insane, Waverley, Mass.|+

UHNE is frequently credited with the authorship of the current
theory that rigor mortis is the effect of spontaneous coagula-
tion of certain proteins of muscle plasma after death. This is in-
correct. The coagulation theory of rigor mortis was advanced, and
in fact generally accepted, several years before Kiihne took up
the subject. Johannes Miiller,? for example, discussed the possi-
bility of muscle rigor being due to fibrin formation of lymph and
blood as far back as 1837. Miiller does not commit himself to the
theory, to be sure, but he clearly favors it as against the earlier
“contraction” theory. Five years later Briicke*® defended the view
that rigor mortis is due to the coagulation of a substance in the
muscle plasma which he conceives to be more or less identical with
the fibrinogen of blood plasma. It is undoubtedly largely due to
this paper of Briicke’s that the coagulation theory came to be the
most generally accepted view.

In 1858, one year before the publication of Kiihne’s first paper
on the subject, Ludwig wrote in his Lehrbuch der Physiologie des
Menschen, p. 472, as follows: ‘There is now scarcely any difference
of opinion in regard to whether muscle rigor is due to the coagula-
tion of a liquid protein substance, while there is every variety of
opinion in regard to the exact nature of that substance. On the same
page Ludwig writes: ‘‘ The objection which has been raised against
the coagulation theory, that no one has been able to press out of
fresh muscles the liquid fibrin substance the coagulation of which
produces rigor, is not well taken, as Briicke points out, because while

' The investigations which form the basis of this paper were begun in Woods
Hole during the summer of 1902, and I wish here to express my obligation to the
U.S. Fish Commission for the use of the laboratories, and for other conveniences
which were freely extended to me during my stay at Woods Hole.

* MULLER: Handbuch der Physiologie des Menschen, 1837, ii, p. 46.

° BRUCKE: Archiv fiir Anatomie und Physiologie, 1842, p. 179.

374

On Rigor Mortis. 375

the muscles are being subjected to the pressure necessary to accom-
plish such a purpose they go into rigor.”

It was at this point that Kiihne! obtained from fresh frog muscles
a muscle plasma which possessed all the properties that had been
demanded by the opponents of the coagulation theory. This plasma
not only coagulated at 40°, the temperature at which frog muscles go
into rigor, but, like blood plasma, it coagulated spontaneously at
ordinary room temperature. By proving the presence in fresh
muscles of such a coagulable protein and by showing that it is no
longer to be found in muscles which have gone into normal rigor,
Kiihne silenced all opposition to the coagulation theory of rigor.

Subsequent investigators, Halliburton, Firth,? G. N. Stewart and
Sollman,? and a very large number of other workers, have brought
out many interesting and important points in regard to the coag-
ulation of the proteins of muscle plasma and the conditions gov-
erning rigor. The fundamental proposition that rigor mortis is
due to the coagulation of certain muscle proteins remains, however,
much as Kiihne left it. And his statement (vc. cit., p. 773) that
“rigor mortis is a coagulation, and all other opinions in regard to it
are nothing but pure speculations because there is not a single reason
for assuming that any other process than coagulation is involved in
its production,” may still be considered practically unchallenged.

While the coagulation theory may thus be considered as well nigh
universally accepted it can scarcely be said that it has gained in
strength or clearness since it was first advanced as an explanation of
the phenomenon of rigor mortis. The admitted fact that muscle
rigidity may begin before any visible coagulation of the plasma has
occurred seems strangely inconsistent with the idea that coagulation
is the cause of the rigidity. The fact that the onset of rigidity may
be prevented and rigidity already produced may be made to disappear
by simply gently bending or pressing a muscle between the fingers
would seem far more in harmony with a “contraction theory” than
with the coagulation theory of muscle rigor. The disappearance of
rigor which by the above theory should be due to a redissolution
of the coagulated “myosin fibrin” inside the muscle fibres has
become, as Stewart and Sollman (doc. cit, p. 456) truly observe,
‘‘more mysterious than ever” since recent investigations have com-

1 KUuHNeE: Archiv fiir Anatomie und Physiologie, 1859, p. 788.
2 FirtTH: Beitrage zur chemischen Physiologie und Pathologie, 1903, iil, p. 543.
8 STEWART and SOLLMAN: Journal of physiology, 1899, xxiv, p. 427.

376 Otto Folin.

pletely failed to find any adequate explanation of this part of the
phenomenon}

The experiments to be described in this paper were undertaken for
the purpose of showing that the coagulation theory of rigor is in-
correct. The investigation was begun with fishes at Woods Hole,
but the facts here stated refer exclusively to later experiments made
with frogs’ muscles, since the original experiments of Kiihne which
have been taken to prove the coagulation theory were made with
that material.

Many physiologists, especially the writers of text-books, seem at
present to include in the term “rigor mortis” several different pro-
cesses, such as the evolution of carbon dioxide, of acids and of heat,
the shortening of muscles and their becoming ‘‘cloudy” or opaque.
As far as I am aware, however, it has never been shown that any of
these phenomena except the last stand in any causal relationship to
the fundamental phenomenon of rigidity and hardness or to the
coagulation process which is supposed to produce the latter. The
term “rigor” is therefore here taken in a more restricted sense, z. é.
as representing only that peculiar stiffening and hardening of muscles
occurring during or after cessation of life.

To prove that rigor mortis in this sense is not a coagulation
phenomenon I produce rigor in frog muscles by subjecting them to
a temperature of —15° to —20° C.

The fact that fresh muscles can be made to go into rigor by simply
lowering their temperature does not seem to be generally known.
It is, however, no new discovery. In 1842 Briicke (doc. cit., p. 186)
described rigor produced by cold and maintained that the rigor so
obtained is in every way identical with normal rigor. Briicke’s ex-
periments with cold as a means of producing rigor received, curiously
enough, no attention from subsequent investigators, and as far as I
have been able to learn the discovery has been forgotten.

Rigor mortis obtained in this way appears, however, much better
adapted to throw light upon the process that gives rise to it than is
“normal” rigor or heat rigor. In cold rigor we seem in fact to have
the fundamental phenomenon of rigor mortis isolated, so to speak,

‘ On this point see especially Firru’s last paper (doc. c7t., p. 549). FURTH
shows that the disappearance of the “muscle clot” is not accounted for by the
increased acidity of the muscle serum, nor can he find the least evidence for

HALLIBURTON’S view that this end is brought about by means of some ferment
reaction.

On Rigor Mortes. 377

from those other processes which are usually associated with the
onset of rigidity. That this should be so is not to be wondered at,
considering under what conditions cold rigor occurs in frog muscles.

The muscles of frogs can be cooled to —7° C., ze. they can be
frozen solid without losing their irritability and contractility when
carefully thawed. If the cooling of such frozen muscles is continued
longer and their temperature lowered to —15°, the power to recover
contractility is lost, and on thawing such muscles they will be found
to have gone into rigor. It seems clear that the chemical changes
involved in the transformation of the frozen but potentially living
muscle into the dead and rigid muscle must be relatively very small.
Such low temperatures and the absence of a liquid medium acting
together would tend to reduce toa standstill all those chemical re-
actions which are associated with constructive or destructive meta-
bolism, z.¢. oxidations, reductions and ferment or other hydrolytic
reactions.

Considering these conditions, it seems indeed remarkable that rigor
mortis should be produced by lowering the temperature of the muscles
from —7° C, to —15° C., but it is not surprising that few if any other
changes occur, and that the rigor so produced should appear to be a
more or less isolated phenomenon. Such is the case. The rigid
muscles are apparently not shortened, they are still neutral or faintly
alkaline to litmus, and they show absolutely no sign of cloudiness or
opaqueness which could point to coagulation as the cause of the
rigidity.

My experience so far indicates that it is not possible to freeze frog
muscles to a point where they have not acquired the rigidity of rigor
mortis and yet have lost the power of recovering contractility on
thawing. Should further experiments verify this observation it
would indicate that in this case the two occurrences — loss of con-
tractility and onset of rigor — may be more or less directly due to the
same cause. Later I hope to make further attempts toward studying
this phase of the problem. The investigations here described are
directed exclusively toward proving that the rigidity and hardness of
muscles in rigor are not due to a coagulation process.

The perfect translucency of muscles which have gone into cold
rigor as compared with fresh living muscles would itself seem to
disprove the theory that this phenomenon is due to a coagulation
process, or it would at least force one to the unwarranted assumption
that the coagulum itself must be translucent. Muscles of this kind

378 Otto Folin.

kept at ordinary room temperature remain both rigid and translucent
until they begin to decompose.

I have, however, not relied upon this transparency alone, but have
made quantitative tests as follows:

All the muscles from the hind legs of two frogs are removed and
divided into two equal portions. Each portion is then placed on a
filter paper or on a watch glass in the bottom of a tin can, and to each
portion is added one gastrocnemius muscle as control. One tin
can is placed on ice for two or three hours. The other is immersed
in a freezing mixture for the same length of time.

The frozen muscles are taken out first, and the control gastrocne-
mius is laid aside to thaw in order to show that rigor has actually
taken place. The rest of the frozen muscles are macerated in a cold
mortar and thoroughly rubbed up with 25 c.c. cold sodium chloride
solution (0.7 per cent). The resulting mixture is filtered and the
residue rubbed up with another 25 c.c. salt solution. The second
filtrate is added to the first.

The other tin can containing muscles kept on ice only is next
placed on the top of the freezing mixture and watched until the
muscles begin to solidify. They are then taken out at once and
macerated and extracted just as in the case of the frozen muscles.
The control gastrocnemius in this case will show that no rigor has
taken place. (Thesame results can be obtained without subjecting
the contents of the second tin can to any freezing at all, but it is
then much more difficult to macerate the muscles completely.)

All this work was done in my experiments in a large ice-room
which was further cooled by liquid ammonia pipes, but such pre-
cautions are not at all necessary for the success of the experiment,
because fresh muscle plasma does not coagulate spontaneously with
any great degree of rapidity.

It is clear that the two saline solutions obtained above should be
quite different if the rigor in the one set of muscles had been pro-
duced by the coagulation of dissolved protein. The two are, however,
absolutely identical except for the fact that the plasma solution ob-
tained from the frozen muscles is usually a trifle clearer than the
other. The solutions are tested as follows:

(a) Samples of each are set aside in test tubes at room tempera-
ture; both show the next day about the same amount of coagulum.

(6) Other samples of each are carefully heated side by side in
test tubes immersed in water. The contents of both tubes coagulate

On Rigor Moris. 379

at the same time at 40°-42° C. and the amount of coagulum is so
nearly alike in quantity as to make it impossible to distinguish one
from the other.

(c) 10 c.c. of each is titrated with tenth normal sodic hydrate
in the presence of phenolphtalein. Both show the same degree of
acidity, usually amounting to 0.4-0.6 c.c. 7 acid (which means
neutral or alkaline reaction to litmus).

(d) Finally the total nitrogen is determined in 5 c.c. of each
extract, and is found to be identical in both cases, thus excluding
absolutely the possibility of one extract containing any more protein
than the other.

The above experiments have been repeatedly performed, and always
with the same results. They prove, it seems to me, that muscle rigor
is independent of the protein coagulation which is usually observed
when muscles go into normal rigor, for since muscles can be made
to acquire the stiffness and hardness characteristic of rigor mortis
without the least indication of any coagulation of protein having
occurred, there is clearly no substantial reason for assuming that
those same characteristics are dependent on protein coagulation in
the case of normal rigor.

The coagulation theory of rigor seems to have originated in a
rather superficial analogy between the coagulation of blood and the
hardening of muscles. The only other evidence in its favor is the
Opaqueness which usually accompanies the rigidity, and the experi-
ments of Kiihne showing that the muscle plasma obtained from
muscles which have gone into rigor is different from the plasma
obtained from fresh muscles in that it does not contain the mother
substance of myosin fibrin the coagulation of which is supposed to
be the cause of the rigor. It is clear that this evidence in favor of
the coagulation theory loses all its force in view of the observations
recorded in this paper, and in view of the fact that the above positive
findings of Kiihne might readily be due to other chemical changes
occurring in frog muscles during the 40-50 hours that one must wait
for them to go into normal rigor.

It would seem that the method of producing rigor by means of cold
would offer a fruitful field for further study, and I mean to continue
the work from the point of view recorded in this paper.

ON THE FORMATION OF ,DEXTROSE IN METABOLISM
FROM THE END-PRODUCIS OF A PANCKEAIS
DIGEST OF MEAT:

By PERCY G. STILES ann GRAHAM LUSte

[From the Physiological Laboratory of the University and Bellevue Hospital Medical .
College. |

INTRODUCTORY.

HE key to the chemistry of the proteid molecule has been sought

not only in the behavior of proteid in the beaker and test-tube,

but also through observations on the metabolism of proteid in plant
and animal life.

An extended review of the literature concerned in our problem is
manifestly unnecessary, because of the completeness of Langstein’s
article, “ Die Bildung von Kohlehydraten aus Eiweiss,”! of Cremer’s
article on glycogen,” and of Pfliiger’s > more recent monograph upon
the same subject. Brief mention of important points only can be
given in this place.

Kiihne and Chittenden believed that the result of pancreatic pro-
teolysis was a production in equal amounts of hemi- and antipeptone,
of which the hemipeptone only was further resolved into amido acids.

Kutscher* showed that prolonged tryptic digestion completely
changed proteid into a mixture of amido acids such as may be obtained
by boiling proteid with acids. Cohnheim® discovered erepsin, a fer-
ment which rapidly splits albumoses into amido bodies within the
intestines ; and Loewi,®° after feeding a pancreatic digest consisting of
amido bodies, has been able to maintain a dog in nitrogen equilibrium
almost as well as after feeding proteid. Loewi concludes that he has

LANGSTEIN: Ergebnisse der Physiologie, 1902, i, p. 63.
CREMER : /éid., p. 803.
PFLUGER: Archiv fiir die gesammte Physiologie, 1903, xxix, p. I.
KUTSCHER: Die Endprodukte der Trypsinverdauung, Strassburg, 1899.
COHNHEIM : Zeitschrift fiir physiologische Chemie, 1901, xxxili, p. 451.

* Loewr: Archiv fiir experimentelle Pathologie und Pharmakologie, 1902,
xviii, p. 303.

1
2
3
4
5

380

On the Formation of Dextrose in Metabolism. 381

proved that a proteid synthesis from amido bodies takes place within
the organism.

The fact that various proteids yield amido bodies different in kind
and amount has led Emil Fischer to make a prolonged series of inves-
tigations in this direction. Kossel! first drew attention to the fact
that cleavage products of proteid, such as leucin, histidin, lysatinin,
lysin, and arginin, each contained the same number of carbon atoms
as dextrose, and compared an aggregation of amido acid radicles
forming proteid, with the analogous polysaccharides. At the same
time Kossel, in conversation with one of the present writers (L.),
declared his belief that these amido bodies are the mother substances
of the proteid dextrose found in diabetes. This idea was later ad-
vanced in a paper by F. Miiller,? who stated that if proteid could yield
50 per cent of leucin, as was found by Cohn, it could hardly contain a
sugar radical equal to 60 per cent, which was the amount of dextrose
formed from proteid in phlorhizin diabetes as determined by Reilly,
Nolan, and Lusk.?

Lusk# at this time suggested that leucin, an amido fatty acid, might
arise from dextrose molecules in the proteid complex. This view was
discussed by F. Miiller,> who considered it improbable, although inti-
mating that it was difficult to disprove.

The proteid of meat and gelatine yield 60 per cent of dextrose in
diabetic metabolism,® and Halsey? has shown the same to be true of
casein. But Halsey could not show that feeding leucin increased the
dextrose excretion in phlorhizin diabetes. Cohn ® believed that the
liver glycogen was increased in rabbits after feeding leucin, but these
experiments have never been considered conclusive. Recently
Stookey ® has fed various gluco-proteids to fasting hens and has
observed little glycogen formation in consequence.

At the time this research was commenced the problem offered was:
what amount of dextrose, if any, would be produced after feeding a

1 KossEL: Deutsche medicinische Wochenschrift, 1898, p. 58.

2 MULLER and SEEMAN: Deutsche medicinische Wochenschrift, 1899, p. 209.

8 REILLY, NOLAN, and Lusk: This journal, 1898, i, p. 395.

4 Lusk: This journal, 1899, iii, p. 153.

5 MULLER: Zeitschrift fiir Biologie, tgot, xlii, p. 545.

6 REILLY, NOLAN, and Lusk: Loc. cit.

7 HALsEy: Sitzungsberichte der Gesellschaft zur Beforderung der gesammten
Naturwissenschaften, Marburg, 1899, p. 102.

8 CouHN: Zeitschrift fiir physiologische Chemie, 1899, xxviii, p. 21I.

® STOOKEY: This journal, 1903, ix, p. 138.

382 Percy G. Stiles and Graham Lusk.

phlorhizinized dog with a digestive mixture of proteid decomposition
products? If a mixture of amido bodies yields dextrose, then the
synthetic origin of dextrose in metabolism as advocated by Pfliger,
F. Miiller, Halsey, and others, would be established, whether such
sugar arose directly from the amido bodies fed or indirectly from the
metabolism of proteid formed from them.

After this research was well under way, Knopf,! using the same
method, published an account showing the synthetic production of
sugar from asparagin fed to a dog.

METHOD.

In order to measure the production of sugar, phlorhizin diabetes
was induced. After one or two days of fasting phlorhizin was admin-
istered three times daily by hypodermic injections. The amount
given each time was 2 gms. dissolved in about 20 c.c. of 1.2 per cent
sodium carbonate solution. The injections were given at approxi-
mately equal intervals. On the third day of phlorhizin treatment
(the fourth or fifth of fasting) the metabolism usually reaches a level
which will be maintained with remarkable uniformity through several
subsequent days, the nitrogen elimination being very nearly constant
and the sugar appearing in quantity according to a ratio which does

not vary widely from that previously determined by Lusk (5 = 375).

The even conditions of phlorhizin diabetes gave us a base-level above
which the production of additional sugar from the pancreatic digest fed
stood out conspicuously.

The digest.2 — This was obtained by the pancreatic digestion of
washed meat which had been allowed to continue for fourteen months
with proper precautions against putrefaction. The resulting mixture
of products was a dark, syrupy fluid with but little sediment or sus-
pended matter. The taste and odor were pronounced but not foul.
The mixture gave only a dubious biuret reaction. As prepared for
our use, the material contained 1.33 per cent of nitrogen and gave a
depression of the freezing point of 1.72° C. This had a practical
interest, for the concentration of salts in such digests must have much
to do with the possibility of retaining them in the stomach. If they

1 KNoprF : Zeitschrift fiir physiologische Chemie, 1903, xlix, p. 123.
2 This was kindly furnished us by Prof. W. J. Gies of Columbia University,
to whom we express our thanks.

On the Formation of Dextrose in Metabolism. 383

are too concentrated they are sure to be vomited; if they are too
dilute, excessive volumes must be given. It is difficult to hit upon
a desirable mean. It was always necessary to give the fluid by a
stomach-tube, as none of the dogs would take it voluntarily.

Difficulties. — Only two of five experiments ran a successful course.
In two cases failure of the kidneys occurred, and death followed closely
upon the suppression of the urine. In another instance stubborn
vomiting compelled us to give up the trial. Fortunately the two
experiments which were carried through without mishap were closely
concordant.

EXPERIMENTS.

Experiment I, —The animal was a bitch weighing about 16 kg. On the
third day of diabetes the urine was analyzed and the sugar was found to
be high in proportion to the nitrogen, the ratio of 4.15 indicating that
the preliminary sweeping of sugar from the system had not been accom-
plished. ‘The urine of the following night gave a ratio of 3.94, and it
was considered that the diabetic condition had become sufficiently settled
to attempt the feeding of the digest. Just before this was done a final
sample of the urine from the bladder gave the D: N ratio of 3.66. ‘The
first portion of the digest introduced amounted to 200 c.c. (N, 2.66 gms.).
After an hour and fifteen minutes the dog vomited 150 c.c. of fluid,
which was saved, and an hour later returned to the stomach together
with 200 c.c. of the original mixture. The dog had now received 5.32
gms. of nitrogen in the form of simple digestive products. ‘There was
no more vomiting for four hours, when about 250 c.c. of fluid, only
slightly colored, was ejected. This was not returned, but analyzed for
nitrogen, and found to contain but 0.18 gm. ‘The quantity of nitrogen
fed and retained was estimated at 5.14 gms. The stomach remained
irritable, and there was vomiting during the following night, but apparently
only of water and mucus. The dog defeecated once in the middle of
the feeding period; the faeces were moderate in amount and did not
differ from those passed during fasting. ‘The analyses of the urine are
tabulated on page 384.

Experiment JJ, —The animal was a large bitch weighing 32.6 kg. After
three days of diabetes and six of fasting the urine was collected and
analyzed for two preliminary periods of twelve hours each. At the
beginning of the third period, 400 c.c. of the digest (N, 5.32 gms.) was
given and retained without any subsequent sign of digestive disturbance
or any diarrhcea. The urine of the feeding period and the twelve hours
following was analyzed. The findings are given on page 384.

384 Percy G. Stiles and Graham Lusk.

EXPERIMENT I.

be
D. -

I. Preliminary 12 hours! . 7.04 27.78 ear
II. Feeding period 12 hours 11.38 39.58 3.48
III. After period 12 hours. 8.22 28.00 3.40

1 The analyses in the preliminary stage were actually for a period of sixteen hours.
The figures have been reduced to a twelve-hour basis.
2 Ratio in last sample of bladder urine = 3.66: 1.

EXPERIMENT II.

First preliminary 12 hours .

Second preliminary 12 hours .

Feeding period 12 hours
After period 12 hours

DISCUSSION OF THE TABLES.

In both cases the nitrogen fed seems to have been quantitatively
eliminated. In Experiment I the excess of nitrogen excreted in the
feeding period as compared with fasting is 4.34 gms., and in the after
period the excess is 1.18 gms., a total of 5.52 gms. nitrogen excreted
for 5.14 gms. fed. In Experiment II the excess of nitrogen elimina-
tion in the feeding period over the preceding twelve hours is 3.93 gms.,
and in the after period 1.65 gms.,a total of 5.58 gms. nitrogen excreted
for 5.32 gms. fed.

In both experiments the sugar rose in the feeding period, in Experi-
ment I by 11.80 gms., in Experiment II by 12.62 gms. It is noteworthy
that the excess of nitrogen elimination extends over two periods, while
that of sugar is limited to one. This circumstance accounts for the
low D: N ratios in the after periods. The sugar had fallen to the
fasting level, while the nitrogen remained markedly above it. If

On the Formation of Dextrose tn Metabolism. 385

all the excess of nitrogen had appeared in the feeding-period, we should
have had ratios for that period of 3.10 and 3.33 respectively, and a
return in the final period to values of 3.97 and 3.80 in the two experi-
ments. It may be that a return to a higher ratio could have been
demonstrated if the urine had been collected during an additional
twelve hours. It is to be regretted that the desirability of doing this
was not recognized at the time.

CONCLUSIONS.

It appears that 5 gms. nitrogen fed in the form of the products of
pancreatic digestion may give rise to the formation of about 12 gms.
of dextrose, or D: N::2.4:1. The same amount of nitrogen fed as
native proteid would be expected to produce 18 to 19 gms. of sugar.
No light is thrown upon the question whether the sugar in our experi-
‘ments was formed after a proteid synthesis had occurred or more
directly from the amido-bodies. Neither have we any evidence as to
the relative importance of the several digestive products which were
fed. Further trials, made with the individual bodies present in such
mixtures and in pursuance of Knopf’s plan as applied to asparagin,
may clear up this matter. The experiment shows that it is impossible
for a large sugar radical to exist in the proteid molecule. It will be
noticed that the amido nitrogen fed was quantitatively eliminated, and
did not protect the body’s proteid as do meat and gelatine under
similar circumstances. It is interesting to possess new facts which
show how closely parallel is the course of the sugar metabolism after
feeding amido-bodies with that which follows upon a proteid diet.

ON THE QUESTION OF PROTEID SYNTHESIS
THE, ANIMAL, BODY.

By YANDELL HENDERSON anp ARTHUR L. DEAN.
[From the Sheffield Laboratory of Physiological Chemistry, Yale University.]

HAT the animal body is endowed with very limited capacities
for synthetic processes is a current and habitual assumption
rather than a principle demonstrated by experiment. For while the
corollary of this statement — that plants are solely devoted to synthe-
sis — has been found insufficient by botanists, physiologists, mainly
perhaps because of the experimental difficulties involved, have con-
tinued to accept the dictum of Liebig so far as it applies to animals
with little modification, and until recently with no serious question.

The recent discovery by Cohnheim! of the ferment ‘‘erepsin ” is
full of suggestion. Cohnheim’s investigations show that in the well-
known experiment of Neumeister — the supposed reconversion of
albumose to coagulable proteid by the action of intestinal mucosa —
the primary products of digestion are in fact further decomposed into
the ultimate products of proteolysis — into simple crystalline nitroge-
nous substances. If such an alteration take place to any consider-
able extent in the normal course of absorption, it is evident that
within the animal body there must be a mechanism capable of resyn-
thesizing these simple substances to form the proteids of the blood
and tissues.

The question thus raised has been put to the direct test of experi-
ment by Loewi. In the preliminary report of his experiments Loewi
stated that he had succeeded in maintaining a dog in nitrogenous equi-
librium on a diet in which all albuminous material had been replaced
by the products of the prolonged self-digestion of pancreas. In the
account of his completed investigations published recently, Loewi?
gives the details of repeated experiments all yielding a similar result.
This is the more striking because it was accomplished in spite of the

? COHNHEIM : Zeitschrift fiir physiologische Chemie, 1901, xxxiii, p. 451.
* Loewr: Archiy fiir experimentelle Pathologie und Pharmakologie, 1902,
xlviii, p. 303.
386

Proteid Synthesis in the Animal Body. 387

disturbances of the alimentary tract which ordinarily occur on a diet
of predissolved material. So successfully was this difficulty over-
come in one of the later experiments that the animal exhibited a gain
in weight and a plus balance between the income and output of nitro-
gen. Loewi interprets this retention of nitrogen as indicating that
the proteid cleavage products of the diet were resynthesized to supply
the needs of the tissues for proteid. The importance of the subject
seemed to warrant an immediate test of the validity of this view.
For this purpose— rather than as a mere repetition of Loewi’s
experiments — we have employed in the experiment detailed below
not the products of pancreatic digestion, but those resulting from
the action of a mineral acid on proteid by prolonged boiling.

Four kilos of lean beef were carefully freed from fat and connective
tissue and run through the meat-chopper. This material was placed
in a flask with two litres of water containing 240 c.c. of concentrated
sulphuric acid. The flask was heated in a steam sterilizer for twenty
hours, and then upon a sand bath, where its contents were boiled for
two hundred and fifty hours. The volume was kept at about five
litres, so that the strength of acid was from seven to nine per cent.
Repeated tests made during the boiling showed that a biuret react-
ing substance was continually being split off in small amounts from
the antialbumid which remained undissolved in the fluid. Finally,
therefore, this antialbumid was filtered off, washed by twice suspend-
ing it in water and filtering; and the combined filtrates were concen-
trated to three litres. At the end of this treatment the last trace of
the biuret reaction had disappeared, — a point which was determined
with the utmost care in all dilutions and with every possible modifi-
cation of the test, and among others that of Neumeister relied upon
by Loewi. The sulphuric acid was removed by means of hot satu-
rated baryta water and the slight excess of barium by a few drops of
dilute sulphuric acid. The fluid was concentrated to about two litres.
Its content of nitrogen was then determined by the Kjeldahl method,
and 100 c.c. pipetted off into each of eighteen small flasks. These
were stoppered with cotton wool and placed in the steam sterilizer
for an hour.

As it was essential that the animal’s calorific needs should be fully
supplied, and that the material used should be free from any albumi-
nous substance, we employed for the purpose the best quality of lard
and arrow-root starch. In each there was in the amounts fed daily
a barely detectable trace of nitrogen. The food was prepared each

Vandell Henderson and Arthur L. Dean.

388

‘eoyleip § Suljtw0 j

‘eoyiieip § Sur1m0 A
”

“BOY ILI

‘eOYlieIp § SuWo A

‘VJOS sox

‘WJOS sa0m@y f SuIZIUIO A

‘jeu
-10U sadaxy f SuIWIO A

‘TeWOU sI0x

“SUITWO A

”

‘Suse

*‘souod poy

80°0—

cL:0—

“WerIp Iod “ae

8S

co's

“wetp 1ed*ae

—_—_—_ ee

oO
N © tT 1m SO Eee conic

SZ +7 ”
€z judy

”

eNae
aourleq

“"N
‘jndjno
1230L

pSselo[ed

feel a

"YOIeIS

‘doo

“LNAWTYadXa AHL AO SLTASHAY GNV SNOILIGNOD DNIMOHS HTaV.L

‘CO6I
‘ALVC

Fasting.

35

iT

“ce

0.83

85

Fed bones.

“

Proteid Synthesis tn the Animal Body. 389

1 The figures for the calorific value of the food are calculated for the starch and lard only,— that of the nitrogenous substances

as
mo
a
Q,
=)
n
te
3
a
o
o
=
~
1S)
ovo
is
n
o
pa
wn
-—
io}
—
c
-
n
ae)
o
o
i=
A
cS
=|
=
=
fon]
o
S
=)
ra)
Ss
=)
~
=
°
a
n
°
pat
>
—_
o
o
2
=
=
vo
>
=
on
o
-
os
4S
=
fos]
a
Ee
(e}
(=|
a.
<
=
on
=
Oo
Q

For the sake of a safe basis of calcu-

2 The total nitrogen in the feces between April 24 and May 13 (nineteen days) was 9.6 grams.
lation it is assumed in the table that one twelfth of this amount, or 0.80 gram, was contained in the faeces of each day of feeding.

utilization of the nitrogen of the food was very poor,
filter paper in the food amounted to 154 grams in all.

The

The

The total weight of the air-dry feces was 174 grams.

only 83 per cent.

6, 5 per cent; K,HPO,, 70 per cent;

Fe,Cl

3 On and after May 6 the food contained each day 2 grams of the following mixture:

NaCl, 15 per cent ; CaH PO,, 6 per cent; MgCly, 4 per cent.

day by heating the contents of
one of the flasks in a casserole and
adding to it the starch which had
previously been ground and mixed
with a small quantity of cold water.
While this mixture was still warm
the lard was stirred in, together
with enough finely divided filter
paper to give a decidedly thick
consistency to the mass. During
the greater part of the period of
feeding the animal ate this food in
two portions daily, voluntarily, and
even with apparent enjoyment, in
spite of the intensely bitter taste
of the proteid products. Toward
the end of the experiment, how-
ever, considerable persuasion, and
finally force, was necessary to in-
duce complete ingestion.

For the investigation we em-
ployed a lean fox-terrier bitch. It
was kept in a cage suitable to col-
lecting any urine that might be
passed. Catheterization was per-
formed each morning. The feces,
consisting largely of filter paper,
were united for analysis, bones
being fed as a means of demarca-
tion at the beginning and end of
the nineteen days the animal
was under observation. Whenever
vomiting occurred the material
was returned to the animal's
stomach immediately, — which ex-
cept for the filter paper would
have been impossible. Although
vomiting and diarrhoea occurred
on the majority of the days of
feeding, these conditions were not

390 Vandell Henderson and Arthur L. Dean.

so marked as to invalidate the essential accuracy of the data
presented.

Before and after the period of feeding the animal was allowed
to fast for a few days in order that from the amount of nitrogen
eliminated in the urine under these conditions the “ hunger mini-
mum” or waste of the tissue proteids might be determined.

The results of the investigation are given in the accompanying
table (page 388).

The numerical results here presented are essentially similar to
those obtained by Loewi. Even in the earlier days of feeding the
nitrogen elimination was not increased by the full amount of the
nitrogen ingested. From the ninth to the thirteenth day of feed-
ing the animal was in nitrogenous equilibrium. Body-weight was
maintained.

It seems fair to conclude, therefore, (1) that the nitrogenous sub-
stances in the diet were not immediately and wholly converted into
urea and excreted; (2) that they were on the contrary to a con-
siderable extent retained; and (3) that that portion which was
expended (appearing in the urine) exerted a marked proteid-sparing
action. These reactions seem to us, however, to afford a sufficient
explanation of the facts without invoking the more radical hypothesis
of proteid synthesis. The diminution in the nitrogen excretion from
1.9 grams in the fore period to 0.8 gram in the after period of fasting
suggests that the protoplasmic waste had not been made good, and
that the retention of nitrogen is not in itself a proof of proteid
synthesis.

OBSERVATIONS ON THE URINE OF THE MUSKRAT
(FIBER ZIBETHICUS).

By ROBERT BANKS GIBSON.
[From the Sheffield Laboratory of Physiological Chemistry, Yale University.]

Ov knowledge of the comparative chemistry of the urine is so
fragmentary that it seems desirable to present a few observa-
tions which the writer has had the opportunity of making at Professor
Mendel’s suggestion on a rodent, Fiber zibethicus. Except in the
case of a few of the domestic animals, only occasional statements
regarding the composition of mammalian urine are to be found in
physiological literature. It is assumed, from the data available, that
in general the urine of carnivora corresponds in chemical makeup
with that of man; whereas the urine of herbivora is characterized by
an occasional deficiency in phosphoric acid, a reaction usually alka-
line to litmus and a high content of hippuric acid and other aromatic
compounds — all depending largely on the peculiar character of the
diet of these animals.!_ That specific peculiarities in the metabolism
of individual species may occur is illustrated in the case of the dog
by the occurrence of kynurenic acid, a substance not yet found in
any other animal.

Regarding the dietetic habits of the muskrat there seems to be
a difference of opinion among writers.2 Our specimen, an appar-
ently full-grown animal weighing 800 grams, readily ate such vege-
table foods as apples, carrots, lettuce, and corn; and it likewise
consumed animal tissues in the form of lean meat, pancreas, and
liver,

General characters of the urine. — The dazly volume of the urine
varied from 54 to 205 c.c., the average of many days being 118 c.c.
This quantity is considerably larger per unit of body weight than

1 Cf. HOPKINS: SCHAFER’S Text-book of physiology, 1898, i, p. 637-638.

2 Cf. MENDEL and JACKSON: This journal, 1808, ii, p. 27.

3 “Tt dives with great facility, feeding on the roots, leaves, and stems of water
plants, and on the fruits and vegetables growing near the margins of the streams it
inhabits” (Encyclopedia Britannica). BEDDARD writes: “Thus it appears that
this rodent, like so many others, is largely carnivorous” (Cambridge Natural
History, 1902, x, p. 478).

391

392 Robert Banks Gibson.

the output of the common laboratory mammals. Thus Alezais! noted
an average urinary secretion amounting to 8 per cent of the body
weight in the guinea pig, a rodent somewhat comparable with the
muskrat in regard to its dietetic habits and its structural characteris-
tics. Whether this peculiarity is in any way related to the muskrat’s
aquatic habits could not be determined definitely. It was occa-
sionally noted that after a bath the animal eliminated somewhat
larger volumes of urine than at other times. But Sacc? has called
attention to the relatively large volume of urine secreted by a
terrestrial rodent, the marmot (Arctomys). He obtained as much
as 25 to 36 per cent of the body-weight, and suggests that this
noticeable renal activity is dependent on the insignificant loss of water
by way of the lungs and skin in the rodents. Corresponding with
the large output of urine in the muskrat, the specific gravity of the
secretion was usually rather low, being in general below 1.010, and
varying from 1.006 to 1.016. The urine emitted a faint musk-like
odor which became more pronounced after heating. The color of the
fresh urine was pale yellow, except during fasting or on a meat diet,
when a deeper color was observed, even in dilute urine. When kept
with thymol in a stoppered vessel exposed to the light, the urine
gradually grew darker, becoming almost black. In the dark, how-
ever, no appreciable change in color was noticed. The reaction to
litmus was acid during fasting and after carnivorous diet; at other
times it was only faintly alkaline or neutral, always remaining acid to
phenolphthalein. The average acidity to the latter indicator was
equivalent to 70 c.c. #4 acid per litre. Neumeister® has observed
the urine of the primitive mammal, Echidna aculeata, to remain
neutral after feeding meat and eggs, although it is usually assumed
that the urine, even of herbivora, becomes acid in reaction to litmus
on a carnivorous diet. The muskrat urine was ordinarily free from
any sediment; in the concentrated alkaline secretion crystals of
triple phosphate and of calcium oxalate were occasionally found.
In the acid urines uric acid crystals separated out at times.
Composition of the urine. — Several litres of urine collected during
a period of mixed diet yielded the analysis reported below.* Further

' ALEzAis: Archives de physiologie, 1897, ix, p. 576.

2 Sacc: Comptes rendus, 1872, Ixxv, p. 1839.

% NEUMEISTER: Zeitschrift fiir Biologie, 1898, xxxvi, p. 77.

* The samples were preserved with thymol which Cronheim (Archiv fir
Physiologie, 1902, p. 262) has lately shown to be most satisfactory.

«

Observations on the Urine of the Muskrat. — 393

data were obtained by collecting the urine in daily periods with various
experimental diets.

The total N was estimated by the Kjeldahl-Gunning process; uric
acid by the Ludwig-Salkowski method; urea by the method of
Morner and Sjoquist, which seemed trustworthy in the absence of any
considerable quantities of hippuric acid ; 1 ammonia by Folin’s method ;
oxalic acid by the process of Autenrieth and Barth; * and the other
constituents by the commonly described methods.

ANALYSIS OF MUSKRAT URINE.

Grams per litre. Grams per litre.
PotalgNieees 50s) a Sea Oe Phosphoric acid (P,0;) . 0.69
(Urea a 6 ee peion meas P,O; with alkali earths . 0.45
Hippuricacid . .. . 0.25 Sulphuric acid (SO3) . . 0.48
MRCIACIG Meta cuwsl ose) 0:22 EitherealsS@xs-) a-8 0:02
Oxalicacid’ ~ =... . =. 0:04 Chiorine (as NaCl) . . 0.40

To ascertain something regarding the average daily output of the
more important constituents, the urine was carefully collected for
four days during a mixed diet somewhat larger in quantity and richer
in meat than in the case above. The output for this period was
423 c.c., having a specific gravity of 1.011. The data obtained
follow.

Grams per day. | Grams per litre.

otal nitrogen 52 2 «4. . 0.808 7.62

WEARE MC ete) hea «ate Scobie 1.635 15.43
Wirieracidy 52.25) ose 0.037 0.35
LSTHTONES G WARREN ceo A eome 0.008 0.08
Manchin iDOGIeS: sie: « s) seoc 0.01
Phosphoric acid (P,0;) . . . ! 0.83

1 Cf. MoOrneEr: Skandinavisches Archiv fiir Physiologie, 1903, xiv, p. 279.
2 AUTENRIETH and BARTH: Zeitschrift fiir physiologische Chemie, 1902,
XXXV, Pp. 327.

394 Robert Banks Gibson.

The distribution of nitrogen in the various types of nitrogenous
constituents in the two experiments was as follows: —

Exp.1. Exp. 2.

per cent. per cent.

Urea nitrogen (sah E>. 23) 91.3 94.3

Winieacidinitrogen™ <0 -ee e 15 1.5
Ammonia nitrogen. . .. . Suse 0.9
Hippuric acid nitrogen

Xanthin bodies nitrogen .

Undetermined

Comments on the analytical data. — From the figures presented the
nitrogenous metabolism of the muskrat is seen to resemble quantita-
tively that of the guinea pig, the average output per hundred grams
of animal being 0.1 gm. N.! The small proportion of nitrogen in
the form of ammonia compounds (one per cent) corresponds with
previous experience on other herbivora? and with prevailing ideas
regarding the origin of ammonia in metabolism. Although uric acid
was excreted in noticeable quantity, the figures are no larger than
those obtained by Mittelbach for the urine of various herbivora.2 He
found quantities ranging from 90 to 450 mgm. per litre. It will be
noted above that an increase of uric acid in the muskrat urine
followed the more liberal introduction of meat into the diet. Neumeis-
ter‘ failed to find uric acid in the urine of Echidna under similar
conditions. The elimination of phosphoric acid noted is of interest
in view of the current statements regarding the low output in the
herbivora.®

Since in the dog and cat the ingestion of tissues rich in nucleic
acid is followed by a large elimination of allantoin, pancreas feeding
was tried in the case of the muskrat, with negative results. Similar

1 ALEZAIS: Loc. cét., p. 579:

? HupPPERT: Analyse des Harns, 1898, p. 42.

* MITTELBACH: Zeitschrift fiir physiologische Chemie, 1888, xii, p. 466.

4 NEUMEISTER: Loc. cit.

° Cf. Sacc: Loc. cit. (marmot). Also BERTRAM: Zeitschrift fiir Biologie,
1878, xiv, p. 335.

Observations on the Urine of the Muskrat. 395

experience has followed earlier experiments on rabbits in this labora-
tory! Kynurenic acid was also missing and creatinine could not be
isolated. The urine contained urobilin ; at times bile pigments were
present, giving Huppert’s test and imparting a greenish tint to the
fluid. Proteids and sugar were never detected.

* MENDEL, UNDERHILL, and WuITE: This journal, 1903, Vili, p. 399.

SALIVARY DIGESTION IN THE STOMACH|!

By W. B. CANNON aAnp H. F. DAY.
[From the Laboratory of Physiology in the Harvard Medical School.]

N stating the functions of saliva the common reference is to its
importance as a lubricant in facilitating the movement of the
parts of the mouth upon one.another and in aiding the passage of the
food through the cesophagus. There is a widespread impression
that the chemical function is slight. Saliva can indeed change starch
to sugar, but during mastication there is little time for amylolysis,
and in the stomach the action of ptyalin is soon stopped by the acid
gastric juice. Such is the view commonly expressed.2_ This paper
is a report of a critical inquiry into the reasons for this view, and an
experimental study of the degree of salivary digestion in the cardiac
and pyloric ends of the stomach.

To support the conclusion that salivary digestion in the stomach is
slight, certain studies of gastric contents after feeding starches are
cited, and attention is called to the common conceptions of the effect
of gastric peristalsis. Briicke ® found in the stomach contents of dogs
killed one to five hours after eating starch paste or mush, no sugar
or only slight amounts, unless it was already present in the food
ingested. In 1877 v. Mering* concluded, after feeding starch paste
to dogs, that the saliva of dogs was without digestive influence,
evidently because ptyalin was absent, and that in man, although
ptyalin must be considered, it is of secondary importance. Obser-
vations on human beings by Ewald and Boas® in 1886 seemed to

1 A preliminary report of this research was presented to the American Physi-
ological Society in December, 1902, and was printed in the proceedings of the
meeting, This journal, 1903, viii, p. xxviii.

2 See, for example, NEUMEISTER, Lehrbuch der physiologischen Chemie, Jena,
1897, pp. 287, 288.

8 BRUCKE: Sitzungsberichte der kaiserlichen Akademie der Wissenschaften,
Wien, 1872, Ixv, p. 144.

* v. MERING: Archiv fiir Physiologie, 1877, p. 393.

® EwaLp and Boas: Archiv fiir pathologische Anatomie und Physiologie, und
fiir klinische Medicin, 1886, civ, p. 296.

396

Sahvary Digestion in the Stomach. 397

sustain v. Mering’s contention, for after giving patients starch paste
to drink only slight amounts of sugar were found in the stomach.
The next year Seegen+ reported experiments in which he fed dogs
carbohydrate food of diverse forms (flour cakes, potatoes, rice), and
yet in the gastric contents failed to find more than a trace of sugar.

These chemical experiments tending to prove that saliva has only
slight action in the stomach were supported by observations and
theories of the churning effect of gastric peristalsis. Beaumont’s?
well-known description of the movement of the bolus along the gastric
walls, and Brinton’s? theory of peripheral and axial streaming of the
food in the stomach offered reasons for believing that all the food is
rapidly mixed with gastric juice and is thus in a short time made acid
throughout. The duration of salivary digestion must therefore neces-
sarily be brief.

The conclusion from the above evidence that salivary digestion in
the stomach is slight is not above criticism. In the first place the
experiments of Briicke, v. Mering, and Seegen were performed upon
dogs. The absence of sugar from the stomach, noted by these
observers after giving a carbohydrate meal, is amply explained, as
v. Mering admitted, by the fact that the dog’s saliva does not possess
a diastatic ferment. Even with ptyalin present in the saliva, which
was the case in the experiments of Ewald and Boas on human beings,
the starch paste as a test food would be objectionable. It is not
only not palatable, but is in a state to be swallowed immediately
without being chewed. There are thus two reasons for its not being
mixed with sufficient saliva to cause noteworthy amylolysis. In-
asmuch as there was no ptyalin, or almost no ptyalin, mixed with the
food given in all these experiments, it is clear that they do not furnish
direct evidence that salivary digestion in the stomach must be slight.

The indirect evidence against gastric amylolysis derived from
Beaumont’s observations and Brinton’s reasoning as to mixing cur-
rents, is also subject to criticism. In 1898, Cannon ® brought forward
proof that neither Beaumont’s nor Brinton’s account of the movement
of the food in the stomach was correct. A marked difference was
observed between the effects of the mechanical activities of the pyloric

1 SEEGEN: Archiv fiir die gesammte Physiologie, 1887, xl, p. 46.

2? BEAUMONT: Physiology of digestion, Plattsburgh, 1833, p. IIo.

8 BRINTON: The diseases of the stomach, Philadelphia, 1865, p. 24.

4 See GRUTZNER: Archiv fiir die gesammte Physiologie, 1876, xii, p. 285.
5 CANNON: This journal, 1898, i, p. 378.

398 W. B. Cannon and H. F. Day.

and the cardiac portions of the stomach. In the cardiac portion there
is no peristalsis, and the food, held in the tonic grasp of the gastric
musculature, shows no signs of movement ; in the pyloric portion, on
the other hand, gastric peristalsis mixes the food thoroughly with the
digestive juices. Little pellets of starch paste lying in the cardiac
end immediately after the food’ is ingested may be seen with the
X-rays in the same relative positions for almost two hours; the
pellets in the pyloric end are moved to and fro by the passing waves
of constriction. There is moreover a marked difference in the
appearance of the contents of the two parts of the stomach after
digestion has proceeded for even thirty minutes: the food in the
pyloric end is usually far advanced toward chymification ; the food in
the cardiac end has still its original appearance, which it may retain
even an hour and a half after ingestion. This observation that the
food in the cardiac end of the stomach is not moved, and therefore is
not mixed with the gastric juice, suggested that the cardiac contents
might retain their original chemical reaction for a considerable period.
Tests made on several cats and dogs, from one to one and a half hours
after feeding, showed that the contents of the pyloric end were in-
variably strongly acid; the surface of the cardiac contents was also
acid, but the internal mass in the cardiac end remained unchanged
in its reaction. Inasmuch as salivary digestion may continue so
long as free acid is absent, the conclusion was drawn that salivary
digestion might proceed in the fundus for an hour and a half or
longer without interference by the acid gastric juice.

The observations and the conclusion just detailed have been con-
firmed by Oehl,! and still more completely by Heyde working under
Griitzner? in Tibingen. Rats, rabbits, guinea pigs, and cats were fed
by Heyde with different kinds of food in definite amounts and killed
at different intervals after eating. The stomachs were carefully
removed and frozen; sections were made through the frozen gastric
contents, and acid indicators mixed with the food revealed at once
the extent of acidification. The inner layers of the food in the
cardiac end retained for hours a neutral or weakly alkaline reaction.
Only the outer layers were slightly acidified and digested. Thus the
diastatic ferment of the saliva may continue its action in the fundus
for a long period wholly unhindered.

The difference in motor activities between the cardiac and pyloric

' OEHL: Archives italiennes de biologie, 1899, xxxii, p. 114.
* GRUTZNER: Reprint from Deutsche Medizinal-Zeitung, 1902, No. 28.

Salivary Digestion tn the Stomach. 399

ends of the stomach has been noted also in human beings. Beau-
mont’s observations on Alexis St. Martin hint at a difference, and
v. Pfungen! and Moritz? have made it certain that in man the
peristalsis is confined to the pyloric end. Since Beaumont’s time
few observations have been made on the movements of the food in
the human stomach. In 1897, however, Hemmeter ? stated that after
many experiments on animals and investigations through fistula into
the human stomach, he found no confirmation of Beaumont’s and
Brinton’s views as to mixing currents. In 1900 Hemmeter* repeated
this statement and affirmed that Cannon’s assertion that the food is
churned by peristaltic waves only in the pyloric portion was correct,
After thus bringing confirmatory data to prove the fundus free from
peristalsis and after entirely rejecting the idea of mixing currents, he
still declares that the contents of the fundus are soon mixed with the
gastric secretions. The only explanation offered for this conclusion is
that food taken from the stomach by means of an Einhorn bucket
always contains gastric juice. Since the dorsal wall of the stomach
in man slopes forward from the cardiac opening, a tube introduced
through the cesophagus comes immediately in contact with the dorsal
wall, and will remove first the food lying in proximity to the secreting
mucous membrane. Clearly this faulty evidence allows no conclusion
to be drawn as to the presence of acid in the interior of the cardiac
mass, and such evidence cannot be accepted as contradicting the
accumulated testimony, from much more exact methods, that the
internal food in the fundus long remains unmixed with acid.

The discussion thus far has led to two conclusions: (1) that the
reasons commonly given to prove salivary digestion possible in
the stomach for only a short time are by no means convincing, and
(2) that in the fundus, a reservoir in which the food may rest for
hours unmoved, salivary digestion may indeed continue for a long
period without interruption.

Several researches not widely quoted lend support to the view that
salivary digestion in the stomach may beof importance. As long ago
as 1880, von den Velden ® pointed out that free hydrochloric acid did
not appear for almost an hour after eating breakfast and for almost two

1 y. PFUNGEN: Centralblatt fiir Physiologie, 1887, i, p. 220.

2 Moritz : Zeitschrift fiir Biologie, 1895, xxxii, p. 339-

8 HEMMETER: Diseases of the stomach, Philadelphia, 1897, p. 86.

4 HEMMETER: Loc. cit., second edition, 1900, pp. 85 and 88.

5 vy. D. VELDEN: Deutsches Archiv fiir klinische Medicin, 1880, xxv, pp. 105
and IIt.

400 W. B. Cannon and H. F. Day

hours after eating a full mid-day meal. Austen! in 1899, after study-
ing one of his students, stated that for a period of at least one hour
after an ordinary meal the albuminous matter of the food united with
the hydrochloric acid of the gastric juice as fast as the acid was se-
creted and thus prevented it from stopping the digestion of starch.
Hensay? and Miiller® were the first to present quantitative measure-
ments of the amounts of sugar and dextrins which might be formed
in the stomach when food is carefully chewed. These observers fed
rice mush made pleasing to the taste with meat extract and butter.
After a certain time, usually one half hour, they pressed out the gas-
tric contents as completely as possible and secured the remnant by
washing. They found that a large part of the expressed carbohydrates,
59.4 to 79.6 per cent, had been made soluble by the saliva. Of the
dissolved carbohydrates over one half, even two thirds, consisted of
maltose and of dextrins closely related to maltose, and the remainder
of dextrins more nearly related to starch.

The four reports just mentioned combine to show that the common
teaching that saliva must be unimportant in the stomach, because its
action is quickly stopped, is not correct. Saliva certainly may be im-
portant in the digestion of starch if the food is well chewed and thus
thoroughly mixed with the ferment. The observers who have brought
forward these positive results have not, it will be noted, regarded the
differences between the pyloric part of the stomach and the fundus.
It was the purpose of the research here reported to study the condi-
tions of salivary digestion particularly in the two portions of the
stomach: in the active pyloric portion where the food is soon mixed
with the acid juice; and in the quiet cardiac reservoir where the food
lies unmoved for a long time until it begins to be forced gradually into
the churning mechanism at the pyloric end.

METHOD.

The cat was the animal used in this investigation. The stomach
of the cat, as observations with the X-rays have proved, is like the

' AUSTEN: Boston medical and surgical journal, 1899, cxl, p. 325.

2 HENSAY: Miinchener medicinische Wochenschrift, 1got, xlviii p. 1208.

® MULLER: Sitzungsberichte der physikalisch-medicinische Gesellschaft, Wiirz-
burg, Igo!, p. 4.

* CANNON: This journal, 1903, viii, p. xxii; Roux and BALTHAZARD:
Archives de physiologie, 1898, xxx, p. 85; WuILLIAMS: The Réntgen ray in
medicine and surgery, New York, 1901, pp. 360-372.

Sahvary Digestion in the Stomach. 401

stomach of the dog, rat, rabbit, guinea pig, and man, in being separ-
able into two parts, — the quiet cardiac end and the active pyloric end.
Moreover the mucosa of the cat’s stomach resembles that of the dog
and of man, not only in structure but also in pouring out an active
secretion from almost every part of its surface. The cat may be
regarded, therefore, as a fairly typical animal for the purposes of the
present investigation.

From twenty-four to thirty-six hours previous to an experiment the
cat was allowed to fast, in order that the stomach might be free from
food and prepared to receive the test meal. Crackers, examined and
found free from sugar, were used as the test food. Unless otherwise
stated a uniform amount of coarsely powdered crackers, 30 gms., was
mixed with a uniform amount of filtered human saliva, loo c.c. The
mixture has a consistency of thick mush, similar to that of food chewed
to a degree suitable for swallowing. The food thus prepared was
given in one of two ways: it was either mixed in small amounts and
fed immediately to the animal, or the total amount of crackers was
mixed with the total amount of saliva and introduced at once into the
stomach by a stomach tube. As the results were the same with both
methods, the latter, because more expeditious, was usually employed.

After the food was in the stomach the animals were allowed to
live one half hour, one hour, one and a half hours, or two hours.
At the end of the time the animal was quickly etherized, the abdomen
opened along the middle line and along a line toward the left parallel
with the greater curvature of the stomach. With a curved hook a
ligature was passed around the stomach at the region where the
peristaltic waves start toward the pylorus,? and tied firmly. Thus
the contents of the cardiac and pyloric portions were separated.
The pylorus and cardia were next ligatured and the stomach re-
moved with as little handling as possible.

Openings were now cut in the stomach wall and the contents of each
part were emptied into an evaporating dish. The food in the pyloric
end invariably had a consistency of thin mush. Occasionally hair or
remnants of meat were found in the pyloric end; cases in which these
disturbing factors were present have been excluded from this report.

1 OprEL : Lehrbuch der vergleichenden mikroskopischen Anatomie der Wir-
belthiere, erster Theil, Der Magen, Jena, 1896, pp. 408, 443, 463; CARVALLO:
Article “ Estomac,” Dictionnaire de physiologie, edited by Richet, Paris, 1902, v,
p. 818. :

2 CANNON: This journal, 1898, i, p. 364.

402 W. LB. Cannon and Ff. fF. Day.

The consistency of the food in the cardiac end was very different
from that of the food in the pyloric end. It was not fluid, like the
food in the pyloric end, and it often retained its shape sufficiently
to permit the surface food to be separated from the internal mass.

The contents in the evaporating dishes, if not fluid, were slightly
diluted with water, and the enzyme action quickly stopped by heating
to the boiling point. The food was now evaporated to dryness by
steam heat; the dry residue was ground to a fine powder in a mortar
and kept in a desiccator until its weight was constant. One gram,
removed from each part of the dried and powdered stomach con-
tents, was mixed with 100 cc. distilled water. After standing for
about one half hour the mixture was filtered, and the filtrate was
poured several times through the residue on the filter. In each in-
stance 25 c.c. of the filtrate were taken for the sugar test. The sugar
content, determined according to Allihn’s method,! was estimated as
maltose.2 It was thus possible to know the percentage of sugar in the
dry residue of the stomach contents.

FACTORS TO BE CONSIDERED.

The first and most important of the factors to be considered is the
presence of free hydrochloric acid. As already stated, experiments in
1898 proved that free acid was absent from the middle of the cardiac
mass for one and a half or two hours after eating. Similar tests made
during the present research have confirmed the earlier observations.
At the end of one half hour after a carbohydrate meal, free hydro-
chloric acid is present in the pyloric end of the stomach, but at the
end of two hours there is no free acid present in the middle of the
cardiac food.

Another factor to be regarded is the rapidity of salivary digestion.
The common statements as to the speed of the change from starch
into sugar are based upon observations on starch paste.? Tests with
other forms of starchy food show in some cases a much slower rate.
Observations made on the test material used in this research, under-

ALLIHN : Zeitschrift fiir analytische Chemie, 1883, xxii, p. 448.

* MuscutLus and v. MERING: Zeitschrift fiir physiologische Chemie, 1878-9,
ii, p. 409; EwaLp and Boas: Loc. cit., p. 297.

* CHITTENDEN and Ey: Journal of physiology, 1882, iii, p. 327.
HAMMARSTEN: Jahresberichté tiber die Fortschritte der Thierchemie, 1871,
isi pologe

Salivary Digestion in the Stomach. 403

going salivary digestion zx vitro at 38° C., show that in seven minutes
four fifths of the amount of sugar found at the end of an hour is al-
ready present. To be sure, under these conditions the accumulation
of the products of the digestion inhibits the rapidity of the action of
the ferment as time passes,! but it is clear that salivary digestion
is sufficiently rapid to cause a considerable amylolysis before the acid
is secreted even in the pyloric end to a degree preventing further
activity of the ptyalin.

Control experiments were made to discover if in the stomach or in
the treatment of its contents there was any factor which, aside from
the saliva, would result in any considerable change from starch to
sugar. Crackers mixed with distilled water were fed to animals, and
after an hour the stomach contents were treated in the usual routine.
Under these conditions only the slightest traces of a reducing agent

were discovered.

RESULTS.

The difference in salivary digestion in the two ends of the stomach
after different periods of time is shown in the table on page 404.

Several matters are to be noted in reference to the figures in this
table. There is a remarkable diversity in the amount of maltose
present in different cases having the same period of digestion, and
there is no very uniform increase in the figures as the period of
digestion is prolonged. Probably numerous agents are concerned in
producing this diversity, but control experiments carried on zz vitro
indicate that the most important cause is the variation in the activity
of the saliva at different times, — a fact to which attention has been
called by several observers.? It is evident that the figures in these
cases cannot be compared absolutely ; in each instance, however, the
ratio of the change in the two parts of the stomach may be taken and
these ratios will serve for comparisons.

The ratios derived from the figures show that there is in every
case a greater percentage of sugar present in the cardiac end than in
the pyloric end. Two considerations serve to prove that the actual
differences between the Zotal amounts of sugar produced in the two
ends of the stomach are notably greater than the differences between

1 LEA: Journal of physiology, 1890, xi, p. 239.

2 See HoFBAUER: Archiv fiir die gesammte Physiologie, 1897, Ixv, p. 503;
CHITTENDEN and RICHARDS: This journal, 1898, i, p. 461; OEHL: Memorie del
reale istituto lombardo di scienze e lettere, 1902, xix, p. 135.

404 W.B. Cannon and FH, F. Day.

the sugar present in unit volumes of the dried contents. In the first
place, as previously stated, the pyloric contents are invariably more
fluid than the cardiac contents: it follows, therefore, that, since a

In this and in subsequent tables each number represents the fractional gram
of maltose present in one gram, of dried contents of different parts of
the stomach, after different periods of time.

Period. Pyloric end, | Cardiac end.

+ hr, 0.352 0.363
0.171 0.209
0.338 0.401

Average ratio .

July 18 0.293
Aug. 1 0.095
Jan. 30 0.111

Average ratio .

1 hrs. July 11 0.360
CoS 0.244
ienZ6 0.262

Aug. 1 0.135

Jan. 14 0.291

Average ratio .

0.315
0.291

Average ratio

larger amount of fluid must be evaporated from the pyloric food than
from the cardiac food, in order to secure the dried masses, there is
relatively less sugar in a unit volume of the original pyloric food in

Salivary Digestion in the Stomach. 405

comparison with a unit volume of the original cardiac contents than
the ratios given in the tables would indicate. For this reason, there-
fore, the actual disparity of sugar production in the two ends of the
stomach is greater than that seen in the dried contents. The second
factor increasing these differences is the inequality in the cubic con-
tents of the two ends of the stomach. After a full meal the cardiac
end holds by far the greater part of the focd. Several observa-
tions on the relative capacity of pyloric and cardiac portions of the
stomach, as shown by their contents after the ingestion of about
100 c.c. of food, brought forth the average ratio of 1 to 5. In order
to get the ratio of the total amounts of sugar in the two regions it is
necessary to multiply the cardiac figure by 5; from this factor alone,
therefore, the ratios given in the above table change so that they are
as 10 to 57.5 for one half hour, as 10 to 88 for one hour, as 10 to 71
for one and one half hours, and as 10 to 63.5 for two hours.

It is to be observed that after one half hour the amount of sugar in
equal parts of the dried contents of the two regions of the stomach is
almost the same, that the greatest difference between the two regions
appears in the estimates for one hour, and that the ratios for an hour
and a half and for two hours again approach unity. Without much
doubt the reason for the percentage sugar content being almost the
same in the two portions of the stomach at the end of a half hour is
that within that time the sugar formation takes place at almost equal
rates in the two portions, z.¢. for some time after the ingestion of
food the ptyalin is not inhibited even in the pyloric region by the
appearance of free acid. Every examination of cases in which diges-
tion had proceeded for a half hour demonstrated that free acid was
already present in the pyloric contents. Salivary digestion in the
pyloric end of the stomach was therefore at a standstill. That salivary
digestion does not cease, however, in the pyloric end until toward the
end of the half hour is shown by the large amount of sugar produced
compared with that in the cardiac end. On the other hand, the fact
that there is a difference between the sugar percentages in the two
portions shows that the free acid in the pyloric end has had an inhib-
itory effect. This inhibitory effect is more marked if the speed of
action of the ptyalin is retarded (as, for example, by diluting it), for
then the acid has opportunity to check the activity before consider-
able change has occurred. Thus in the stomach of an animal fed 30
gms. crackers mixed with 100 c.c. diluted saliva (saliva 25 c.c., water
75 c.c.), the ratio of sugar percentages at the end of a half hour was

406 W.B. Cannon and H. F. Day.

not near 10: II.5, the average when undiluted saliva was given, but
was as IO to 20, a ratio approximating that found at the end of an
hour when undiluted saliva is used. Under ordinary circumstances
in human beings the saliva is thus diluted by drinking at meals. If
the results secured with the cat may be transferred to the human
subject (a matter to be considered later), the factor of dilution would
certainly enhance the relative value of the fundus as a region for
salivary digestion.

The change in the ratios in the above tables, from 10: 17. 6 at
the end of an hour, to 10: 12.7 at the end of two hours, can be con-
sidered only after evidence of other changes has been presented. The
evidence for these changes will now be given.

It is certain that the sugar formed is in solution, and it is highly
probable that this solution may diffuse from a region in which it
is more concentrated into a region in which it is less concentrated.
It may also pass to the lowest position which a fluid may take in
the stomach. That such changes in the sugar concentrations in the
stomach occur is evidenced by the following observations. In four
animals digestion was allowed to proceed for one and a half hours.
The contents of the cardiac end were then removed in two parts: an
external mass from near the ventral wall of the stomach, which was
lowest during digestion ; and the internal mass, dorsal in its relation
to the first mass. When these masses were dried the sugar percentages
were as follows:

External food Internal food
(from lowest part). | (from higher part).

Ol 0.500
0.410 0.357
0.326 0.263
0.360 0.281

These results indicate that as the sugar solution is formed it passes
into a dependent portion of the stomach. Further data on this point
were secured by having an animal rest on the left side during a
digestive period of an hour. Since the cat’s stomach lies more
nearly transverse than longitudinal in the body, the position on the
left side caused the cardiac end to lie below the pyloric. The sugar

Salivary Digestion in the Stomach. 407

percentages of the food in the most dependent portion and in the
interior are as follows:

External food Internal food
(from lowest part). | (from higher part).

0.398 0.301
0.482 0.468
0.481 0.422

When an animal lies on its 77g#¢ side during an hour of digestion,
opposite results are secured from the cardiac end, thus:

External food Internal food
(from highest part).| (from lower part).

0.291 0.423

As already noted, at the end of an hour or an hour and a half
the food near the wall in the cardiac end of the stomach almost inva-
riably shows the presence of free acid, while the internal food of this
region retains its original reaction. Evidently in the cases cited the
internal region, favorable for amylolysis, contains less sugar than
the dependent region, in which the process is hindered or prevented.
The most reasonable explanation for a larger amount of sugar near
the dependent gastric wall than in the interior of the food mass, is
that the sugar solution runs into the lowest region from the effect of
gravity.

Inasmuch as the sugar solution passes thus from the midst of the
food in the stomach to the gastric wall, it is probable that diffusion
also occurs from the cardiac region into the pyloric region, in which
sugar is present in less amount. The fact of diffusion can readily be
demonstrated by feeding first meat, and later starchy food mixed with
saliva. Under such circumstances the meat is found crowded against
the greater curvature, the starchy food lies along the lesser curvature,
and between the two foods a perfectly clear separation is possible.
If the meat by itself causes no reduction of copper, whatever reduc-
tion it may cause subsequent to its being in the stomach alongside
of food undergoing amylolysis must be due to diffusion of sugar

408 W. B. Cannon and H. F. Day.

into the meat. An animal was fed thus with shredded canned salmon
and later with crackers mixed with saliva; and the meat taken at the
end of an hour from near the surface in the cardiac region, fully
a centimetre and a half from the starchy food, although originally
giving no reaction for sugar, contained about 0.7 per cent sugar cal-
culated as maltose. Since the sugar in the starchy food was only
27 per cent, the relative per cent of sugar in the meat was almost
3 per cent. Diffusion of sugar into the meat had evidently taken
place to a considerable degree.

A further result of the solution of the sugar is the possibility,
in this state, of its being absorbed. Brandl! and v. Mering? have
proved that sugars may be absorbed from the stomach, and have shown
that within limits the rate of absorption increases with the increase
of concentration of the solution. Inasmuch as the sugar is present
in the cases under consideration in somewhat high concentrations, it
is certainly very likely that some sugar is absorbed directly from the
stomach.

Only these facts of diffusion and absorption will explain certain
observations repeatedly made in the course of the experiments,
namely, that control digestions zz vztro almost invariably resulted
in the production of more sugar than could be found in the cardiac
end of the stomach. Typical examples of the difference are as
follows:

Control food Cardiac food.
(22 vitro).
July 18 0.473
26 0.309
ee | 0.436

Aug. 1 0.309

In these instances there is no free acid present in the cardiac end
to prevent the action of the ptyalin; and, moreover, the action of the
ptyalin is not retarded by its products, for there is not so great
an accumulation of sugar in the fundus as in the control dishes. The

1 BRANDL: Zeitschrift fiir Biologie, 1892, xxix, p. 277.
2 vy. MERING: Verhandlungen des xii Congresses fiir innere Medicin zu Wies-
baden, 1893, p. 471.

Salivary Digestion in the Stomach. 409

facts already presented justify the conclusion that there is a continuous
formation of sugar in the fundus, that as the sugar is formed it diffuses
from the region of greater concentration into regions of less concen-
tration, and that it is absorbed as it comes in contact with the gastric
walls. Thus the differences between sugar percentages in the two
ends of the stomach would become less as time elapsed, and thus
there might readily be less sugar present in the cardiac end of the
stomach than in an artificial digestion from which the products do
not pass away.

It is probable that as the sugar diffuses from the region of greater
concentration into regions of less concentration a certain amount of
the ptyalin is carried away with it. That this loss is not serious
is shown by making a watery extract of the cardiac contents and
testing its amylolytic power. Such extracts were made of pyloric
contents, and of food near the surface and in the interior of the cardiac
end after digestion had proceeded for an hour. Tested with starch
paste, blue with iodine, these extracts gave the following results:
pyloric food, no change of color in four hours; surface of cardiac food,
slight change of color; interior of cardiac food, complete disappear-
ance of color in a short time. Manifestly ptyalin was still active in
the cardiac end of the stomach.

Further evidence of absorption was secured by feeding two animals
with equal amounts of the same food mixed with equal amounts of the
same saliva, and examining the stomach contents at different intervals
after the food was given. The experiment was tried twice with the
following results :

Period. Cardiac food. Pyloric food.

+ hours 0.363 0.352
0.455 0.315
0.209 0.171
0.244 0.135

In these two cases the cardiac sugar content increases as time goes
on; naturally it would be expected that as the cardiac food passes
into the pyloric end the pyloric sugar content also would increase;
instead, however, as time goes on there is a decrease in the percentage
in the pyloric end. These two cases are not sufficient to allow sure

410 W. B. Cannon and H, F. Day.

conclusions to be drawn, but they indicate that sugar passes out of
the pyloric end of the stomach more rapidly than the undissolved
portions of the food. A selective action at the pylorus is hardly
to be expected, for the food that comes to it is not a mixture of hard
particles in fluid, but a uniform creamy substance. Active absorption
at the pyloric end, for which the peristalsis is especially favorable,
is suggested as an explanation of the results which these two cases
furnish.

Effect of position on sugar formation.— It is. important to know
if position has an effect on the mixing of the gastric contents with
the gastric secretions, and if thereby salivary digestion is modified.
Observations were made on animals caused to rest on the right side
or on the left side during the period of digestion. As the full
stomach in the cat lies nearly transverse, the long axis of the organ ©
was nearly vertical in either case. The results secured were as
follows :

ANIMALS ON LEFT SIDE.

Pyloric end. | Cardiac end. Ratio.

0.191 | 0.350 10 : 18.3
0.435 0.475 : 10.9
0.227 0.451 LOS

Averame ration.) sees sn

ANIMALS ON RIGHT SIDE.

Pecwljnsak 0.306
JENIN IEP 6 os 0.251

Average ratio

Examination of these results reveals no very marked differences
in the ratios of sugar content when the long axis of the stomach is
reversed in direction. If the animal is on the left side, with the
cardiac end lower than the pyloric end, there is, however, more sugar
present in the food in the fundus than is the case when the opposite
relation of the parts is maintained.

Salivary Digestion tn the Stomach. All

Effect of variations in the food, and massage. — When the food
is liquid, it is to be expected that the gastric juice will be more
readily and more uniformly mixed with the food than when the
food is present in somewhat viscid masses. Crackers (30 gms.)
were ground to a coarse powder and mixed with 150 c.c. filtered
saliva. The mixture was about as fluid as the chyme ordinarily seen
in the pyloric end of the stomach. After one hour the percentage
sugar content in dried food from the two ends of the stomach was as
follows :

Pyloric end. Cardiac end.

0.427 0.519

The average ratio for one hour with the usual test food was
10: 17.6; with the liquid diet there is a considerable decrease in
the disparity.

Similarly small amounts of food in the stomach would naturally
be more quickly and uniformly permeated by the gastric secre-
tions, and the differences to be observed when large amounts are
given would not under these circumstances be expected. Animals
were fed one half the usual test meal,—15 gms. crackers, 50 C.c.
saliva, — and digestion allowed to proceed for one hour. The fol-

lowing results show the sugar production in the two parts of the
stomach :

Date. Pyloric end. Cardiac end. Ratio.

ANUS Si 2a) oc 0.369 0.349 10 : 9.4

Jens BUN Bia =e 0.296 0.284 10 : 9.6

Averase ratio! panies een LOKIOIS

In these cases more sugar is present in a unit weight of dried
pyloric content than in a unit weight of dried cardiac content, a fact
difficult of explanation.

A difference in the ratios similar to that observed when the amount
of food was small was seen in two instances in which the upper
abdomen was massaged at intervals of about five minutes during one

412 W. B. Cannon and FH. F. Day.

hour of digestion. The sugar content in these instances was as
follows:

Pyloric end. Cardiac end.

0.484 0.391
0.493 0.423

Average ratio

In these instances, as when liquid food or a small amount of food
was given, the sugar content of the two ends of the stomach was
more nearly the same than in the usual cases with large amounts
of semi-solid food resting in the stomach undisturbed. The larger
percentage of sugar in the pyloric end, when little food was given
and when the stomach after a full meal was massaged, was not
expected and cannot at present be explained.

Effect of combining proteid with carbohydrate food. — From the
work of Pawlow and his school! it is to be expected that after the
introduction of flesh food, the character of the secretion of the gastric
juice will change; there will be in the early stages of digestion a
more abundant flow of gastric juice, with a concomitant greater pro-
duction of hydrochloric acid, than is normal when carbohydrates alone
are ingested. On the other hand, as pointed out by Chittenden and
Smith,” proteid not only favors the diastatic action of ptyalin by com-
bining with hydrochloric acid so as to delay the appearance of the free
acid, but also seems to act, in the form of a small percentage of acid
proteid, as a direct stimulant to diastatic action. The effects of pro-
teid favorable to diastatic activity might therefore counterbalance the
extra production of acid destructive to that activity. Such in fact
was the result of experiment. A cat was fed 45 gms. of thoroughly
mixed fish and crackers (25 gms. salmon, 20 gms. crackers) and
100 c.c. diluted saliva (50 c.c. saliva, 50 c.c. water). After one hour
the stomach was examined in the usual manner. No free acid was
found in the cardiac end, and in the pyloric end there was only a slight

' PawLow: The work of the digestive glands, London, 1902, pp. 31, 34;
and 365.

? CHITTENDEN and SMITH: Studies from the laboratory of physiological chem-
istry, Sheffield Scientific School of Yale College, 1885, i, p. 33.

Salivary Digestion in the Stomach. A13

discoloration of the test paper. The food, treated in the manner
already described, yielded the following amounts of sugar:

Internal cardiac, 0.153 External cardiac, 0.152 Pyloric, 0.147

The ratio of sugar production in the two parts of the stomach when
flesh is added to the carbohydrate food —about Io to 10.3 — need
only to be compared with the ratio obtained when carbohydrate food
alone is given—- 10 to 17.6-—to see that the proteid protects the
ptyalin from the acid in the pyloric end and permits the diastatic
action to continue for an hour at least at a rate equal to that in the
cardiac end. It should be remarked that the protection afforded by the
proteid is really effective only in the relatively small pyloric portion
of the stomach and to some‘extent on the surface of the cardiac portion.
The greater mass of the food, lying in the fundus, undergoes unin-
terrupted amylolysis, not because the proteid protects the ptyalin, but
‘because the food in this region is not mixed with the gastric juice.

Change of starch into dextrin. — Starch not changed to maltose in
the stomach may be changed in considerable degree to dextrin.
Inasmuch as dextrin is not directly fermentable, the amount of
dextrin thus produced represents just so much carbohydrate food
preserved from possible loss to the organism by fermentation. The
amount of dextrin formed was roughly calculated in several cases as
follows. One gram of the dried stomach contents was taken, about
100 c.c. water were added, and after one half hour the mixture was
filtered through a weighed filter, on which the residue (starch and
proteid) was several times washed, then dried, and residue and filter
together weighed again. The loss of weight was taken to represent
the soluble carbohydrates. From the total weight of soluble carbo-
hydrates was deducted the weight of the sugar, and the remainder
indicated the dextrin formed. The salts present were not deter-
mined. The dextrin thus roughly estimated varied in several
instances between seventeen and forty per cent. The following ex-
ample illustrates the relations between the various products in the
different parts of the stomach:

Aug. 1 Amount taken.| Starch. Dissolved. Maltose. Dextrin.

Internal cardiac 1.0 0.354 0.646 0.250 0.396
External cardiac 1.0 0.441 0.559 0.237 0.322

Pyloric 1.0 0.687 0.313 0.135 0.178

414 W. B. Cannon and FH. F. Day.

In all cases there was a large amount of dextrin in the internal part
of the cardiac contents, to which the hydrochloric acid penetrates
last. Evidently fermentation may proceed for a long time in this
region unchecked by acidity. It is of obvious advantage, therefore,
to have in this region a considerable amount of the starch ingested
preserved to the organism by being changed to a form fermentable
with difficulty or not at all.

DISCUSSION OF RESULTS.

In reviewing the observations here recorded on salivary digestion
in the cardiac and pyloric ends of the stomach it is to be observed
that invariably under normal conditions, when ordinary amounts of
food are given, more sugar is produced in the cardiac than in the
pyloric end. The conditions present in the cardiac end of the
stomach are wholly in agreement with the contentions of observers
who have urged that saliva may continue its chemical function for
some time after being swallowed; the conditions present in the
pyloric end, however, are very different, for the food there soon
becomes acid and the action of the saliva stops. In 1880 von den
Velden divided gastric digestion into two periods; in the first period,
before free hydrochloric acid appears, the saliva swallowed with the
food is still effective; in the second period, after free hydrochloric
acid appears, only the pepsin continues activity. It is clear, however,
from the results presented, that these two periods are different in the
two ends of the stomach; in the pyloric end the first period is short,
in the cardiac end it may last three or four times longer than it lasts
in the pyloric end. In the early stages of digestion in the stomach,
therefore, the cardiac end serves chiefly for salivary digestion; the
pyloric end, after a brief course of salivary digestion, is thenceforth
the seat of the strictly peptic changes. Later, as the cardiac con-
tents become penetrated by gastric juice, diastatic activity ceases,
and the stomach contents as a whole are subjected to the action of
proteolytic ferments.

It is of interest to note that the results reported in this paper are in
close agreement with the results obtained by Ellenberger and Hof-
meister ‘in their observations on the horse and pig, and by Hohmeier
in his observations on the rat. The cardiac end of the stomach in

* ELLENBERGER and HoFMEISTER: Archiv fiir wissenschaftliche und prak-
tische Thierheilkunde, 1884, vii, p. 6, and 1886, xii, p. 126.
* HOHMEIER : Inaugural-Dissertation, Tiibingen, gol.

Salivary Digestion in the Stomach. 415

the horse, the.pig, and the rat is to a great extent lined with pave-
ment epithelium and with “cardia glands,” distinguished from the
fundus glands in not having an acid secretion.1 According to these
observers this region becomes, in the absence of gastric secretion, the
seat of prolonged amylolysis. That there is this similarity in salivary
digestion between animals without acid secretion in the cardiac end
and animals with free secretion there, indicates that the division of
the stomach into two regions with functions chemically different
during the early stages of gastric digestion is a general fact. The
important agent in either case is the mechanical agent — the absence
of peristalsis in the cardiac end and the consequent quiescence of
the food in this region. As already pointed out (p. 401), the distinc-
tion between the active pyloric and the inactive cardiac end of the
stomach has already been observed in many animals including man.
It is altogether probable, therefore, that in man as in other animals
the cardiac end serves chiefly for the action of ptyalin during the
early stages of gastric digestion.

The recent observations of Miiller and of Hensay, together with
the results here presented, emphasize strongly the importance of
mastication. Only by mastication is the food properly mixed with
saliva and properly broken up so that all parts of it can be pene-
trated readily by the saliva. When the food has thus been thoroughly
jnsalivated it will undergo to a great degree salivary digestion in the
cardiac end of the stomach.

SUMMARY.

The evidence that the action of ptyalin is inhibited in the stomach
soon after the ingestion of food is inconclusive. ‘The support for this
evidence from the commonly accepted accounts of mixing currents in
the stomach is not well founded. Observations show that in many
animals, including man, gastric peristalsis occurs only in the pyloric
end of the stomach; the cardiac end remains undisturbed by the
waves. Food in the pyloric end is soon mixed with the gastric se-
cretions, but food in the cardiac end of the stomach is not mixed with
the acid gastric juices for two hours or more, and in this region,
therefore, during that time salivary digestion may go on undisturbed.

Examination of the dried contents of the pyloric and cardiac
portions of the stomachs of cats, after carbohydrate food mixed with

1 OppeEL: Lehrbuch der vergleichenden mikroskopischen Anatomie der Wir-
belthiere, erster Theil, Der Magen, Jena, 1896, pp. 240, 337, 346, 397-

416 W. B. Cannon and H. F. Day.

active saliva has been given, shows that the percentage of sugar
present is about the same in the two portions at the end of a half
hour, and at the end of an hour the cardiac portion contains about
eighty per cent more sugar in unit volumes than the pyloric portion.
The actual amount of sugar present in the fundus is relatively much
greater than this ratio would indicate, for the fundus contains after an
ordinary meal about five times as much food as the pyloric portion.

After an hour the ratio of the sugar percentages in the two parts -
of the stomach begins to approximate unity again. This change is
probably due largely to diffusion of sugar from the fundus into the
pyloric end, and to some extent to absorption.

The diffusion of the sugar does not to a marked degree remove the
ptyalin from the food.

Position does not very notably affect the differences in sugar pro-
duction between the two parts of the stomach, although with the
fundus lower than the pyloric portion slightly more sugar is found in
the fundus than when the opposite relation is maintained.

When liquid food is given, when small amounts of food are given,
and when the stomach is massaged, sugar percentages in the two
parts of the stomach are nearly the same.

Mixing proteid with carbohydrate food protects the ptyalin from the
action of free hydrochloric acid in the relatively small pyloric part of
the stomach and on the surface of the cardiac contents. The greater
mass of the food, lying in the fundus, undergoes uninterrupted
amylolysis, not because the proteid protects the ptyalin, but because
the food in this region is not mixed with the gastric juice.

Much of the starch not changed to sugar is changed to dextrin,
and thus, since dextrin is not readily fermented, the food is saved
to the organism. The especial value of this process lies in the fact
that it occurs to the greatest degree in the fundus, in which region
the hydrochloric acid, inhibiting the action of many of the organized
ferments, does not for some time make its appearance.

In the early stages of gastric digestion, if food has been properly
masticated, the fundus serves chiefly for the action of the ptyalin;
the pyloric portion, after a brief stage of salivary digestion, is there-
after the seat of strictly peptic changes. Later, after two hours or
more, as the contents of the fundus become acid, the food in the
stomach as a whole is subjected to the action of proteolytic
fermentation.

NUCLEIN METABOLISM IN LYMPHATIC LEUKEMIA.

By YANDELL HENDERSON anp GASTON H. EDWARDS.
[from the Physiological Laboratory of the Vale Medical School.]

LTHOUGH the metabolism of the nucleins has been the subject

of a considerable number of investigations during the last few
years, the net result has left much of uncertainty regarding the
processes involved in the various forms of leucocytosis, physiological,
experimental, and pathological. The first broad comparative study
of the variations of metabolism involved in these conditions was that
of Milroy and Malcolm.! On the basis of their own researches and
those of previous investigators they drew a sharp distinction between
forms of leucocytosis which previously (and on the basis merely of
the similarities in the blood counts) had been regarded as essentially
similar. They showed that while in the leucocytosis produced by the
injection or ingestion of nuclein the increased number of the leucocytes
in the blood is associated with an increased excretion of uric acid
and phosphates (the latter in greater amount than could originate in
the nuclein absorbed), in chronic leukaemia on the other hand the
excretion of these substances is rather below the average for normal
individuals. To these conclusions v. Moraczewski2 has added the
observation that in the cases of leukaemia studied by him there was
a marked retention of nitrogen, phosphorus, chlorine, and calcium,
and an alteration in the relative amounts of sodium and potassium
excreted. In acute leukemia, on the contrary, Magnus Levy ® has
found that the excretion of nitrogen, uric acid, and phosphates is
enormously increased, coincident, however, with only a slight leucocy-
tosis. Accordingly White and Hopkins,‘ in discussing the results
of these investigators and adding observations made by themselves,

1 MiLroy and Matco.tm: Journal of physiology, 1898, xxiii, p. 232; and 1900,
XXV, Pp. 105.
2 vy. MORACZEWSKI: Virchow’s Archiv fiir pathologische Anatomie, 1898, cli,
Pp. 22.
ELEY Y = Lizd,, 1899; eli, p. 107.
4 Waite and Hopkins: Journal of physiology, 1899, xxiv, p. 42.
417

418 Yandell Henderson and Gaston H. Edwards.

are led to the conclusion that there is no “ necessary proportionality
between the number of circulating leucocytes and the excretion of
those products (P,O; and alloxuric bodies) which result from the
breakdown of nucleins.”

Recently there has been under treatment in the public clinic of the
Yale Medical School a case of lymphatic leukaemia which, through
the kindness of Dr. O. T. Osborne,! we have been enabled to observe.
As the clinical aspects of this case have already been discussed else-
where no further description of them is needed here than to say that
the patient was a male, sixty-four years of age, and presented a case
of leukaemia typical of the purely lymphatic variety of slow develop-
ment and progress. It was therefore especially suited to the study of
metabolism in this form of leucocytosis.

Although no attempt was made to regulate the diet a careful record
was kept, and it was found to vary only within very narrow limits,
and to be almost nuclein free. It consisted of oatmeal, bread, pota-
toes, milk, butter, cheese, eggs, and occasionally beef, tea, coffee, and
a small amount of whiskey. The urine was collected for twenty-four
hours in bottles containing a few cubic centimetres of toluol, and was
brought to the laboratory next day for analysis. Total nitrogen
was estimated by the Kjeldahl-Gunning method; uric acid by
Hopkins’s method, weighing the crystals; phosphoric acid by titra-
tion with uranium acetate, using potassium ferrocyanide as indicator ;
total acidity and chlorides by the ordinary methods. In order to
discover whether there might not be an elimination of phosphorus
in organic combination a number of determinations were made of
the total amount after evaporation of a sample of urine and ignition
with sodium hydroxide and potassium nitrate. The results, however,
agreed entirely with those obtained by the titration method. Allan-
toin was looked for by evaporating 400 c.c. of urine toa syrup, adding
a little strong acetic acid, and after several weeks examining for the
crystals. The results were wholly negative. The unreliability of
some, and the difficulties of all of the methods at present available
for the estimation of the xanthin bases, and the fact that there is no
reason to suppose that the relative amounts of uric acid and the other
alloxuric bodies vary in any considerable degree dissuaded us from
the attempt to determine this excretion. The urine was at all times

wholly free from albumin or sugar and in all other respects entirely
normal.

1 OsBoRNE, O. T.: American medicine, 1902, iv, p. 533-

Nucleen Metabolism in Lymphatic Leukemia. 419

The results of observations extending over more than six months
are given in the accompanying table, together with the calculated
ratios of the principal excretives, and such other data as seem of
importance. The significance of the figures is perhaps best seen
in the charts showing their variations. The data presented show
that in spite of the enormous leucocytosis (175000—380000 per cubic
millimeter, of which 96 per cent were lymphocytes), and in spite
of the alternations of great increase and equally marked diminution
in the number of circulating corpuscles, the excretion of uric acid
and phosphates was at no time excessive. The leucocytosis seems
to be due not to a general increase in nuclein metabolism, but to
a failure in the normal destructive processes. A diminution in the
number of circulating leucocytes was accompanied by a considerable
increase in uric acid excretion, although the actual amount excreted
"was small compared with the reduction in the number of circulating
leucocytes. Their diminution was therefore probably due to an abstrac-
tion from the blood and storage, rather than to a destruction of cor-
puscles. This view is supported also by the changes observed in the
size of the lymph glands.

More detailed consideration reveals two periods in which the course
of metabolism was markedly different. During these the average
daily amounts and ratios were as follows:

October and November. January to April.

N 12 grs. N 9 grs.
Uniemcid Ni <li 23) = Orictacid=:INe: slew
E507 SNiss eS POP Ni ce BL.

Waac-acid| ESO: le: 3 Wixiesacid) -"P5Or = lessleS

In the first period the total nitrogen was, if anything, slightly sub-
normal, the uric acid slightly above the normal, and the phosphates
distinctly subnormal, —the normal ratio of P,O,; to nitrogen being
about 1 to 5. Contrasted with these conditions the second period
shows a distinct diminution in nitrogen excreted, a more than pro-
portionate increase in the ratio of the uric acid to nitrogen (due to
a slight increase in the absolute amount of uric acid), and a further
great reduction in the ratio of phosphates to nitrogen in spite of the
diminution in the latter. “ Taking the normal ratio of uric acid to

Hi. Edwards.

Yandell Henderson and Gaston

420

000‘0+8' | 000‘OSE

000‘962°€ | 000‘E0%

000‘TE6‘E

000‘61L°2 | OO0'SLT

Th 1
Sci T
go? 1
Geel:
8o?T1
Let
6d: 1
orm |
Sie cal
(Gaol
site |b
AG II
ONG
Sie cal

"SOuY AA

"0%d

prov 119)

S6
sy)

SOE =
$C :
T&Z:
LoS *
OST:
0S :

0080

cer 87.

OOor
ell
rol
9 +1
SPU

OL

"Sued

JNUTIEIOL

_N [P10
prov ou)

‘SLNQOD GooTg

‘SOLLVY

OE NGSE

AN TSO.L | sap0(yD

Sle
80°C
LOT
£9'T
Os'T
eLT
09'T
08'T
9L'T
Wl
Os T
eS'T
trl
LST

"survIs

OH se
Aypioy

“HANIY(]) HO SHSATVNY

9201 008

*MOJIEUL
auog

‘otuasly 48

9
c

VINWMONUT OILVHdNAT NO SNOILVANHSAO AO ATAVL

A421

1a.

Lymphatic Leukem

27

Nuclein Metabolism

0} payqopul av 9A,

‘pasn ,, plow o1ajonu,, yy OF ‘OD ® stavq Ye ‘sissay[
‘SUOISN[DUOD IO LIVp dy} UL PauIIDUOS jou SI ynq ‘ssauajza{dwo0d Jo dyvS BY} IOJ pa}BoIpUT St JuWyVOI] BY, 7

000°9T9‘¢
000‘ZI1'E

000‘0Tt*+

000'090‘F

000'299'¢

00089
000°8EZ

000‘26Z

Chleeal
Spueal

601
6

aL

MS |Le |
Sr: 1

060
Orl

ScL'0
éSL'0

se9'0
S+9'0
LSS'0

+86

S3°8
+18
ce Tl
+96
Is‘01
$88

SL
cs

697
67

FIOL
6101

OOET
OOOT

‘ple
o1a/9N N
L
|
fc

"MOTEL
quod

Z your
82

A422 Vandell Henderson and Gaston H. Edwards.

P,O, to be 1 to 4 or 5, the tendency to a progressive diminution
in the excretion of phosphates is strikingly illustrated by the figures
presented above. Although the diet was not regulated it remained
the same in kind, and there was no reason to believe that the amount
was diminished during the second as compared with the first period,
—certainly not to the extent suggested by the nitrogen eliminated.
The explanation of the small nitrogen of the second period is probably
to be found in a nitrogen retention, a general diminution of proteid
metabolism. Furthermore the amounts and the ratios of the phos-
phates show that of this excretion also there was aretention. Indeed,
in this case the retention was much more distinctly indicated and
was much greater relatively than with the nitrogen. That this reten-
tion was not due to a diminished breaking down of nuclear material
during the second as compared with the first period, is shown by the
fact that there was a coincident slight increase in uric acid output.
Nor can the low phosphorus output through the urine be explained
by an increased excretion through the intestine. The ratio of P,O, to
nitrogen in the faeces was indeed so high as to suggest such an avenue
of excretion, but the absolute amounts of P,O,; in the faeces were
insufficient to account for more than a part of the urinary deficiency.

ANALYSES OF FACES OF SEVEN-DAY PERIODS.

Date. Dry solids. Total N. | Total P,O;. |N. per diem sepia

Jan. 24-31
Feb. 21-28

grams.

80.0
76.0

grams.

4.27
3:53

grams.

6.76
Dalv/

grams.

0.61
0.51

grams,
0.97
0.74

In the chlorides also there was strong evidence of a retention during
the second period; and the small volume of the urine and the occur-
rence of cedema in the ankles during the latter half of February
suggest a considerable retention of water. The contrast afforded by
the conditions existing during October and November with those dur-
ing January and thereafter is the more interesting because the former
are essentially similar to the conditions observed by Milroy and
Malcolm, and the latter are equally the counterpart of those found by
v. Moraczewski. The differences in the results of these observers
are probably to be assigned, therefore, to the different stages to which

the leukzemic conditions had developed when they came under
study.

CHARTS SHOWING THE RESULTS OF URINARY ANALYSES AND
BLOOD COUNTS.

October November
i nlieeo ot 26S 14 17'19 21 2426 4 '8 12 26 28 3/ 22 24 26 28 2

'
i)
|
1
!

AN

Total Nitrogen

\ '
Base Line 00 :
Total Acidity HCL

] H
15 grams “A

'
‘
!
'
\
|

Base Line 00

1
Base Line [000
Leucocytes

250,000

| Uric Acid
=

O5 gram

1

1
1
'
{

00 Base Line

00° Base Line

The points indicated in these charts are: —

1. The curves for total nitrogen and acidity are very similar. If they were com-
pletely superimposable an excretion in the uniform ratio of 10: 1.5 would be demon-
strated.

2. The curves for uric acid and P,O; are very similar during the “first period”
(see text), and together with the sp. gr. are approximately parallel to total nitrogen.
For the “second period”’ there is no such parallelism.

3. The curves for chlorides and volume are strikingly similar, but show no relation to
those before mentioned.

4. Except during the later days of November the number of leucocytes bears no
apparent relation to any of these curves.

424 Yandell Henderson and Gaston H, Edwards.

Referring to the charts, it will be seen that throughout both periods
the total acidity was closely parallel to the nitrogen, which accords
with the view that the total acidity is chiefly an expression of the
SO, resulting from the proteid metabolism. Accepting this idea,
the parallelism in the two curves is an indication of the lack of any
marked disturbance in proteid metabolism. During the first period the
specific gravity ran quite closely parallel to the total nitrogen (urea),
without regard to the volume, while during the second period the
specific gravity shows no relation to either volume or nitrogen. The
chlorides varied more than any other excretive (from 2.8 to 21.0grams
per day), yet throughout both periods was closely paralleled by the
figures for the volume of the urine. On the other hand, these two
curves exhibit no relation to those of nitrogen and acidity. During
the first period, as already stated, the uric acid and phosphates were
closely parallel, and their variations when studied in connection with
the curves for the number of circulating leucocytes and the excre-
tion of chlorides (especially during the time between the eleventh
of November and the fourth of December) are very suggestive.
They seem to show that during this time there occurred (1) a marked
increase in nuclear break-down, followed by a return to the usual
rate; (2) an increase in the number of circulating leucocytes (probably
thrown out of the glands where they had been stored) followed by a
diminution in their number; and (3) coincident with these changes, or
slightly preceding them, a considerable increase in the excretion
of chlorides followed an immense diminution.

ie EFEECT OF DIURETICS, NEPHRITIC POTSONS,
mn, OTHER AGENCIES ON FRE CHLORIDES
OF> Erie OistN Er:

By TORALD SOLLMANN.

[From the Pharmacological Laboratory of Western Reserve University, Cleveland, Ohio. |

CONTENTS.
Page
PEPUTUNOCUCE OE. Gels fo ware so ME ay seat oe ee soy Sy ay eee Seep 426
II. Methods:
ilo LMECH CHGS OMNIS) oc 6 Ss 6 SoBe 6 cha ooo oe es oun & 4S
Pap Micthodsiot analysich ‘citar! en ne Rats) Sonn in) Ws teak, eye Whee aoe
ILI. Results of injection experiments . .. . a. ; ao AHS
1. Effect of sodium sulphate injection on chitarides: A, A niarenen = Zo
2. Saline diuretics which lower the per cent of chlorine in the urine. . ~-429
3. Does the injection of the salts stimulate the retention of chlorides? . 434
4. Saline diuretics which do not cause chlorine retention. . . . . . 435
Sa biheCHo mother diGneticragentse.) Ws) Gi 4 (tenet eae te OS
Om Etech ofnephniticmapentss. cs. 05 9-0 co" iy so) ee ed a
7. Otherfactors . . - 443
IV. Discussion of the effects of she various rier on ae aieeaee of ne urine,
and the light which pi throw on the mechanism of chlorine
MeELentiones| 9 ps ed C8 Cae Aa
1. What is the essential factor in nthe chloginie enone syadke 447
2. The per cent of chlorine in the urine is not determined 2 the total
per cent of chlorine inthe serum . . . : : : 447

3. Is the urine poor in chlorine secreted poor in ithied ion, Or is it seoreted
with the normal content of chlorine, but modified after its secretion,
by absorption of chlorine or by the secretion of water? . . . 449
4. What is the mechanism by which urine with a low per cent of citrine
is formed from sera containing a low per cent of free sodium
chloride, but a normal or only slightly subnormal per cent of total

sodium chloride? .. . : : 449
5. The saline diuretics increase ihe absolute amount of ehlovine: w hile
they diminish its percent . . . 54 See Ses oro 50)
6. The effect of sodium nitrate, iodide, ‘ad EnipHionyanade LN ee ees S|
7. Phlorhizin and nephritic poisons. . Re 451
8. The difference in the behavior of alba ad jee? Ridaeys as re-
AAS, UNS Geen Of Caine 6 6 o-oo 6 oo 6 o b oe a BOI
Name @onclisions. castle gms tt ay Mead Minar ee ccs ce Sethe ase ne 452

425

426 Torald Sollmann.

I. INTRODUCTORY.

| Ea a previous paper! I pointed out the importance of the phe-

nomenon of chloride retention to the theory of the mechanism
of urine secretion, and I discussed there the data which we possessed
at that time. It was seen that these were sufficient to define the
problem more or less sharply, but that they did not justify any defi-
nite conclusions. In the present investigations I have attempted to
approach the problem, in the first place, by studying the effect of
various classes of diuretic and nephritic agents upon chloride excre-
tion. After the research had been planned in detail there appeared |
a series of papers from Filehne’s laboratory? and a research by
Loewi? which dealt with some of these problems. These were,
however, to some extent mutually contradictory, and covered only a
part of the questions which I intended to investigate. I therefore
carried out the research as planned, with the result that the cause of
the contradictory results was cleared up, and that the facts are now
sufficiently numerous for safe generalizations.

II, MeETHODs.

1. Injection experiments. — The experiments were made exclusively
on dogs. As it was my purpose to investigate as many conditions as
possible, no precautions were taken to keep the animals on a uniform
diet. Some urines were accordingly rich in chloride, others were poor
in this ion. This did not influence the results. The animals received
a large dose of morphine, and were kept under a light ether anaesthesia
throughout the experiments. In Experiments VIII to XXIX, both
vagi were divided to abolish any inhibitory effect of the vagi on urine
secretion, such as is claimed by Corin, and confirmed by the experi-
ments of Anten.! The ureters were exposed by a small incision
through the linea alba, and narrow glass cannulze were introduced
near the bladder. The urine was collected continuously, the col-
lecting vessel being changed every ten to twenty minutes. The in-
jections were made into the femoral vein. The fluids were heated to

1 SOLLMANN, T.: This journal, 1902, viii, p. 155.

2 FILEHNE, W.: BIBERFELD, RUSCHHAUPT, PoToTzky, and ERCKLENTZ:
Archiv fiir die gesammte Physiologie, 1go2, xci, p. 565.

® LoEwt!, O.: Archiv fiir experimentelle Pathologie, 1902, xlvili, p. 410.

* ANTEN, H.: Archives internationales de Pharmacodynamie, 1901, vill, p. 455-

Effect of Diuretics, etc., on the Chlorides of the Urine. 427

+ 38° C., and injected in two to four minutes, usually an hour apart.
The first injection was of 35 c.c. per kg., the succeeding injections
each usually of 25 c.c. per kg. Three or four injections were made.
The urine was collected for about an hour before the first injection.
It was found that the composition of the bladder urine bears no rela-
tion to that secreted after the operation.

The saline diuretics were used in the uniform strength of 4%. The
following solutions were employed : !|—

Sodium salts. — Acetate, 1.94 per cent crystals; ferrocyanide, 7.43 per cent
crystals ; iodide, 2.16 per cent ; phosphate (Na,HPQ,), 5.10 per cent crys-
tals; sulphate, 4.6 per cent crystals ; sulphocyanide, 1.16 per cent.

Non-electrolytes. — Glucose (C. P. Dextrose), 2.57 per cent; urea, 0.886 per
cent, in water or in % sulphate.

Non-saline diuretics. — Alcohol, 3 per cent in water or in 4 sulphate ; juniper
oil, o.4 per cent, and alcohol 1.6 per cent, in sulphate ; caffein, 0.04 per
cent of citrated caffein in ” sulphate; phlorhizin, 0.4 per cent in water

7

or in ” sulphate; methylene blue, 0.5 per cent in 4% sulphate. ‘The
citrated caffein was also used hypodermically in 1 per cent solution.
Nephritic agents. — These will be discussed in the text.

2. Methods of analysis. — The chlorides were determined by evapo-
rating the urine (usually 5 to 20 c.c.) with about half a gram
Na,COg, carbonizing, and fusing with sufficient NaNO,; dissolving,
neutralizing with HNO, and titrating with AgNO, (1 c.c. = 1 mg.
NaCl), using chromate as indicator.

In the presence of iodides, the method of Salkowski? was used.

Neubauer and Vogel® quote this paper of Salkowski as reference for the state-
ment that the method can also be used for bromides. The quotation is
incorrect and the method not applicable: 10 c.c. of NaBr solution
require before treatment, ro c.c. of 7 AgNO,; after treatment (when
they should require no AgNO;) they use 7.3 ¢.c. in one test, 8.0 c.c. in
another.

In a few experiments, in which the freezing point and nitrogen
were determined, this was done by the method of Beckmann and of
Kjeldahl. |

1 For a to per cent solution, ro grams of the salt are dissolved in 90 grams of
distilled water.

2 SALKOWSKI, E.: Archiv fiir die gesammte Physiologie, 1872, vi, p. 214.

3 NEUBAUER and VoGEL (H. Huppert): Analyse des Harnes, oth edition,
Wiesbaden, 1898, p. 713.

428 Torald Sollmann.

III. ResuLts oF INJECTION EXPERIMENTS.

1. Effect of sodium sulphate injection on chlorides, A, and nitrogen. —
Three preliminary experiments were made to ascertain whether it

EXPERIMENT I.

Sa Mg.in10m.| 1
paresis.) AL )|) Ne. | Cl \Ateele |= =
N Cl
Ns per cent. | per cent.
Bladder urine ... . ee 0.790 | 0798 | 0.243 | .... scciena.. jl erceretetal OTR
Before injection . . . 0.2
After injection :
OtonliimS. 14.7 O775 | 0127 | 0:035 | 11-39") TS8i7SeoelSe eee
Wf ion Bie Tato a Gg 8 11.4 0.790 | 0.094 | 0.020 9.01 | 10.69 | 2.28 | 4.7
58m.tolhr.48m. . 10.0 0.765 | 0.120 | 0.060 7.65 | 1204 | 6.00 | 2.0
1h. 48m. to 2 h. 34 m. 10.9 0.640 | 0.123 | 0.055 TASs| USASH GOO 2rZ
Zhen S4ometo.end. 6 nee eee OnLS

EXPERIMENT II.

BetOyennyECtiOnl seem 2.0 BOO URVY er) souma mi acos. || NOS

After injection :
SNS. ae oe 11.4 0.895 | 0112 | 0.040 | 10.12 | 12.77 | 4.56 | 2.8
Binidaee helest oe ee k pies sada || KO)Otess}

EXPERIMENT III.

Bladdermnines 0s ne shee ae | ooo || ade Ae |) Ale

Before injection . . . 0.6

After injection :
2EREORD Sa No 8.9 0.740 | 0.140 | 0.040 | 6.59 | 12.46 | 3.56 | 3.5
So, oO wise G cs 6 Piped 0.765 | 0.112 | 0.020 | 17.36 | 25.42 | 4.54 | 5.6
60m.tolhr. 35m. . 22 0.740 | 0.118 | 0.020 | 1680 | 2679 | 4.54 | 5.9

1 h.35 m. to 2h. 29 m. 250 0.660 | 0.122 | 0.020 | 1650 | 30.60 | 5.00 | 6.1
2h.29m.to 3h.14 m. 22.7 0.740 | 0.158 | 0.020 | 16.50 | 35.85 | 4.54 | 7.9

3 h. 14 m. to 4h, 04 m. 14.7 0.790 | 0.193 | 0035 | 11.61 | 28.40 | 5.15 | 5.8
. 04 m.to 4h.45 m.

0.835

Lofect of Diuretics, etc., on the Chlorides of the Urine. 429

would be necessary to determine the changes in nitrogen and the
freezing-point, as well as the chlorides.

The three experiments gave very uniform results, the urinary con-
stituents during active diuresis varying as follows : —

Depression of freezing-point: 0.640 to 0.790° C. (This factor varied be-
tween o.85 and 1.30 in Experiments VI and VII, in which 35 c.c. per
kg. Na,SO,, 25 c.c. urea, and 25 c.c. glucose were injected an hour
apart.) Per cent N, 0.0938 to 0.1232; per cent NaCl, 0.020 to 0.40;

ay ZO tO” 7.9%

The increased diuresis, as usual, tends to lessen the concentration
of the urinary constituents, whilst it increases their absolute quantity.
The per cent of chlorine and the freezing-point are but little affected
by the degree of diuresis. The diuresis reaches its maximum in
twenty to sixty minutes, and is still quite perceptible in five hours.

The nitrogen and chlorine change generally in the same direction,

: N :
but by no means in the same proportion, the factor Nac] V@tying

between 2.0 and 7.9; the difference is seen very strikingly in the last

two urines of Experiment III. The factor is independent of the

N
NaCl
diuresis, and does not vary inversely to the diuresis, as would be
demanded by the reabsorption theory.

Since the nitrogen and A vary quite independently of the chlorine,
it is evident that they are controlled in part by other factors, which
would complicate the problem. I deemed it better, therefore, to
neglect them in the further experiments.

It was noticed incidentally that the depth of the color of the urines
varied in the same direction as the per cent of nitrogen, so that the
two would seem to be excreted by the same mechanism.

2. Saline diuretics which lower the per cent of chlorine in the urine.
—This class comprises solutions of the acetate, ferrocyanide, phos-
phate, and sulphate of sodium; as also urea, glucose, and water.
With the exception of the very small diuresis produced by the last
substance, the phenomena are exactly alike in all cases.

The diuresis sets in very shortly after the injection. The per cent
of chlorine falls at the same time to a very low figure (to less than
0.1 per cent, mean about 0.020 per cent). The chlorine is usually the
lower, the poorer the original urine was in chlorides. The fall of the

430 Torvald Sollmann.
per cent of chlorine is entirely independent of the diuresis in quantity
and in duration. The low content in chlorine always persists with
little change for an hour, whilst the diuresis becomes very much less.
If further injections are made the chlorine remains low or is lowered
still further. The absolute amount of chlorine is uniformly increased
by a considerable amount, varying with the diuresis.

The following abstracts of the experiments will illustrate these
conclusions : —

Sodium sulphate. — Twelve experiments were made, the injections ranging

from 25 to 75 c.c. per kg.. These reduced the per cent of NaCl (from
0.019—0.555, mean 0.150) to 0.15-0.050, mean 0.018. Examples : —

EXPERIMENT V.

Minutes
after
injection.

Diuresis
in 10 min.

NaCl.

Total quan-
tity of NaCl
in 10 min.

Before injection

Injection of 35 c.c. per kg. of SO,

0-15
15-30
30-45

c.c

0.75

8.0

per cent.

0.280
0.015
0.015
0.015

mgms.
2.10

1.20
3.30
1.95

EXPERIMENT VI.

Before injection

Injection of 35 c.c. per kg.

OFS
30-45

Before injection

Injection of 35 c.c. per kg.

Liffect of Diuretics, etc., on the Chlorides of the Urine. 431

EXPERIMENT XV.1

Time. Diuresis. Per cent.

Beroresmjetuiony. 4 =< . ayers 6.0 0555

Injection of 35 c.c.perkg. . . 0-20 36.0 0.110

20-60 18.0 0.118

60-90 5.3 0.050

EXPERIMENT XVI.

Before injection

Injection of 35 c.c. per kg.

Sodium phosphate. — Three experiments, with injections of 25 to 35 C.c.
per kg. The per cent of NaCl is reduced (from 0.020-0.160) to 0.017—
0.056; the absolute amount per 10 min. is increased (from 0.40-2.40) to
I.10-I1.20 mg. Example : —

EXPERIMENT XVIII.

Time. Diuresis. Per cent.

Beforeimjection 5. 2 . 3 = - seek 0.25 0.160

Injection of 35 c.c. per kg. of |
Name Og ec es, se ne 0.045

0.080
0.063
0.056

1 The columns in these tables correspond to those of Experiments V to XI on
the preceding page.

432 Torvald Sollmann.

Sodium ferrocyanide. — Two experiments, with injections of 25 and 35 c.c.
per kg.

EXPERIMENT XXVII.1

Time. Diuresis. Per cent.

‘
35 c.c. per kg. of 3% alcohol in
9

water; 25 c.c. per kg. Na,SO,4?

c.c. per kg. NagFe(CN), . .~ 0-10
20-40
60-80

EXPERIMENT XXVIII.

Oneal WHE Go a 6 Ss sc mtaiere 16 0.704 iss
35 c.c. per kg. NagFe (CN),-- - 0-10 33.0 0.089 AY EW
10-20 24.0 0.038 9.12

20-40 16.0 0.030 4.80

1 The columns in these tables correspond to those of Experiments V to XI on
page 430.
2 Injections made an hour apart, before the injection of the ferrocyanide.

Sodium acetate. — Two experiments: the diuresis is quite small, but the effect
on the per cent of chlorine is the same.

EXPERIMENT XI.

Time. Diuresis. Per cent.

SoNC G4 PCIe Nas SO ee anes GIO ifs) 0.021

Z5/C:C, per kg. NaCoHaOnns cr 20.5 0.015

13.0 0.022
6.0 0.020

Effect of Diuretics, etc. on the Chlorides of the Urine. 433

EXPERIMENT XVII.

Per cent.

Time. ]iuresis.

@iamalvarne <<. 2S Ls. Bests 16 0.070
go-e-c per ke. NaCjH,05 - . . 2.0 0.065
4.3 0 044

Urea. — This substance, dissolved in water (three experiments) or in sulphate
(one experiment), reduces the per cent of NaCl (from 0.015-0.055) to

0.012—0.046. Example : —

EXPERIMENT VI.

Time. Diuresis. Per cent.

Boece pen ke. NajSOy =. ©. Sone : 0.015
25 c.c. per kg. urea in water . . : 0.015

0.012

The following experiment illustrates how a urine of very low chilo-
rine content may show a slight temporary increase. This was occa-
sionally seen with other salts, but was always negligibly small; it
is probably to be explained by the theory of Cushny, z. e. by a lesser
reabsorption of chlorine due to the diuresis.

EXPERIMENT XXIV.

Na,SO,, cantharidin

25 c.c. per kg. urea in Na,SO,

Time.

Diuresis.

Per cent.

11.0
36.0
6.5

0.027
0.046
0.017

434 Torvald Sollmann.

Dextrose. — Three experiments with injections of 25 or 35 c.c. perkg. The
per cent of NaCl is reduced (from o0.012-0.256) to 0.012-0.061.
Example : —

EXPERIMENT XIX.

Time. Diuresis. | Per cent. Mg.

LeXenrovee’ TNS Ge Saers 2.0 0.256 5.12

35 c.c. per kg. of glucose . . . 0-20 10.0 0.082 8.20

20-40 5.0 0.061 3.05
0.103

Water. — One experiment: this reduces the per cent of chlorine in the same
way as Salts, but as it causes no diuresis, the absolute amount is also
reduced.

EXPERIMENT XXV.

Time. Diuresis. Per cent. Mg.

Watery alcohol, NagSO, . . . dai 15.0 0.036 5.40

25) CC perm ke Tor waters meen 0-15 16.5 0.029 4.79

10.0

0.029

3. Does the injection of the salts stimulate the retention of chlorides ? —
To answer this question, the chlorine content of the urine was com-
pared after injection of 0.5 per cent sodium chloride, dissolved in
water and in 2.3 per cent sodium sulphate crystals. Three animals
were used, the watery solution being injected first in two of these
experiments, last in the other. There was no difference in the per
cent, sodium chloride in water reducing the per cent in urine (from

Lifect of Diuretics, etc. on the Chlorides of the Urine. 435

0.050-0.380) to 0.025-0.180; sodium chloride in sodium sulphate
reducing the per cent in urine (from 0.029-0.180) to 0.020-0.045.
The diminution of the chlorine is, therefore, not due to the presence
of a foreign salt, but to the dilution of the blood. However, the total
amount of chlorine is greater when sulphate is injected, on account
of the greater diuresis. Example: —

EXPERIMENT XII.

Time. | Diuresis. NaCl.

per cent.
IFS yA So ee One ates zs 0.050

SOMeC-ukco. 05% Nacl 5 0.060

0.029
0.021
0.023
0.038

0.033

4. Saline diuretics which do not cause chlorine retention. — Under
this heading come: sodium nitrate, sulphocyanide, and iodide (and pre-
sumably also the bromide; but this was not tested, for want of a suit-
able method). These salts differ from the preceding, not merely by
failure to cause chlorine retention, but they even increase the per
cent of chlorine greatly if this has been artificially lowered. It is also
remarkable that the injected iodide reacts toward sulphate injection
just like the chloride, and that these ions were excreted (in the con-
ditions of the experiment) in approximately equimolecular ratio.
This was not investigated in connection with the other ions. Since
Loewi! had already demonstrated the effect of the nitrate ion, I con-
tented myself with a former experiment, already quoted in part?
and which is inserted here somewhat more in detail, in addition to
the experiments made with the other ions mentioned.

1 Loew! O.: Archiv ftir experimentelle Pathologie und Pharmakologie, 1902,
xlvili, p. 410.
2 SOLLMAXNN T.: This journal, 1902, viii, p. 155.

Torvald Sollmann.

Nitrate experiment.— Dog of 16 kg., injection of 75 c.c. per kg. of 1.23
per cent NaNO, solution (A 0.481), in 10 min.

Time after
injection,

URINE AFTER INJECTION.

Quantity
in 10 min.

NaCl.

NaNOs.

NaySQOjq.

10 min. .
24 min. .

42 min. .

per cent.

0.40
0.38
0.37
0.35
0.28
0.26
0.24

SERUM.

NaCl.

per cent.

1.481

NaNOs.

per cent.

0.217
0 067
0.064
0.051
0.079
0.096

Na,SO,.

Total
solids.

per cent.

4193
2.780
2.570

2.910
2.580
2.640
3.745

Serum
in blood.

Before injection
After injection:

] min.

15 min. .

Thr:

23 hrs. .

per cent.

0.37
0.38
0.43
0.48

per cent.

0.023
0.261
0.219
0.122
0 146

per cent.

0.159
0.131
none
trace

0.023

per cent.

56.22
72.22
65.96
61.58
58.10

It is seen that the urine after nitrate injection has a high per cent

of chlorine, slightly higher than that of the serum.

Liffect of Diuretics, etc. on the Chlorides of the Urine.

Sulphocyanide : —

EXPERIMENT XXVI.

4

Time.

Diuresis.

Nac

Original urine .

Aqueous alcohol, Na,SO4, NayFe (CN),

25 c.c. per kg. NaSCN .

0-10
10-30
30-50

per cent.

0.300
0.027

0 278

0.115
0.054

EXPERIMENT XXVIII.

Original urine .

Na,Fe (CN),

The sulphocyanide has raised the
to 0.300 per cent.

25 cc. per kg. NaSCN .

25 c.c. per kg. NaSCN .

chlorine of the urine (from 0.030)

438 Torald Sollmann.

lodide : —

EXPERIMENT XIV.

Equivalents.!

NaCl | Naci|Nal +

: ; Nal X
+Nal.| X c.c. Pir

p. cent, | p. cent.

INGDSIOVE S tom ge f 01027 00) 0!562))| 2 = O:562" | S1SiGs eso
2 e.c ke. Nal: 0.106 | 0.245 | 1.813 3.432 | 68-9.-|_130:3
0.100 | 0.243 | 1.710 SoH) || SO) |) WO227/
0.133 | 0.212 | 2.280 3.681 | 45.6 | 73.6

25 c.c. kg. phlor-
hizin in water. 0.060 | 0.064: 1.026 1.449) 921058) Noor

0.096 | 0.094 | 1.642 2.263 30.5

EXPERIMENT XV.

NavSOly fe Mate 0 | 0.050) 0.0 | 0.860
25 c.c.kg. Nal. . 0.248 | 0.773 | 4.241
0.200] .... | 3.420

25.c.c. kg. NagSO,

Per cent X 100

molecular weight

1

equals, for NaCl, per cent X 17.1; for Nal, per cent X 6.61.

Sodium iodide (two experiments) raises the percent of chlorine in
the urine. With the quantities which were used, the chlorine and
iodine are excreted in approximately equimolecular proportion, but
not quite so. Injection of sulphate or water cuts down the iodine in
the same way as it does the chlorine.

5, Effect of other diuretic agents. —- Alcohol, oil of juniper, ether,
methylene blue, caffein citrate, or phlorhizin have no effect on the

per cent of chlorine, beyond that of the vehicle in which they are
administered.

Liffect of Diuretics, etc., on the Chlorides of the Urine.

439

Alcohol. — Given intravenously in 3 per cent solution, in water (two experi-
ments) or in % Na,SO, (one experiment), in dose of 25 and 35 c.c.

per kg.

excretion of chlorine may also be diminished.

EXPERIMENT XXVII.

When given in water there is much less diuresis, so that the total
Example : —

Time.

Diuresis.

NaCl.

Before injection

35 c.c. per kg. of 3% alcohol in
WitkC ami etei cr ios mie slate

25 c.c. per kg. of Na,SO, .

8.00

5.00

per cent.

0.300

0.103
0.043
0.092

Funiper. — A mixture of 0.4 per cent juniper oil and 1.6 per cent of alcohol

in “” Na.SO, is given intravenously in the dose of 35 c.c. per kg.

EXPERIMENT XVII.

Time.

Diuresis.

Per cent.

NaC,H30,; alcohol in NagSO, .

Juniper mixture

17.0
40.0

325

0.017
0.019
0.019

Ether. —Vhis was injected hypodermically ; 10 c.c. per kg., divided into
6 doses, being distributed over 50 minutes,*when the animal died.

EXPERIMENT XXV.

Time.

Diuresis.

Per cent.

Alcohol in water, NagSO4; water, Na,SO,; ether

0-10
40-50

0.015
0.012

440 Torald Sollmann.

Methylene blue. — 25 c.c. per kg. of a o.5 per cent solution in 4% Na SO,
was injected intravenously. Two animals died, with anuria and strong
convulsions, within 20 minutes after the injection. The following

succeeded : —-

EXPERIMENT XVI.

|
|

Time. Diuresis. Per cent.

Na,SO,, phlorhizin, caffein . . ciel 2.5 0.014
Methwieiesblueiy-s annem 5.0 0.024
8.0 0.039

ZD 0.057

Caffen. —This was used hypodermically in two experiments (10 to 20 mg.
per kg.) and intravenously, dissolved in % Na,SO,, 25 c.c. per kg. of
0.04 per cent in two others. Both vagi were cut.’ When used hypo-
dermically it caused an insignificant rise if the chlorine was already
low.

EXPERIMENT XIX.

Time. Diuresis. Per cent.

Glucosé?.... 2 55 ee rar ifs 0.103

Caffein intravenously . . . 0-20 : 0.045

20-40 ‘ 0.024
40-60 oh 0.022

' ANTEN, H.: Archives internationales de Pharmacodynamie, 1901, vill, p. 455.

Effect of Diuretics, etc., on the Chlorides of the Urine. 441

EXPERIMENT XIII.

‘Time. Diuresis. Per cent.

Urea . 2.0

Caffein hypodermically . . 6.0
3.0
1.5

Phlorhizin. — This was given intravenously, 25 c.c. per kg. of 0.4 per cent
solution, dissolved in one experiment in water, in another in 4% Na,SO,.

Example : —

EXPERIMENT XIV.

Time. Diuresis. NaCl. Nal.

per cent. per cent.

INES On INE 95 Sooo 4p 6¢ 195 0.133 0.212
21.0 0.060 0.064

Watery phlorhizin
13.5 0.096 0.094

442 Torald Sollmann.

6. Effect of nephritic agents. — Arsenate of soda, mercuric chloride,
(one experiment each), and cantharidin (three experiments) did not
affect the chloride retention. Examples :—

Arsenate of sodium in 1 per cent solution hypodermically.

EXPERIMENT XIX.

Time. | Diuresis. | NaCl. Proteid.

per cent.

Glucose; catfemyNajsOj 7. 4 sue Shae 5.5 0.022 | None.
ATSenatcsel Ohm Sap etal. een: ee 0.5
Z51C-C. Pelakp ys NasS ©, 7mm ane 13.0 Trace.
Arsenate, 20) mp. per ko ie Less.

Trace.

IRepeatedilastidoses.as-aa. vr aeeunnnn: Considerable.

Repeated last dose .

25 c.c. per kg. NagSO, .

Arsenate, 20 mg. per kg. intravenously

The urine stops 10 minutes later, but the animal remains alive for
an hour longer.

Effect of Diuretics, etc., on the Chlorides of the Urine. 443

Mercurie chloride. —The dog has received daily hypodermic injections of
5 c.c. of o.r per cent HgCl, for the five days preceding the operation.
Weight at operation, 8.9 kg.

EXPERIMENT XXIV.

Time. Diuresis. NaCl. Proteid.

per cent.

Bladderuriné <9. « 2 « eles waists 0.050 ‘race:
Secreted while on table . . ote 1.6 0.156 Trace.
35 c.c. per kg. of NagSO, . 0-10 34.0 0.018 Very faint.

10-20 31.0 0.014 Very faint.
Cantharidin ! (10 mg. per kg.) 0-10 24.0 0.018 Very faint.
10-30 11.0 0.027 Slightly more.

EXPERIMENT XVIII.

N,.HPO,, caffein, NagSO, . Sore 4 None.

Cantharidin, 5 mg.perkg. . : Faintest trace.

Cantharidin, 10 mg. per kg. . é Appreciably more.

Died.

1 ] per cent solution in acetic ether, given hypodermically.
Pp g yp df

Experiment XXI.— 5 mg. per kg. of cantharidin was injected on the day pre-
ceding the operation. The bladder urine contains considerable proteid,
but no sugar. Very little urine secretion follows the injection of saline
and other diuretics, but the per cent of chlorine is reduced as usual. The
bladder urine contains 0.354 per cent NaCl; after injection of salines,
the urine contains at first 0.108, later 0.ogo per cent.

Kidneys suffering from aloin nephritis (produced by the hypodermic
injection of 5 c.c. of 5 per cent aloin, daily, for three days) can also
secrete a urine with a chlorine per cent lower than that of serum (0.27
per cent). The animal died before injections could be made.

7. Other factors. — Some other factors, which might be supposed to
have an influence on the chloride excretion were observed inci-
dentally.

444 Torvald Sollmann.

Diuresis. —'The extent of the diuresis did not in any case materially modify
the per cent of chlorine in the urine. Examples : —

EXPERIMENT VII. EXPERIMENT XI. EXPERIMENT XVI.

Diuresis. NaCl. Diuresis. NaCl. Diuresis. NaCl.

per cent. per cent. per cent.

14.0 0.025 0.019 68.0 0.013
3.0 0.025 0.020 mes 0.014

The per cent of chlorine is just as low with a small as with a large urine flow
Examples : —

EXPERIMENT X. EXPERIMENT XIV.

Max. Diuresis. NaCl. Max. diuresis. NaCl.

per cent.

per cent.

c.c,
0.021 33.0 0.027

Time. — The low per cent of chlorine in the urine outlasts the diuresis very
much. In seventeen experiments it remained uniformly low for the hour,
or longer, during which it was observed. In seven other experiments it
shows a slight rise with time. This was so small (never exceeding o.1
per cent) that it seems superfluous to risk any explanations. It may be
due to the return of the composition of the blood toward the normal.
Example : —

EXPERIMEN?T I.

Time after injection. NaCl.

per cent.

Q-17 min. 0.035
17-58 “ 0.020

SealGts) & 0.060

Lifect of Diuretics, etc., on the Chlorides of the Urine. 445

Laking. — This was observed in five experiments, as the result of the injection
of watery solutions of urea, phlorhizin, or alcohol. It did not affect the
chlorine in any of these experiments. Examples : —

EXPERIMENT VII.

NaCl.

per cent,

Na,SO,, glucose. . No hemoglobin . . 0.030

Watery urea . . . Hemoglobin . . . 0.025

EXPERIMENT XIV.

Na SO; Nalges. = No hemoglobin

Watery phlorhizin . Hemoglobin

EXPERIMENT XXVII.

NaCl.

per cent.

Original urine. . . No hemoglobin . . 0.300
Watery alcohol . . Hemoglobin . . . 0.103
0.043
NaS Ofer een Muchless hemoglobin 0.092

Be b 0.019

Ferrocyanide . . . No hemoglobin . . 0.025

IV. DIscussIoN OF THE EFFECTS OF THE VARIOUS FACTORS ON
THE CHLORIDES OF THE URINE, AND THE LIGHT WHICH THEY
THROW ON THE MECHANISM OF CHLORINE RETENTION.

These factors may be divided into four classes : —

Class I. Those which diminish the per cent of chlorides (to 0.008,
0.060 per cent), but which through increased diuresis cause a greater
absolute amount of chlorine to be excreted in the urine. This class
comprises: Solutions of acetate, ferrocyanide, phosphate, and sul-
phate of sodium, urea, and glucose; as also the drugs of class III, if
these are administered in solutions of the above salts.

446 Torald Sollmann.

Class IT. Those which diminish the per cent of chlorine to the same
degree, but which do not increase the urine secretion, and which there-
fore lower also the absolute amount of chlorine excreted by the urine.

This class comprises water, and we could add, deficient chlorine
income.

Class IIT. Those which do not affect the per cent of chlorine mate-
rially; namely -

(a) The diuretics which do not dilute the blood: caffein, phlorhizin,
juniper oil.

(6) Nephritic agents: arsenic, cantharidin, mercury, ether, aloin.

(c) The degree of diuresis, laking of the blood, and (within the
limit of the experiments) the time, and the molecular concentration
and the quantity of fluid injected.

Class IV. Those which greatly increase the per cent, and the ab-
solute amount of chlorine in urines originally poor in this ton. This
class comprises: Nitrate, sulphocyanide, chloride, iodide (and probably
bromide) of sodium.

The results are so constant, both qualitatively and quantitatively,
that the mechanism by which they are produced must be compara-
tively simple. They agree entirely with those obtained through dif-
ferent methods by Loewi! with sodium nitrate, phlorhizin, and water,
on dogs.

Very different results are obtained on the rabbit. Cushny? finds that in this
animal the injection of Na,SO, often increases the per cent of chlorine
in the urine. Pototzky® finds that this is the case if the per cent of
chlorine is originally low, whereas the per cent of chlorine is dimin-
ished if it was originally high. In either case the effect of the injection
is to approximate the per cent of chlorine in the urine to that of the
serum. The same results are seen in the rabbit, with caffein,* diuretin,®
sugar,® urea,’ phlorhizin,® sodium phosphate, and nephritic agents.’

1 LoEwI, O.: Archiv fiir experimentelle Pathologie und Pharmakologie, 1902,
xlvili, p. 410.

2 CusHny, A. R.: Journal of physiology, 1g02, xxvii, p. 429.

8 Porotzky: Archiv fiir die gesammte Physiologie, 1902, xci, p. 565.

* KatTsuyamMa, K.: Zeitschrift fiir physiologische Chemie, tgor, xxxii, p. 235.

5 KatsuyaMa, K.: Loc. cit.; PototzKy: Loc. cit.

6 POTOTZKY: Mopac:

7 KasuYyAMA, K.: Loc. cét.; Pototzky: Loc. cit.

* RuscHHAuPT: Archiv fiir die gesammte Physiologie, tg02, xci, p. 565;
Cusuny, A. R.: Loc. cit.

*SLOEWL, Onn iocncer: 10 RUSCHHAUPT: Loc. cit.

Lifect of Diuretics, etc., on the Chlorides of the Urine. 447

Water and salt hunger being the only agents which uniformly lower the
per cent of chlorine in rabbits’ urine.

The difference between the results of Cushny and those obtained by
Magnus and by myself is therefore explained, as I supposed (p. 16r)
by the different animals which were employed. Zhe mechanism of
chloride excretion reacts differently in rabbits and dogs. I shall recur to
this difference later (page 451) and confine myself at present to the
experiments with dogs.

1. What is the essential factor in the chlorine retention? — If we con-
fine our attention for the present to the first class of diuretics which
diminish the per cent of chlorine in the urine, it will be seen that
they possess the following factors in common: A dilution of the
blood serum, and consequently a diminished per cent of chlorine in
the serum; the presence of a foreign substance in the serum and in
the urine; and an increased diuresis.

A comparison of Class I with Class II permits the elimination of
all but one of these factors as unessential.

The increased diuresis is not essential, for it does not exist in
Class II, and even in Class I the per cent of chlorine is quite inde-
pendent of the diuresis. The presence of a foreign substance is not
essential, for none is present in Class II; nor is the per cent of
chlorine in the urine any lower if a mixture of chlorine and SQ, is
injected, than if a pure chlorine solution is administered. The
dilution of the serum is not in itself essential, for it does not exist
in the chlorine retention produced by salt-starvation.

The essential factor in the production of the low per cent of chlorine
in the urine ts therefore the lowered per cent of this ton in the serum.

2. The per cent of chlorine in the urine is not determined by the total
per cent of chlorine in the serum. — The conclusion stated in the pre-
ceding paragraph is not supported by direct determinations of the
chlorine in the serum. On the contrary, it is found that the urine
remains poor in chlorine when the per cent of this ion in the serum
has returned to normal; and that in some conditions (7. e. if NO, is
introduced) the chlorine of the urine may be high when that of the
serum has been lowered. I quote from some previous experiments.”

1 SOLLMANN, T.: This journal, 1902, viii, p. 155.
2 SOLLMANN, T.: Archiv fir experimentelle Pathologie und Pharmakologie,
TQS, xiv, p.l.

448 Torald Sollmann.

EXPERIMENT VII (former series).

NaCl in NaCl in
serum. urine.

per cent. per cent.

SOS SHH. 6 6 = ee 5 5 6 0.57 0.105

50 min. after injectionof SO, . . . 0.45 0.040

34 hours after injection of SO, . . . 0.54 <0.030

EXPERIMENT VI (former series).

Beforennijection ga. ie eee 0.56
3 min. after injection of SO, . . . 0.43

3% hours after injection of SO,. . . 0.56

NITRATE EXPERIMENT.

1 min. after injection of NO,

1 hour after injection of NO;

In view of this contradiction, it is necessary to have recourse to
Forster's hypothesis that the greater part of the serum chlorides
ordinarily exists in the form of combinations (probably with the
proteids) which are not capable of excretion by the urine, and that
only the free chlorides can be excreted. It is true, as I have pointed
out! (page 166), that the existence of such combinations has not
been directly demonstrated. Until this is done, the considerations

1 SOLLMANN, T.: This journal, 1902, viii, p. 155.

* BuFFA (Archives internationales de Pharmacodynamie, 1900, vii, p. 425)
argues that there is such a combination from the fact that if serum is precipitated
by ammonium sulphate and the precipitate is redissolved in the original quantity
of water, this solution has the same A as the original serum. This shows, he
believes, that the ammonium sulphate displaces the chlorine from its compound
with the serum proteid. It seems to me that the experimental disposition, as far
as can be judged from his meagre description and data, is too crude to allow any
such far-reaching conclusions. His result might be pure coincidence. However,
even if it be accepted, it would not explain the chlorine retention, for according to
him the ammonium sulphate liberates the chlorine and would therefore make it
filtrable.

Lifect of Deruretics, etc. on the Chlorides of the Urine. 449

here laid down are amongst the strongest arguments for assuming
their existence. The conclusion stated in the preceding paragraph
must therefore be modified: The essential factor in the production of
a low per cent of chlorine in the urine ts the lessened amount of unbound
chloride in the serum.

3. Is the urine poor in chlorine secreted poor in this ion, or is it
secreted with the normal content of chlorine, but modified after its
secretion, by absorption of chlorine or by the secretion of water? — If
the urine were diluted after its secretion, the per cent of chlorine
should vary according to the diuresis; if the low chlorine were due
to a reabsorption of chlorine, then increased diuresis, leaving less time
for this reabsorption, should give a urine richer in chlorine; if the low
per cent were due to a secretion of water, then an increased diuresis,
in which the secretion of water is increased, should yield a urine
poorer in chlorine.

Since the diuresis does not affect the per cent of chlorine, the urine
must be secreted poor in this ion. The extremely rapid urine forma-
tion also speaks against secondary changes. We may conclude that:
The urines poor tn chlorides have a low per cent of this salt when they
are first formed, the low per cent, in other words, ts not due to reab-
sorption or to secondary dilution.

4. What is the mechanism by which urine with a low per cent of
chlorine is formed from sera containing a low per cent of free sodium
chloride, but a normal or only slightly subnormal per cent of total
sodium chloride ?

The independence of the chlorine in the urine from the per cent of total
chlorine in the serum has been sufficiently shown, so that it is not neces-
sary to discuss a theory that the kidneys are impermeable to a low per
cent of chlorine, but permeable to a large per cent.

Accepting then that the chlorine of the urine is determined by
the free chlorine of the serum, we have the following theories:

1. The excretion of chlorine occurs by filtration, the chlorine of
the filtrate corresponding to the free sodium chloride of the serum.

2. The excretion of chlorine occurs by secretion, the presence of
free sodium chloride stimulating the renal (glomerular ?) cells to the
secretion of this salt.

A number of facts speak strongly against a pure filtration theory:
(a) No physical filter is known which will effect this separation !
(page 166).

i SOLEMANING dis 3 LoGenGre

450 LTovald Sollmaun.

(6) The separation of such a urine would demand a greater filtra-
tion pressure than exists. This would not hold in the present case of
salt injections, for quite enough foreign salt is present to make the
total concentration of the urine superior to that of the serum. How-
ever, in water diuresis, this is not the case’ (page 172).

(c) My former experiments? on chlorine injection also speak
against a simple filtration. In these it is seen that the per cent of
chlorine in the urine is superior to that of the serum, at a time when
the quantity of serum in the body has returned to normal, and when
therefore a considerable amount of the chlorine of the serum must
be bound.

For instance: In Experiment IV one and one-half hours after in-
jection the quantity of serum has returned to normal, the per cent of
sodium chloride in the serum is 0.703; in the urine is 0.836.

This could only be explained on the filtration theory by assuming
that a considerable reabsorption of water has taken place. If this
were so, then the per cent of chlorine in the urine should be inversely
proportional to the diuresis, for the greater the diuresis, the less
would be the chance for the reabsorption of water. As a fact, how-
ever, the per cent of chlorine in the urine is practically independent
of the degree of diuresis.

If filtration does not suffice to explain the phenomena of chloride
excretion, we are forced to assume a vital mechanism.

I attempted to demonstrate this by studying the effect of renal
stimulants and irritants on the chlorine excretion, which, if the pro-
cess were a vital one, might be supposed to have an influence upon
it. The attempt was unsuccessful, for no such influence was per-
ceptible. This, of course, does not prove that the process-is physical.

It still remains to explain several phenomena.

'5, The saline diuretics increase the absolute amount of chlorine, while
they diminish its per cent.— The saline diuretics, whilst they lower
the per cent of chlorine in the urine, never cause its complete dis-
appearance, and indeed increase the absolute quantity of chlorine
excreted:

In accordance with the above theory this can only be explained by
assuming that the diuretics increase the total amount of free sodium

1 SOLLMANN, T.: Loc. czt.

? SOLLMANN, T.: Archiv fiir experimentelle Pathologie und Pharmakologie,
19go!, xlvi, p. I.

Liffect of Diuretics, etc., on the Chlorides of the Urine. 451

chloride. Since the per cent in the urine is independent of the
amount of diuresis and of the amount of chlorine which has been
removed, it would seem that the liberation of sodium chloride took
place constantly. In other words, when the per cent of free sodium
chloride in the serum tends to fall below a certain minimum, a further
amount of sodium chloride is liberated from its combination.

6. The effect of sodium nitrate, iodide, and sulphocyanide.— Loewi,!
who was only acquainted with the effect of sodium nitrate, assumed
that it caused the passage of chlorine from the tissues into the serum.
This is disproven by analysis of serum, which contains only 0.4 per
cent sodium chloride. That this action plays only a small, if any
part, is also shown by the fact that nitrate brings no more chlorine
out of corpuscles than does sulphate.

A sample of fresh defibrinated dog’s blood is mixed with 14 volumes of 10

per cent Na,SO, crystals; another sample with 1 volume of the sulphate
and 4 volume of 15 per cent NaNO;. The samples are centrifugalized.
The sulphate serum contains 0.200 per cent NaCl, the nitrate serum 0.196
per cent.

The action must be something different. The similarity of these
anions is suggestive: they are all monovalent, and dissociate in equal
degree. It would seem that they are able to displace the sodium
chloride from its unfiltrable combination, which the other anions are
unable to do. The fact that sodium iodide and sodium chloride are
excreted in practically equimolecular proportions, and that sodium
iodide is affected in the same way as sodium chloride by diuretics,
favors this view.

7, Phlorhizin and nephritic poisons. — The fact that phlorhizin
and nephritic agents, which increase the per cent of sugar and of
proteid in the urine, do not affect the chlorine, shows that they do
not increase the general permeability of the kidneys and that their
action is specific.

8. The difference in the behavior of rabbits’ and dogs’ kidneys as
regards the excretion of chlorine. — The differences, which have been
pointed out on page 446, would admit of the following explanations : -—

I. The sodium chloride does not exist in unfiltrable combinations
in rabbits’ serum. This explanation cannot be the true one, since

1 Lorwl, O.: Archiv fiir experimentelle Pathologie und Pharmakologie,
1902, xlviii, p. 410.

452 Torvald Sollmann.

water, or salt-starvation, diminishes the chlorine in the rabbits’ urine
just as in dogs’ urine.

2. The unfiltrable chlorine compounds, either in the serum or
tissues, are broken up by all diuretics. Against this speaks the fact
that diuretics /ower the per cent of chlorine in the urine, if it was
previously abnormally high. 3

3. Diuretics break down the resistance of the kidney to the excretion
of combined sodium chloride. This seems the most likely hypothesis,
and if we consider the greater susceptibility of the rabbits’ kidneys to
nephritic agents, it is not at all unlikely.

It would be interesting to know to which class the human kidney
belongs? This could easily be determined by administering diuretics
(except nitrate) on a milk diet.

V. CoNCLUSIONS.

1. Effect of sodium sulphate injection on chlorine, A, and nitrogen.
The increased diuresis, as usual, tends to lessen the concentration
of the urinary constituents, whilst it increases their absolute quan-
tity. The per cent of chlorine, and the A, are but little affected by
the degree of diuresis. The nitrogen and chlorine vary generally in
the same direction, but by no means in the same proportion, the
factor a varying between 2.0 and 7.9. This factor is independ-
ent of the degree of diuresis.

2. Effect of various factors upon the chlorine of the urine, in the dog.
These factors may be divided into four classes.

I. Diminishing the per cent (to about 0.020 per cent) but increas-
ing the absolute amount: solutions of urea, glucose, alcohol, sodium
acetate, ferrocyanide, phosphate, and sulphate.

II. Diminishing the per cent and the absolute amount : water, salt-
starvation.

III. Without effect: nephritic agents, caffein, phlorhizin, laking,
degree of diuresis, and (within the limits of the experiments), the
quantity or concentration of the injected fluid.

IV. Increasing the per cent of chlorine if this has been low: solu-
tions of sodium nitrate, iodide, and sulphocyanide.

3. Mechanism of the chloride retention. The essential factor is
the lowered quantity of unbound sodium chloride in the serum (not the

Lifect of Diuretics, etc., on the Chlorides of the Urine. 453

absolute amount of sodium chloride in the serum, nor the dilution of
the serum, nor the presence of foreign salts, nor the diuresis).

The low per cent of chlorine is mainly due to the urine being
secreted poor in this salt, and not to secondary dilution, nor to
reabsorption of chlorine.

The uninjured renal cells secrete only free sodium chloride, not
combined sodium chloride. This property is not affected in the dog
by diuretics or by nephritic poisons, whereas in the rabbit these
agents cause the excretion of combined sodium chloride.

The sodium chloride is displaced from its combination by the
nitrate, iodide, and sulphocyanide ions, but not by acetate, ferro-
cyanide, phosphate, sulphate, urea, or glucose.

Increased permeability of the kidneys to glucose or proteid is not
necessarily accompanied by increased permeability to chlorides.

In conclusion, I wish to thank Mr. C. A. Lenhart and Dr. R. A.
Hatcher for much valuable aid in carrying out the experiments.

THE COMPARATIVE DIURETIC EFFECT OF SADEINE
SOLUTIONS!

By TORALD SOLLMANN.
[#rom the Pharmacological Laboratory, Western Reserve Medical College, Cleveland, Ohio.]

CONTENTS.

Page
I. Non-secretion of urine . . MPEP Ss cs IS
II. Influence of glycosuria on the teal faites PRET sks dc a) UNE
Ilf. The time-relations of the saline diuresis .-. . . . .) 9.) )sh gee
IV. Comparative diuretic effect of various diuretics .-. . . . . #005 aueetom
V. Causes of the difference in diuretic effect . . . Sh!

VI. How does the molecular concentration of the tageteesd salen eoaeuee
diuresis? . . aes tie A «45,0.
VII. The comparative aeete alte af diluted and andiluted bloat . (le sens
VITI. “Conclusions 5:08 2 6 ee ee

I. NON-SECRETION OF URINE.

ye is common in a series of diuretic experiments on dogs, a num-

ber of the animals secreted practically no urine. This occurred
in seven out of twenty-four experiments (29 per cent) in which
diuretic injections were made. The urine of six of these non-diuretic
animals contained proteid; in only one animal which gave a poor
diuresis (Experiment X) was the urine free from proteids. In two
of the animals (XX and XXI) the nephritis was due to the previous
injection of nephritics (bichromate and cantharidin) ; in one (VIII),
to the inspiration of liquid ether ; in three (XIII, XXIX, and XXX)
it was not accounted for.

It seems from these results that the main cause of anuria is
nephritis. It can be seen, however, that not every form of albu-
minuria causes anuria. No anuria exists in the acute stage, imme-
diately following the injection of the nephritic agent (cantharidin,
Experiments XVIII and XXIV; alcohol, XVII; arsenate sodium,

' The experiments described in the preceding paper offered an opportunity for
studying the rate of diuresis produced by these salts. The investigation was also
extended to the phenomena of urine filtration which may be observed in the
excised kidney.

454

Comparative Diuretic Effect of Saline Solutions. 455

XIX; ether, XXV); and in subacute mercurial nephritis (Ex-
periment XXIV)’

The resistance of non-secreting kidneys to diuretic agents 1s prac-
tically absolute. As examples I may quote Experiments XX and
ool Phe animals had received: XX, 5 c.c. of 5 per cent
ie@r,0,; XXI, 1 c.c. per kg. of 0.5 per cent cantharidin, hypo-
dermically, a day before the operation.

Neither animal secreted any urine during an hour under anesthesia.
They then received each the following injections, per kg., in the
course of three hours: 140 c.c. Na,SO, solution ; 30 c.c. Na,HPOy,
solution (= 170 c.c. of total fluid); 0.032 gm. citrated caffein ; 0.4
gm. chloral; 0.016 gm. diuretin; 0.18 gm. urea. As the result of
these, dog XX secreted in three hours 10 c.c. of urine (= 1.6 c.c. per
kg.) and had a serous diarrhoea ; XXI secreted in two and one-half
outses.5 G.c. of urine (— ‘1 c.c. per kg:).

In one experiment (X) section of the vagi relieved the anuria some-
what; in VIII this was not effective; in the other five the vagi were
divided before the injections were made.

Experiment X.—
11.25-12.20. No urine.
12.22. Injection of 35 per c.c. kg. Na,SO,; no urine to 12.41.
12.41. Injection of 35 per c.c. kg. NagSO,; no urine to 1.10.
1.10. Divided both vagi.
1.17. Lurst drops of urine.
1.37. 9.5 c.c. of urine in 20 min.
1.57- 15 c.c. of urine in 20 min. Injected 25 c.c. per kg. NagH PO,
2.17. 13 C.c. of urine in 20 min.
2.37. 6 .c. of urine in 20 min.
257. 2:¢.c. Injected 25 c.c. per kg: NaC,H;O;,
3.37. No urine.

It was also noted, in the only experiment in which it was tried,
that Ayperisotonic injections relieve the anurta somewhat.

1 HELLIN and Spiro (Archiv fiir experimentelle Pathologie und Pharmakologie,
1897, xxxviii, p. 368) found that arsenate of potassium, and cantharidin, entirely
suppressed the caffein and the phlorhizin diuresis in rabbits, whereas bichromate
had no effect. This is another illustration of the difference between the dog’s and
rabbit’s kidneys.

456 Torvald Sollmann.

Experiment XXX (only one kidney used). —

Before injections ; no urine in 40 min.

Injection of 40 c.c. per kg. of 4.2 per cent Na,SO, crystals per kg. in
30 min. ; average urine in Io min., 0.67 c.c.

Injection of 35 c.c. per kg. of 4.2 per cent Na,SQO, crystals per kg. in
26 min.; average urine in ro min., 1.0.

Injection of 10 c.c. per kg. of 32.0 per cent Na,SO, crystals per kg.
in 50 to 65 min.; average urine in 1o min., 6.0.

II. INFLUENCE OF GLYCOSURIA ON THE INITIAL DIURESIS.

It was noticed that almost all good diuretic animals had a pro-
nounced glycosuria, and vice versa. The sodium chloride per cent of
the original urine, on the other hand, seemed to have no effect on the
diuresis.

The influence of the glycosuria is well seen from the following
epitome of the urines which were secreted before any injections were
made:

Vo sugar: g cases; diuretic factor’ from o to 1.9, mean 0.7.
Small amount of sugar: 3 cases; diuretic factor from 0.7 to 2.8, mean 1.5.
Considerable sugar: 11 cases ; diuretic factor from o to 6.6, mean 2.0.

II]. THe TIME-RELATIONS OF THE SALINE DIURESIS.

The diuresis sets in very quickly after the injection of the saline
solution, reaches its maximum usually in the first ten minutes, is
maintained for about thirty minutes, and then declines quite rapidly.
The amount secreted in ten minutes is often as much as 30 to 35 C.c.
The onset of the diuresis occurs frequently during the injection.
This is seen very strikingly in Experiment XIV. No urine had been
secreted in the twenty minutes preceding the injection, but within a
minute after starting the injection there appeared also the first drops
of urine. In a few instances there was some delay, as in Experiment
XVI, in which methylene blue in sodium sulphate was injected; the
diuresis and blue appeared in ten minutes.

Repeated injections, after an interval of an hour, again restore the
diuresis, and the same phenomena are observed as after the first
injection. The diuresis is usually about the-same for each injection,
if the same solution is used; sometimes it is rather greater, and some-

' Cubic centimetres of urine secreted in one hour, divided by the weight of the
animal in kilograms.

Comparative Diuretic Effect of Saline Solutions. 457

times rather less in degree (Experiment XV), or jess lasting (Ex-
periment XVI). This is presumably due to the lowered circulation
from the prolonged anesthesia.

IV. CoMPARATIVE DivuRETIC EFFECT OF VARIOUS DIURETICS.

The uniform technic employed in these experiments offers an
excellent opportunity of comparing the diuretic effects. For this
purpose I chose as déuretic factor of the maximum rate of diuresis
the maximum number of cubic centimetres of urine secreted in
forty consecutive minutes, divided by the weight of the animal
in kilograms.

This factor disregards the rate of urine secretion before injection, for this has
no appreciable effect upon the quantity of urine secreted in response to
the diuretics. The period of observation is sufficiently long to eliminate
accidental variations. The period of maximum diuresis is chosen rather
than a definite period after the injection, to take account of the variations
in the onset of the diuresis. As will be seen, this factor is really very
constant, with different animals, if the same solution is used ; it is much
more uniform than any other factor which I have tried to apply, and it
varies in a constant manner with different solutions. “The factor is only
computed to 0.5. All animals, the diuretic factor of which remained less
than 5 after saline injections, are excluded as abnormal.

Variations tn the quantity of the injected solution (between 25 and
75 cc. per kg.) have little effect upon the rate of diurests «
75 c.c. per kg. of Na,SO, solution: 3 experiments ; diuretic factor: extremes,
6tO 14-2; mean, 13-
25 and 35 c.c. per kg. of Na,SO, solution: 9g experiments ; diuretic factor:
extremes, 10 to 20; mean, 13.

The same holds of other salts, or of successive injections; 2. ¢. if
several injections are given an hour apart, as was done in my experi-
ments, the diuretic factor is practically the same for each injection.!
This fact is very important, for it permits the comparison of all the
injections, no matter in what order they were given. In the fol-
lowing table, the means of the results obtained on different animals
are compared.

1 It may be well to repeat that this diuretic factor relates only to the maximal
rate of diuresis. It is quite probable that the diuresis is more prolonged and
therefore absolutely greater when larger quantities of solution are injected. My
experiments do not bear on this question.

458 Torald Sollmann.

MAXIMAL RATE OF DIURESIS WITH THE VARIOUS DIURETICS.

25 to 35 c.c. per kg. of the solution, injected into the femoral vein in three minutes.
The braces join those solutions which have practically the same diuretic factor.

Diuretic factor.
Solution. pee
Mean. Extremes.

{Nacl (OG Ane er eke cea ese ce 1 0.5 to 2.5 3
WAleatol SU in water anes <0 1? ile 720) 2
PNAC, HyOx rita ens Bak | 4 15 * 65 2

NaSO, 7% with NaCl05% . | 6 LOe 395 3
pase Pe ee eee 6p be Ri 1
3 Glacese: 7s 9 jc tee Verte Ps ae | 63 6.0 to 6.5 2
NialS CNG) ieee A le 74 6:0) 5585 3
{Nal ei ie: wack Bee MEM : 10 5:0 5° 5 2
Mee Mile Payee eae 10 7.0 “ 12.5 2
fNa2SOq E,W 0 er 13 10.0 “ 20.0 p
UNagH POs?) 0 thee ee 13} 9.5 “17.0 2
(7 Na,SO, + alcohol3% . . | about 153 BO OC 1
| ” Na SO, + caffein 0.04% . . 8 git 15.0 to 20.0 2
| 7 NagSO, + juniper ol 04%. | “ 18 s000 1
| 7 Na gSO, + phlorhizin 0.4 % | 183 eae 1
{Nas Re(CN), Lien meee | 203 16.0 to 24.5 2

! 75 c.c. per kg.

This table, deduced from the means, can be well supported by the
study of individual experiments in which different salts were injected
successively into the same animal:

Water. Less than SO,, Experiment XXV.

Alcohol in water. Less than SO,, Experiments XXV and XXVI.

Acetate. Less than SO, or PO,, Experiments XI and XVII.

Glucose. Less than SO,, Experiment XIX.

SCM. Less than SO, or Fe(CN),; more than alcohol in water, Experi-
ments XXV and XXVI.

Comparative Diuretic Effect of Saline Solutions. 459

fodide. Less than SO,, Experiments XIV and XV.

Urea. Less than SO,, Experiment VII; more than glucose, Experiments
VI and VII.

PO;. Less than SO,, Experiment XVIII; more than (7H. Os and o@7.
Experiment XI.

FeCl). More than SO, and SCN, Experiment XXVI and XXVII.

SO, and alcohol. Greater than C,H,O,, Experiment XVII.

SO, and juniper. Greater than C,H,O,, Experiment XVII.

Caffein in SO, Greater than SO,, Experiments XVI and XIX; than
glucose, Experiment XIX.

Lhlorhizin. Greater than SO,, Experiment XVI.

V. CAUSES OF THE DIFFERENCE IN DIURETIC ILFFECT.

The following table shows that the diuretic effect of the saline
diuretics is generally proportional to the dissociation, z. ¢. is mainly a
factor of the molecular concentration of the injected solution :

In 7% solution,
1 mol. disso-
ciliates into!

|
| Mean diuretic
| factor.

NaC HiOg. 7 «| | 1.03
NaNO; ar. an | 1.8
Glucose
NaSCN
Neil
Urea
NaySO,
Na,liPO,

Na,zFe (CN),

1 Compiled from Hamburger (A) and Ostwald (A).

It is seen that a molecule dissociating into 2.6 molecules gives the
highest diuresis; one dissociating into 1.8, less; 1.03, least. There
are, however, breaks in the series: glucose, and especially urea, stand
high; Nal stands above NaNO, or NaSCN.

The conclusion is therefore justified that the diuretic power of
saline solutions is proportional, in the first place, to their freezing-

460 | Torald Sollmann.

point;- and therefore in equimolecular solutions (in the sense of
Hamburger) to their dissociation. Equiosmotic solutions, however,
also vary. If this difference between equiosmotic solutions is ex-
pressed as “specific diuretic’ power, then the specific diuretic power
of urea and glucose surpasses the salts, urea surpasses glucose, iodide
surpasses nitrate and sulphocyanide.

These conclusions appear in agreement with the results of Haake
and Spiro,! although the differences in the methods makes comparison
difficult.

VI. How bors THE MOLECULAR CONCENTRATION OF THE INJECTED
SOLUTION INFLUENCE DIURESIS?

The influence of the molecular concentration on diuresis which
these experiments illustrate, was already shown by Miinzer.2 The
explanation which has been given seems very simple. The hyper-
isotonic solution draws fluid fromthe tissues, so that the result is
the same as if so much more fluid were injected.

This explanation is not sufficient in regard to the maximal rate of
diuresis; for a substance dissociating into 2.6 molecules could in-
crease the injected 25 c.c. of fluid to no more than 70 c.c.; and as
has been shown (p. 457), such an increase in the amount of isotonic
fluid would not alter the rate of diuresis. Some other explanation is
necessary. This could be sought in a vital stimulation of the renal
epithelium by the hyperisotonic solutions. However, it is not neces-
sary to have recourse to such an obscure explanation, for the same
phenomenon is seen on the perfusion of excised and dead kidneys
with salt solutions.

Method of the perfusion experiments.—The kidneys were exposed by a large
incision through the linea alba and another incision along the border of
the ribs. A cannula was inserted into the ureter and another into the
renal artery, toward the kidney, and connected at once with a bulb con-
taining the warm perfusion fluid (saline solutions, with or without the
addition of blood) placed at a height of 140 cm. above the animal. The
perfusion was begun; a cannula was then placed in the renal vein,
toward the kidney, and the kidney excised, the capsule being usually
removed. An outflow of fluid from the ureter began at once, and con-

1 HAAKE and Spiro: Beitrage zur chemischen Physiologie, 1902, ii, p. 149.
* Mtnzer, E.: Archiv fiir experimentelle Pathologie und Pharmakologie, 1898,
xli, p. 74.

Comparative Diuretic Effect of Saline Solutions. 461

tinued indefinitely. The rate of urine flow varied from ? to 22 c.c. in
to minutes. That this fluid did not result from a rupture of the blood-
vessels, is shown by: —

1. The low filtration pressure used.

2. If blood is circulated after the saline, the urinary fluid remains free
from corpuscles.

3. The changes in the composition of the urinary fluid on changing
the perfusing fluid are only complete after some 20 to 4o minutes.

4. The secretion varies with the vein outflow, the injection pressure
remaining unchanged.

5- If blood diluted with Na,SO, is circulated, the per cent of NaCl in
the urine is generally less than in the serum.

The following typical experiment shows that ¢he rate of circulation
and of urine formation varies with the osmotic pressure of the perfusing
solution -

EXPERIMENT XXXVI.

Flow from ureter Flow from vein

Perfusing solution. : : p ;
g in 10 minutes. in 10 minutes.

c.c c.c

NaCl $% + Na,SO, crystals 115% . 6.0 95.0
6.0 40.0

MACWOOU en 2) om Be . 15.0

NaCl 1.95%

NaCl 3%

Na,SO, 8%

The increased urine flow is simply a consequence of the increased
circulation, for others of my experiments, which are reserved for a
later paper, show that in the dead kidney, as in the living, the filtra-
tion of urine depends mainly upon the rate of perfusion, and but little
on the filtration pressure. In the light of these experiments the
superior diuretic action of hypertsotonic solutions 1s explained very

A462 Torald Sollmann.

simply, by their causing a shrinkage of the cells, and through this in-
creasing the lumen of the blood-vessels, and hence the rapidity of the
circulation.

That we are dealing here with a purely physical phenomenon is shown by the
fact that the changes of circulation are the same in kidneys which have
been excised for three days.

Vein. Ureter.
NaGlY, ao Face a lO 3
eel ety icy erica. Semel Co(0, 1
WH 1

Strong alcohol also increases the circulation.

EXPERIMENT XXXVII.

Time. Ureter.

NaCl 3.5% 2.30

2.45
3.00
Alcohol 80% Bals
Bal5
NaCl 2% 3.25

This result explains, on a physical basis, a number of phenomena
which had hitherto to be regarded as proofs of “vital” secretory stimu-
lations.

(a) The superior diuretic action of hyperisotonic solutions, and
the fact that intravenous injection of water has but little diuretic
effect.

(6) Why the addition of a urinary constituent (harnfahige Sub-
stanz) increases diuresis. This has been noted by all observers who
worked with sluggish kidneys. Munk!, for instance, in his perfu-
sion experiments, found very littie secretion unless some such con-
stituent was added. This he interprets as due to a stimulation; as
he increased the concentration (NaCl was added to 2 per cent, etc.),
the phenomenon is identical with that just described.

1 Munk, I.: Archiv fiir pathologische Anatomie, cvii, p. 291.

Comparative Diuretic Effect of Saline Solutions. 463

VII. THE ComparRATIVE DiuRETIC EFFECT OF DILUTED AND
UNDILUTED BLOop.

A blood diluted with saline solution circulates more rapidly, and con-
sequently filters more urine, than an undiluted blood. Examples: —

EXPERIMENT XXXIII.

Flow from Flow from

1 part of defibrinated blood to ; y
vein. ureter.

3 Part isotonic saline solution | 0

ce « “ 3 few drops

1 Really belongs to previous period.

EXPERIMENT XXXVI.

(The kidney has lain for two days.)

|
| Flow from Flow from

vein. Ureter.

Pure saline

Blood 1, saline 6

Pure saline

A64 Torald Sollmann.

These two experiments, selected from a number of others,’ illus-
trate a fact which is really self-evident, but which seems to have been
neglected entirely by experimenters:

(a) Munk? found that ordinary defibrinated blood would not cir-
culate through the excised kidneys, unless it had been diluted with
saline solutions. He seeks the cause of this phenomenon in some
clotting which, however, he could not demonstrate. The cause is
simply that the undiluted blood is too viscid to circulate through the
contracted vessels of the excised kidney.

(6) The diuretic effect of hydramic plethora is partly, at least,
explained by the dilution of the blood and the consequent more rapid
renal circulation; conversely, the absence of diuresis on the injection
of blood (Magnus®*) is explained by the thickening of the blood and
consequent slower renal circulation.

(c) It is not justifiable to estimate the rapidity of the circulation
in the kidney by measuring the kidney volume, as is commonly
done; for by dilution the renal circulation may be quickened without
changing the calibre of the vessels; and by dehydration (hyperiso-
tonic solutions) the calibre of the vessels may be altered without cor-
responding changes in the volume of the kidney.

VUHI. Concriusions.

1. The maximal rate of diuresis is fairly uniform for a given salt.

2. It is not markedly influenced by variations of 25 to 75 c.c. per
kg. in the quantity of fluid injected.

3. In equimolecular solutions it varies generally with the number
of dissociated ions, but urea has a greater diuretic effect than glucose:
either is more active than the salts in equiosmotic solutions; the
iodide is more diuretic than the nitrate or sulphocyanide.

4. The superior diuretic effect of hyperosmotic solutions cannot be
satisfactorily explained by the greater.hydraemia. The perfusion ex-
periments show it to be due to increased circulation through the kid-
neys, produced by the dehydration of the renal tissues. Hypoisotonic
solutions have the opposite effect.

* Some experiments are not very clear, because a solution containing a large
amount of blood often obstructs the vessels so that the following solution does not
circulate for a considerable time.

2 MuNK, I.: Loc. cit.

3 MAGNuS, R.: Archiv fiir experimentelle Pathologie und Pharmakologie, 1gor,
xlv, p. 250.

Comparative Diuretic Effect of Saline Solutions. 465

5. Hydrzmic plethora increases diuresis, amongst other mechan-
isms, by diminishing the viscidity of the blood, and therefore increas-
ing the circulation in the kidney.

6. The volume of the kidney is not a safe index of the circulation
through this organ, even if the arterial pressure remain constant.

7. The urine filtration in the excised kidney depends much more
upon the rapidity of the blood-flow than upon the arterial pressure.

THE RESPONSE OF THE FROG TO: LiGker

BY ELLEN. LOREEDE.

CONTENTS.
Page
I. Introduction. . . : MS moe ASE,
II. Response to white Baht § in the (Gameemie of chs laboratee et Oe eC
Diffuse lights. s-ccs > Se aS ow 8 ee eps Soe ree eS
Directsliitt i. 0a 5 eben se ce. ia Rooting We hiss a as
Diffuse light versus Smo Cr mermrre mie me G5 il
Light reflected from below = % 35. 5 2 = « a0 os ety ee
Light transmitted from above s: “95 29 62) 3) 8 ree ce
Phototaxis:in, water. .) . s. eely .obee 2 oe "Ss en
Orientation . . 56 atin AO ka) ee
Light transmitted dhraush a ‘eeenne Prism.) 2.59 coe aree eee
The orientation of the frog with one eye covered . . . . =» s GN w see ueee
Effect of prolonged light. . . . 3 1a Sy ae eee
III. Response to white light in increased aa in iereree tometer + eer
The effect of increased temperature; . . . .. .- ©) s: susie ee
The effect of lowered temperature . 9. .0 . ss 7. 2) eee eee
Stereotropism . . ‘ Nee she 25/77)
Effect of darkness on upward anal Gawinward’ movements in water . . . 477
IV. Response to monochromatic light .-. . . :; .|. 9. «= 9) s.r
Red i) sce tee ew Jewel ty aoe 4 Be 0 oct oe
Yellow 20 0%. te 8 In - sl pee et ew ke 8)
Green © 20 sole eR ne ep
Blue esse ; 478
Response to diorently colored iehts admitted at opaos onde of a
TECeptacle sn. we 478
Unequal amounts of light tranenvitieds throught the seat an eiveeen ine
blueimediaye cnr: 5 ‘ PE aS)
Response to a red and toa Blue pakerounde eas 480 »
Response when the entire environment is one-half blue aed one- half ied 485
Response when white light is admitted at opposite ends of a receptacle,
one-half of the surface of which is painted red, the other half blue . 485
V. Conclusions... 2. 6) 2 Sl se ek ae tes Sor tc

I. INTRODUCTION.

i his study of light-response and color-sense of animals, Graber!

states that frogs are negatively heliotropic. Loeb,? however, in
his paper on the extension of heliotropic phenomena in the animal
kingdom, finds them positively heliotropic.

‘ GRABER: Grundlinien zur Erforschung des Helligkeits und Farbensinnes der
Thiere, Prag, 1884.
* LoeB: Der Heliotropismus der Thiere und seine Uebereinstimmung mit dem
Heliotropismus der Pflanzen, 1890, p. 89.
466

The Response of the Frog to Light. 467

Loeb does not give a detailed account of his observations, but
Graber gives tabulated results of experiments carried out between
October 10 and 20. He used a large box about 2 cm. high and
divided into two compartments, one of which was dark, the other
illuminated with diffuse light (daylight). Three series of experi-
ments of ten trials each were performed with Rana esculenta, forty
frogs in each trial. The frogs were placed in the boundary between
the light and the dark compartment, and each trial covered a period
of fifteen minutes. The totals of the results are as follows:

I IT III
Sto pee elas 166 174
Darke as 29) 267 Zan 226

indicating a reaction-proportion of 15:10.

Loeb finds that frogs move to the source of light, through whatever
colored medium it be transmitted, a quantitative difference only being
observable between the effects of lights of differing refrangibility.
Graber had found, in three series of experiments of ten trials each,
that 736 responded to the red, and 464 to the blue; the reaction-
proportion being 6:10.

Loeb does not state whether or not he made observations with
the intermediate colors, green and yellow. Graber, however, tested
the response to green but not to yellow. As compared with red, the
results in the two series of experiments were,

I IT
Dark Red) ws. 1450 Briphtzereen) paar oo0
the reaction-proportion being 10:13.
Compared with blue, he found the responses to be,
I I
Brightgreen =. . - 440 Dende ible 6 5 4 AsO)

the reaction-proportion being in this case 7:11. He compares the
“attraction-strength ” !.of the colors, and finds them to be,

Red. Yellow. Green. Blue.
1 ye 0.7 0.5

In order to carry out more detailed observations than have hitherto
been made on the frog, and in order to determine conclusively, if

1 GRABER attributed the response to differently colored lights as an exhibition
of the color-preference of the animal.

468 Ellen Torelle.

possible, its orientation to light, Dr. Morgan, to whom I am deeply
indebted for kindly criticism, suggested that a series of experiments
be performed to include responses of the frog to—(1) diffuse light,
(2) direct light, (3) light reflected from below, (4) light transmitted
from above, (5) light transmitted through a gelatine prism, (6)
orientation with one eye covered, (7) orientation and reaction in
high temperature and in low temperature, (8) response to monochro-
matic light.

The experiments were carried out in the Bryn Mawr biological
Jaboratory from October to February, 1902-03. The material con-
sisted of the two species, R. virescens virescens and R. clamata. The
frogs were kept in an aquarium in the laboratory, where they seemed
to remain in good condition.

In testing responses to light, two boxes were used as receptacles,
the inner surfaces of which were painted a dull black, with the ex-
ception of the glass surface of one wall, through which the light was
admitted. Each of the covers of the boxes contained a longitudinal
slit, about an inch wide, permitting the movements of the frog to be
observed without removal of the cover, a strip of black wadding
which could be noiselessly lifted being laid over the slit.

At first, no time-limit was set for each trial, the frogs being kept
in the box from ten to twenty minutes, or more, a varying numter
of frogs being used in each set of experiments. Later, all these ex-
periments were repeated, a time-limit of ten minutes was set for each
trial, and the same number of frogs was used in each set of experi-
ments, except in cases where lack of conclusive response seemed to
call for the trial of more individuals. Only one frog was used in
each trial.

The results obtained fall into three divisions: Response to white
light in the temperature of the laboratory; response to white light
in increased and in lowered temperatures; response to monochro-
matic light.

II. RESPONSE TO WHITE LIGHT IN THE TEMPERATURE OF THE
LABORATORY.

Diffuse light. — ‘The response to diffuse light was first observed; a
tin box, nine inches long, five inches high, and twelve inches wide,
prepared as above described, served as a receptacle. In the first
set of experiments, six frogs were used. Each was placed at the

The Response of the Frog to Light. 469

rear, which was the darker end of the box, with the head turned
away from the source of light. Though the time of response varied
with each individual, as a rule, from one-fourth of one minute to one
minute sufficed for turning and for moving the twelve inches to the
opposite or light end of the box. While there was usually some
movement from side to side of the box at this light end, the frog
remained here during the rest of the time, which in this first set of
experiments was twenty minutes or more. In all cases, whether
moving or resting, the median plane of the body of the frog was
parallel with the incoming ray.

In a second set, of observations, made a month later, five frogs
were used. Each trial lasted ten minutes, and the results were sub-
stantially as before.

These results seemed to indicate that the species of frog used were
positively phototactic to diffuse light (daylight), and that in diffuse
daylight the orientation of the frog was such that the median plane
of the body was placed parallel to the incoming ray.

Direct light. — The apparatus used in the foregoing experiment
was used also in determining the response to direct light. Five frogs
were used in the first set of experiments, and five in the second.
The sunlight fell into one end of the box only. In each case, the
response was immediate and positive. The animals moved directly
to the illuminated end of the box, where they remained a variable
length of time, from two to four minutes, when they moved backward,
just outside the circle of bright illumination, where they remained
until taken away, the median plane of the body being parallel to the
incoming ray. In most cases, when the sun-illuminated area was
small, the head was not turned from the light during the retreat, which
was accomplished by moving first one side of the body, then the
other, sidewise and backward. In other cases the frogs turned at
right angles to the light, hopped outside the area of intense illumi-
nation, and orientated themselves with their heads in the direction
of the incoming ray.

Since the retreat into the area of less intense illumination might
have been caused by the heat of the sun’s rays, the experiments were
repeated, heat being cut off by placing a glass vessel with parallel
sides three and one-quarter inches apart and filled with water, close to
the glass end of the box containing the frog to be tested. In each
case, the result was practically the same as before. When the frog
was placed in the rear of the box with its head directed from the

470 Ellen Torelle.

source of light, it turned, moved into the sunlight close to the glass
end, where it remained a short interval, and then retreated as before,
remaining in a resting position in the area of lesser illumination until
removed. When placed within the area of intense illumination, with
the head directed toward the source of light, it left this region as
before.

Tests were made out-of-doors, with the animals unconfined and
free to move into a shadow from the sunshine or vice versa. Ten
frogs were tried. Each was placed on a glass plate, covered with a
bell-jar rendered impervious to light, and carried onto the lawn near
the laboratory, where it was deposited about three yards from the
shadow of the building, the bell-jar removed, and the frog left on the
plate with its head turned away from the sun and from the shadow
of the building.

The frog at first hopped forward, then stopped, turned in the direc-
tion of the sun, and hopped well into the shadow, where it remained
quietly for ten minutes. Jt was then moved into the sunshine, in
about its former position. Again it turned and hopped into the
shadow. The results were very much the same in the case of each
frog tried; there was a positive and decided movement from the
sunshine into the shadow.

Since the sun’s rays and the shadow of the building during these
experiments were in exactly the same direction from the frog, it was
impossible to decide whether the movement was due to a response
to the direction of the ray or toward a shadow. Therefore, later in
the day, when the shadow of the building became oblique to the
direction of the sun’s rays, the experiments were repeated, the frog
being placed in such a position that if it moved into the shadow it
must hop at right angles to the direction of the rays. In each case,
the results were substantially these: First, the frog turned in the
direction of the (sun) ray; second, it moved .quickly into the shadow
by a direct path. The experiments were repeated on different bright
days, but the results were always the same as regards movement
from the sun-illumined area into the shadow. In some instances the
frog remained in the grass; in others, it moved close to the wall of
the building.

The question now arose — Does the frog recognize the shadow as
an area of less intense illumination, or would it move toward or onto
a black surface as well if this were placed in the sunlight?

The side of a large wooden box was covered with black cloth, and

The Response of the Frog to Light. A7I

the frog placed near the black perpendicular surface. It hopped close
to this, remained but a couple of minutes, then moved to the wall of
the gray-colored building, where it remained at rest in the angle
formed by the wall and the ground. When placed near the un-
covered box (pine) on the side in full sunlight, there was no move-
ment toward it. When the box was raised on one edge and propped,
so that the other edge was about four inches from the ground, the
frog moved toward the shadow thus formed, crept well under the box,
placed the body between its floor and the ground, where it remained
with its head directed outward.

A black cloth was fastened close to the ground in the centre of a
sun-illumined area, and a frog placed near it moved onto it, crept
along the edge as if seeking cover, then hopped off. A second
frog also hopped onto the cloth, but almost immediately moved off.
Apparently a dark surface, brightly illuminated, does not produce the
effect of a shadow or of diffuse light.

Tests were also made at mid-day on a level tract of ground about
two acres in extent, which contained neither trees nor any object that
could cast a shadow. Six frogs were tried. When freed, each
moved indifferently toward any point of the compass, but usually kept
on moving in the direction in which it began to move. In several
trials no movement resulted; the frog crouched low between short
bunches of grass, its head held close to the ground. When dark
black or dark brown screens were placed in the middle of this area,
and the frogs placed within five yards of them, the movement was
toward and into the shadow of the screen, where they usually re-
mained indefinitely.

Diffuse light versus sunlight. A tin box eighteen inches long, three
inches high, and three inches wide, painted a dull black inside and with
the opposite ends, consisting of glass plates, placed so that the sun’s rays
were transmitted through one end and diffuse light through the other,
was used as a receptacle. Five frogs were tried, each being given
three trials, in each of which the first position in the receptacle was
changed. That is, the frog was deposited first at the end at which
diffuse light was transmitted, then in the middle of the box, then
near the end at which sunlight was transmitted. In each case the
frog turned toward and moved to the end at which the direct ray
was transmitted, but did not remain within the circle of most in-
tense illumination. In some cases it méved to the opposite end of
the box; in others, without turning, it retreated into the area of less
intense illumination.

472 Lillen Torelle.

This experiment corroborates and reinforces the results obtained
with diffuse light and direct light.

Light reflected from below. — In this set of experiments, a tin box,
nine inches long, five inches high, and twelve inches wide, was used, all
of whose surfaces were painted a duil black except the floor, which was
made of window-glass. The box was supported so that the move-
ments of the frog could be watched from below as well as from above.
In the first set of experiments, five frogs were tried; in the second,
fifteen. The results in both cases were the same, and differed with
the amount of light (diffuse) admitted.

(a) When light was reflected from the whole area of the lower
surface, the frog remained in normal resting position.

(0) When light was reflected from one-half of the lower surface, the
frog hopped toward the light area.

(c) When the light was reflected from one-third of the surface, there
was movement toward the light area, but the head was held at a
greater angle to the horizontal.

In all the above trials it was found that the less the amount of
light admitted, the greater the angle of the head to the horizontal
plane of the floor of the box; so that, when light was reflected from
the entire lower surface, a normal resting position was taken, about
two-thirds of the ventral and posterior part of the body resting on
the plate. When two-thirds of the lower surface of the box was
covered (opaque to the light) only one-third of the ventral and pos-
terior part of the body rested on the glass plate. In each case the
frog moved from the darkened area onto or near the lighted area.

Light transmitted from above. — The box used in Experiment III was
used here, the glass plate serving now as the upper side of the box.
Eighteen frogs were tried; the average of results was about the same.
In all cases the response to the direction of the incoming ray was
immediate. The body was raised to an angle of about 45° to the
horizontal. If a portion of the upper surface was covered, the frog
moved to the uncovered side. Frequently, too, the frog jumped
upward, toward the source of light.

Later, it was seen in experiments on five frogs, that the angle of
inclination of the body varies as the distance of the frog from the
upper illuminated surface. Each frog was placed in a tall glass jar
which rested on a black cloth and was covered laterally by an opaque
black cloth. If the entire lower and one-half of the lateral surfaces
were covered, the angle of inclination of the frog’s body was about 45°.

The Response of the Frog to Light. 473

If the entire jar was covered the body was raised so that the forelegs
were as nearly as possible at right angles to the horizontal bottom of
the jar. This made the inclination of the body 60° or over. Fre-
quently the frog assumed an almost erect position, by means of placing
the forefeet against the side of the jar. Some of these results can be
demonstrated at any time by simply placing a frog in a tin pail and
covering the pail with a wire gauze. The results are valuable here,
together with those of the foregoing experiments, as showing that
the frog is positively phototactic to light coming from any direction.

Phototaxis in water. —Is the frog positively phototactic in water?
In order to answer this question a frog was placed successively in
tubes of varying diameters, the smallest being one and three-eighths
inches, one end closed with wire gauze, the tube placed at angles of
inclination varying from 45° to a plane parallel with the floor of the
_ receptacle, the end covered with gauze being held near the wall of
the receptacle. Light was admitted from one end only, and the
tubes were completely immersed in water. Five frogs were tried,
each in three trials. All moved close to the illuminated end of the
tube.

Orientation. — In the first five experiments, the floor of the recep-
tacle was bare, being kept moist by occasional rinsings with cold
water. It seemed desirable to ascertain if the movements and accu-
racy of orientation would be affected by the presence of a bank of
sand or pebbles in the box, between the source of light and the frog.
Upon six inches of the central longitudinal area and across the entire
width of a box nine inches wide, twelve inches long, and five inches
high, a bank of sand two and one-half inches high was made, the sides
of which sloped gradually toward the darker and toward the illumi-
nated ends of the box. Twelve frogs were tried. The movements of
one frog will be followed as illustrative of the response of all.

The frog was placed in the rear compartment, with its head turned
from the source of light. It immediately turned around, moved to
the bank, where it paused, craw/ed, not hopped, up the bank to the
top, then across the plane surface to the opposite edge, where it re-
mained one and one-half minutes, then crawled down the bank and
moved close to the glass at the light end of the box.

The same frog was again placed in the rear compartment after
about one and one-half inches of water had been poured into it. It
swam about at first, then crawled up the bank in the direction of the
light, turned again toward the water, but soon moved to the lighted

474 Ellen Torelle.

edge of the bank, where it remained four minutes, when it was. re-
moved and water poured into the lighted end of the box. Within one
minute the frog had crawled over the bank and into the water at the
lighted end, where it remained during the rest of the experiment, or
nine minutes, moving from side to side of the box with its head
against the glass end.

The responses of the other eleven frogs varied somewhat with the
individual, but were in the main like the one above described.

Light transmitted through a gelatine prism. — A triangular prismatic
plate, three inches in diameter at the base, was made by mixing
lamp-black with dissolved gelatine and allowing it to become firm.
This was then placed in front of the glass end of a box with the
thick end of the prism to one side, so that light of differing intensities
was admitted at the same time into the box. From time to time the
position of the prism was reversed. In all cases the frog hopped to
the side of the box at which the most light was transmitted, z. e.
the thin part of the prism, with the median plane of the body in the
direction of the incoming ray.

The orientation of the frog with one eye covered. — The left eye was
first covered with black cambric of several thicknesses, cut and sewed
together so as to fit smoothly over the left portion of the head, above
the nostril and anterior to the tympanum. This cap-like garment
was fastened to a cambric band passed around the body just posterior
to the forelegs. The frog was placed in the box used in the former
experiment, with its head directed from the source of light. It
immediately turned, with its right eye directed toward the source
of light, z. ¢. with its body oblique to the incoming ray. The angle of
deviation from the direction of the incoming ray differed in different
individuals. Five frogs were tried, but in no case was the orientation
that observed when both eyes were uncovered.

Next, the right eye was covered in the same way that the left one
had been, and now the frog orientated itself so that the left eye was
directed toward the incoming ray. The frog was then removed from
the box and allowed to jump freely on the floor; the movement was
toward the left, and the frog alighted on the floor on one side, in an
uncertain, floundering way.

In these experiments the responses were no doubt modified by the
irritation caused by the covering, of which each frog tried to rid it-
self. That all the movements were due to this cause cannot be con-
cluded, for in each set, when the right or the left eyes were covered

The Response of the Frog to Light. 475

the orientation was characteristically different, as if resulting from
differing causes, and not merely similar movements caused by the
irritation of the covering.

Effect of prolonged light. — Does exposure to light for a prolonged
period alter the response to light-stimuli?

In order to answer this question two frogs were kept confined in
glass-lined boxes two and one-half inches by one and one-half inches,
the ends of which were covered with wire gauze. Since the frog
turns in a very small space, cords were passed through the boxes
from side to side and one-half inch from the top, forming a sort of
fence, which allowed space for up and down movements of the head,
but not for turning around. One frog was kept in the box from
II A.M. to 4 P.M.; the other from 8.20 A.M. to 4.20 P.M. When
freed and placed in the box with one lighted end, the response was

the same as before, z.¢., the frogs were positively phototactic and
also moved close to the lighted end of the box.

The foregoing experiments seem to indicate two different kinds
of response to light. One kind, the response to the direction of
the rays which affect orientation, is unquestionably phototaxis; the
other I shall not venture to call photopathy in the present unsettled
definition of the term.!

III]. RESPONSE TO WHITE LIGHT IN INCREASED AND IN
LOWERED TEMPERATURES.

The effect of increased temperature. — The temperature of the aqua-
rium in which the frogs were kept varied from 12° to 15° C. The
temperature of the room in which the experiments were performed
varied from about 18° to 20° C. In order to observe the effect of
a rise in temperature on the character of the response to light,
the box, before described, was placed within a large box, also having
a glass end, and into which enough water could be poured to come
well up the sides of the inner box. This water was heated by means
of an Argand burner placed under the larger box. - The temperature
could be increased or kept constant as desired. A marked accelera-
tion in time of response was noted in temperatures up to and in-
cluding 25° C. The frog moved immediately and directly from the
darker end of the box to the lighted end, where it remained close to

1 HoiMES, S. J.: This journal, rgot, iv, p. 211; Horr and Lee: This journal,
1901, ix, p. 460.

476 Ellen Torelle.

the glass. Between 25° and 30° C. the frog became restless and
moved about much. Above 30° C. movements toward the darker
end were as frequent as those toward the lighter end, the response
to light being overcome by the effect of heat.

The effect of lowered temperature.—In observing the effect of
lowered temperature upon the response to light, the same apparatus
was used as described for Experiment I b, z. e. a small tin box, twelve
inches long, nine inches wide, and five inches high, containing a glass
end, within a larger box sixteen inches long, twelve inches wide, and
eight inches high, also furnished with a glass end. The bottom of
the smaller box was covered with a layer of sand one-half inch thick,
and the box was surrounded by ice placed in the larger box. The
temperature of the water in the aquarium in which the frogs were
kept was 15° C., the temperature of the box was 8° C. When a frog
was placed in its rear end, head turned from the light, it moved to
the light end once, remained there for one-half minute, but retreated,
turned away from the light, and remained in the rear of the box,
either moving about, its head down as if it were trying to get under
something, or quietly crouching, with the head down during the other
nine and one half minutes of the experiment. When returned to the
aquarium, the above movements continued in the water, the frog
remaining for several minutes on the floor of the aquarium. Five
frogs were tried; three of these did not leave the rear end of the
box at any time.

Reaction in water to lowered temperatures. —In order to test the
reaction of the frog in water to lowered temperatures, a glass jar
sixteen inches by eight inches in diameter was filled with water, and
set in a box containing ice, so that the lower one-third of the surface
was surrounded by ice; in other respects the jar was left entirely
uncovered.

When the temperature of the water in the jar was 8° C. the frog
was put into it. With swimming movements it went down almost
immediately, head foremost, to the bottom of the jar. With legs out-
spread, almost at right angles to the longitudinal axis of the body, it
moved about on the bottom of the jar, from time to time repeating
the movements described as taking place in Experiment IIb, but
rarely coming to the surface.

From Experiments I, IT, IIIb, one is led to conclude that an increase
of temperature to 30° C. lessens the time of response to light, 7. e.
accelerates the rate; that below 8° C. the frog becomes negatively
phototactic, whether it is in water or on a dry surface.

The Response of the Frog to Light. A77

Stereotropism.— If opportunity be given, does the frog burrow in
sand in temperatures below 8° C., or are the movements observed
stereotropic responses ?

In order to answer this question, sand, to the depth of several
inches, was placed in a tall giass jar, the jar being then filled with
water. The sand was arranged so that its upper surface sloped from
side to side. Twelve frogs were tried. When the temperature of the
water in the jar became 10° C. the frogs went down and remained
down, with the body flat and limbs outspread, but no attempt was
made to burrow. The crouching movements, together with the pass-
ing of the head over the surface of the sand as if exercising a sense of
touch, continued with a lowering of the temperature to 4° C., when
they ceased. When a rock was lowered into the jar in such a way
that a small space was formed between it and the wall of the jar, the
frog crawled into this space and remained there. When a space was
formed between the bottom of the jar and the rock, it crawled into
that. This was tested several times, and was also observed when the
temperature of the water in the aquarium in which the frogs were
kept was lowered to 10° C. and below. When this was done, all the
frogs responded, either by flattening their bodies against the stone
floor, or by creeping under the rocks usually kept there. It there-
fore seems that the frog is stereotropic in temperatures between
to. ©. and 4° C.

Effect of darkness on upward and downward movements in water. —
The same jar used in the preceding experiment was used for this.
The upper two-thirds of the jar including the open surface was
covered with a cloth opaque to the light. With the temperature
of the jar at 10° C. five frogs were tried, each being left in the jar
ten minutes. They went immediately to the bottom, but rose to
the top at intervals as before, and their movements seemed the
same as when the jar was left uncovered. When the lower two-
thirds of the jar was covered, no change was produced.

IV. RESPONSE TO MONOCHROMATIC LIGHT.

The response to monochromatic light was studied in different ways.
First, glass vessels with parallel sides three and one-half inches apart
were used to hold solutions of pigment recommended by Davenport.'

1 Davenport: Experimental morphology, p. 157.

478 Ellen Torelle.

An alcoholic solution of fuchsin was used in testing the response to
red; Lyon’s blue, aconcentrated solution of potassium chromate, and
nickel nitrate were used for blue, yellow, and green respectively.
The response to each was first separately observed.

A glass vessel containing a solution of fuchsin was placed close to
the glass end of the tin box used'in some of the previous experi-
ments. This box was twelve inches long, nine inches wide, and five
inches high, and the inner surfaces were painted a dull black; a slit in
the cover, which was overlaid with several thicknesses of wadding,
made frequent observations easy and caused little, if any, disturbance
to the animal.

Red. — (a) When placed close to the red, the frog turned and
hopped to the rear or opposite end of the box. This happened two
out of seven times. Five times when so placed it turned away from
the red, but remained in the front half of the box.

(6) When placed in the rear of the box it remained there six out of
seven times. The seventh time it wandered about back and forth.

(c) When placed at or near the middle of the box, it was indifferent
as to moving backward or forward. It usually remained about where
it was placed.

Yellow. — The concentrated solution of potassium chromate was
used in the same way that the fuchsin had been used. Five frogs
were tried in ten trials. In each case, the frog moved to the source
of the light, but soon retreated, remaining seated usually a short
distance from it, indifferent as to orientation.

Green. — Four frogs were tried in twelve trials. There was much
moving about, to and from the green, but in no case did the frog
remain for any length of time close to the green light.

Blue. — Three frogs were tried in thirteen trials. The reaction was
immediate and positive. Each frog hopped close to the glass,
usually with the tip of its head against it, and frequently remained
so until removed.

Response to differently colored lights admitted at opposite ends of a
receptacle. — The question arose as to how the frog would respond
were differently colored lights admitted into opposite ends of the
receptacle. In order to answer this, a tin box eighteen inches long,
three inches wide, and three inches high, whose inner surfaces were
painted a dull black with the exception of the opposite ends, which
consisted of glass, was used as a receptacle.

Results.—(a) The vessels before described were placed close to the

The Response of the Frog to Light. 479

glassends. One was filled witha solution of fuchsin, the other with
asolution of nickel nitrate. Five frogs were tried. Each moved from
the red to the green or toward the green.

(6) When the green and the yellow lights were opposed, movement
was from the yellow toward the green. ‘The frogs usually remained
a few inches from the green end of the box with heads turned toward
the green light, but they were not always precisely orientated by
the rays.

Five frogs were tried, each in two trials.

(c) Next yellow and red were used in the same way. Five frogs
were tried. The movement was from red to yellow.

(dz) When the blue light and the red light were at the opposite ends,
the response was an immediate movement toward the blue. The

TABLE I.

Time at blue
end.

Time at red
end.

Position of frog at beginning
of experiment.

8 min.

2 min.

Head turned toward red and in

red end of box.
Same as l.

10 “ce O “ce
10 ““c

Same as l.
ee Same as l.

Same as l.

Same as l.

difference between the response to the blue and the responses to
the green and yellow is very marked. Blue not only effects an
immediate response, but the frog remains close to the glass end where
the blue solution is placed, frequently with its head against the glass
and its median longitudinal axis parallel to the incoming ray.

Unequal amounts of light transmitted through the red and through the
blue media. — Other tests were made with a greater amount of light
transmitted through the red than through the blue medium. A vessel
with parallel sides three and one half inches apart was used for the
blue solution, one with its parallel sides one and one half inches apart
being used for the red. In order to be able to make more accurate
comparisons, it was thought best to note the time during which the
frog in each trial remained with its head directed toward one or the

480 Ellen Torelle.

other light, a ten-minute limit for each trial being taken. Since the
responses to green and to yellow had seemed conclusive, and since
previous observers differed as to the response to red and to blue,
attention was confined to testing the response to these colors.

The results when red and blue were opposed are shown in Table I,
showing a reaction-proportion of 4:2 in favor of the blue, even when
more light was transmitted through the red medium.

Response to a red and to a blue background.— One-half of the inner
surface of a tin box twelve inches long, nine inches wide, and five
inches high was covered transversely with blue, one-half with red
cheese-cloth. White light was admitted through the glass, at the
end, which was covered with blue. The results are shown in

able ll:
TABLE, IL

No. of frog.

Time at blue
end.

Time at red
end.

Position of frog at the beginning
of the experiment.

7 min.

3 min.
Di
13
4

Head turned from light in the
rear of red compartment.
Head turned from light in the

middle of red compartment.
Head turned from light in the

rear of red compartment.
Same as 3,

Same as 3.

Then the strips were reversed, white light being admitted through
the glass at the end, which was covered with red. The results are
embodied in Table III.

Time at blue
end.

TABLE III.

Time at red
end.

Position of frog at beginning
of the experiment.

10 min.
10 ce
8 “

O min.

Head turned from light and in
rear of blue compartment.

Same as l.

Same as l.

Same as l.

In the middle of the red area, head
turned toward the light.

The Response of the Frog to Light. 481

The box was lined with the strips laid lengthwise, the white light
admitted equally on both; the position of the frog at the beginning
of each trial, together with the results, being indicated in Table IV.

PAGES IE Ere

Time at blue Time at red Position of frog at beginning
end. end. of experiment.

9} min. > min. In rear enc of red compartment,

head turned from light.
Ora haa | Same as l.

6 : : Same as I.
9 : ce In rear of blue compartment, head

turned from light.
IQ) eS Da Same as 4.

Red and blue solutions were placed in front of the red and blue-
lined sides of the box respectively, with results shown in Table V.

LABCE Ne

Time at blue Time at red | Position of frog when
end. end. experiment began,

10 min. QO min. | In rear of red compartment, head

turned toward the light.

5 ie S) aes | In rear of red compartment, head
turned from the light.

lines 3 | In middle of blue compartment,
head at right angles to incom-
ing light.
OS In rear of blue compartment, head
, toward light.
ORS | In rear of blue compartment, head
turned from light.

The results of the experiments with the colored cloths did not
seem convincing, for the cloth offered an absorptive surface to the
skin of the frog, and the dyes as well might have vitiating influences.
So the external surfaces of three rectangular glass vessels were
painted in the one case blue and red, transversely applied; in the
other, the position of the colors was reversed; in the third, the
paint was longitudinally applied.

The blue corresponded to the tube paint known as new blue; the
red, to vermilion. White light was transmitted at the open end
through an ordinary window-pane.

482 Lillen Torelle.

(a) The vessel on which the paints were longitudinally applied was

first used.
TABLE VI.

Time at blue Time at red Position of frog at beginning of
end. end. experiment.

10 min. O min. In rear of red compartment.
ce

In rear of red compartment.

* In rear of blue compartment.

“

In middle of blue compartment.

In middle of red compartment.

In rear of blue compartment.

In middle of red compartment.

In rear of red compartment.

In rear of blue compartment.

In ant. end of red compartment.
ant. end of red compartment.
rear end of red compartment.

rear end of blue compartment.

The above table (Table VI) shows a response of eleven out of thir-
teen to blue. The response to red may be accounted for by fear,
or by sluggishness.

(6) Next the vessels to which the colors had been transversely
applied were used. The frog was first placed in the vessel in which
the red was next to the white light. It was watched for ten minutes ;
then this same frog was put into the aquarium in which the blue was
next to the white light and again watched ten minutes. The frog
was always placed in the same relative positions in the two vessels
and the positions differed for each experiment.

When red was next to the white light the responses were those
indicated in Table VII: |

The Response of the Frog to Light. 483

ABER WLI

Time at blue | Timeatred | Position of frog when
end. end. experiment began.

10 min. In right-hand, rear corner of blue.
(Moved to red boundary and
stopped.)

10 In anterior part of blue compart-
ment. (Moved to red boun-
dary and stopped.) Remained
in blue compartment.

In middle of red compartment.

In rear of middle part of blue
compartment.

In rear of left-hand corner of blue
compartment.

In middle of red compartment.

When blue was next to the white light the frogs responded as
shown in Table VIII:

TAB EES VUE

Position of frog at

Time on blue. Time on red. Wh scotia 5;
beginning of experiment.

O min. | 10 min. In middle of red compartment.
In middle of blue compartment.
In middle of blue compartment.

In rear of red compartment.

In rear of red compartment.

In middle of blue compartment.

A constancy in the response of the same frog, in vessels in which
the colors are reversed, is here observed, except in frogs 1 and 5
of the last set. The first I can account for, as the glass plate at
the open end fell during the experiment and frightened the frog
into retreat; the case of the fifth I cannot account for except that
its condition was rather sluggish.

A week later the same experiments were repeated with the follow-
ing results:

When red is next to the white light the following results were
obtained:

484 Ellen Torelle.

TAB IE Exe

Position of frog at beginning
of experiment.

Time on blue. Time on red.

10 min. Oiminy In rear of blue compartment.

In rear of blue compartment.

In middle of red compartment.

When blue is next to the white light the results are those indicated

in Table X:
TAR ILE Ss

Position of frog at beginning

Time on blue. Time on red. =
of experiment.

2 min. In rear of red compartment.
In rear of red compartment.

In middle of blue compartment.

The experiments were repeated, using other frogs, and changing
the time of each trial from ten to twenty minutes. When red is next
to the source of light the behavior of the frog is that indicated in

Gable xc
ANB Io Xcle

ena ‘ sit a
Position of frog at beginning

‘Time on blue. Time on red. :
of experiment.

20 mins. | 0 mins. In middle-left of red compartment.

20 | In middle-left of red compartment.

Ze ; In middle-left of red compartment.

When blue was next to the source of light the frogs responded as
indicated in Table XII:

The Response of the Frog to Light. 485

TABLE XII.

Position of frog at beginning
of experiment.

Time on blue. Time on red.

4 min. In middle of blue compartment.

Tie“ In middle-left of blue compartment.

a « In middle of blue compartment.

Calculating in minutes the response to a red or to a blue back-
ground, when white light is admitted, the response to the blue was
three times greater than that to the red, the actual number of minutes
on the blue being 479; on the red, 159.

Response when the entire environment is one-half blue and one-half
red. — A glass plate, one-half of which was painted red, the other
one-half blue, was placed before the opening of the vessel! so that
the red of the plate was adjacent to the red of the vessel, and the
blue of the plate adjacent to the blue of the aquarium. The trials
were of twenty minutes’ duration and gave the following results:

TABLE XIII.

Position of frog when placed

Time on blue. Time on red. =
in the vessel.

16 min. In rear of red compartment.

20s In rear of blue compartment.

In middle of red compartment.

In middle of blue compartment.

In front end of blue compartment.

In front end of red compartment.

In this set of experiments the reaction-proportion is as 14:1 in
favor of the blue.

Response when white light is admitted at opposite ends of a re-
ceptacle, one-half of the surface of which is painted red, the other
half blue. — A tin box eighteen inches long, three inches wide, and

1 In which the strips of red and blue ran lengthwise.

A486 Ellen Torelle.

three inches high, and containing opposite glass ends (three inches
by three inches), was laid off into equal compartments and its inner
walls painted blue and red. In six trials the same frog was used,
being put in the same relative position but upon a different color in
consecutive experiments. In eleven trials the colors were reversed
after each experiment, the frog being placed now on red, now on
blue, and again on the boundary between the two. The results of
the eleven trials are shown in the following table:

TABLE XIV.

Position of frog at

Time on blue. | Time on red. aus >,
beginning of experiment.

10 min. 0 min. In rear of red compartment.

10 0 In front end of blue compartment.
10 . 0 In middle of blue compartment.
In rear of blue compartment.

In front end of red compartment.
In middle of red compartment.
In rear of red compartment.

In front end of red compartment.
In middle of red compartment.
In rear of red compartment.

In rear of blue compartment.

The results in the former case show a greater length of time on
the blue in the case of all except the second frog, which, as indicated
in Table XV, remained in the red compartment for ten minutes when
placed in the red compartment.

The results obtained in response to monochromatic light seem to
illustrate Loeb’s theory “that the more refrangible rays are extraor-
dinarily more active than the less refrangible, which occasionally
remain almost ineffective.” ! According to Abelsdorff,? red rays

affect the pupils of the eyes of some animals like darkness.

1 Logs, J.: Der Heliotropismus der Thiere, p. 20.
2 ABELSDORFF, G.: Archiv fiir Physiologie, 1900, p. 561.

The Response of the Frog to Light. 487

Graber’s result on the frog can be explained only on the ground of
a confusion arising as a result of using so many frogs (forty) at the
same time in one receptacle.

TABLE XV.

No. of | Time on Time on Position of frog at the begin-
trials. blue. red. ning of the experiment.

10 min. 0 min. At glass end of blue compartment.

10 Oe At glass end of blue compartment.

ila At glass end of red compartment.
= At glass end of red compartment.

At glass end of blue compartment.

At glass end of blue compartment.

CONCLUSIONS.

1. At the usual temperature of the laboratory, between 16° and
21° C., Rana virescens virescens and R. clamata are positively
phototactic.

2. They respond to light coming from above and from below, as
well as from the side.

3. They respond differently to different intensities of light. They
move out of the sunlight into the shadow, even when by so doing the
movement is away from, or at right angles to, the direction of the
ray.
4. When one eye is covered, the body is placed with its median
plane oblique to the ray.

5. When a bank of sand is interposed between the frogs and the
light, they crawl over this and move to the source of light.

6. Arise in the temperature to 30° C. accelerates the rate of the
positive response. A lowering of the temperature to 10° C. produces
movements away from the light.

7. When placed in water the temperature of which is lowered to
10° C., the frogs swim downward: (@) In an uncovered glass ves-
sel; (0) Ina vessel, the upper two-thirds of which has been dark-
ened; (c) In a vessel the lower two-thirds of which has been
darkened.

488 Ellen Torelle.

8. The frogs turn away from red light and move toward blue light.
They move toward green and toward yellow light, but are not defi-
nitely orientated by either.

g. When red light is admitted at one end and green light at the
other end of a receptacle, the frogs move from the red to or toward
the green. When red and yellow lights are opposed in the same
way, movement is from the red to the yellow. When red and blue
are opposed, movement is immediately toward the blue.

10. When white light is admitted at one end of a receptacle, and
the frogs are given a choice of a red or of a blue environment, they
move, in most cases, into the blue, and remain in it longer than they
do in the red.

11. When one-half of the entire receptacle is blue and the other
half is red (no white light being admitted), movement is from the red
to the blue.

ENDEX TO VOL:

2, ae J. J. On the true elementary

composition of purified adrenalin

and the relation of the substance to
epinephrin, xvii.

ApAms, G. P. On the negative and posi-
tive phototropism of the earthworm Allo-
lobophora foetida (Sav.) as determined
by light of different intensities, 26.

Adrenalin, composition, xvii.

Alcohol, excretion of uric acid, xi.

Agitation, effect on Arbacia eggs, 245, xviil.

Apnea, 24.

Arbacia eggs atfected by agitation, 245.

Auditory affected by visual stimuli, 116.

ACTERIA, chemistry, xviii.

BEEBE, S. P. The effect of alcohol
and alcoholic fluids upon the excretion of
uric acid in man, xi.

Blood, coagulation affected by adrenalin,
35:

, coagulation, influenced by formal-
dehyde, etc., 187.

——,, electrical potential, 262.

, Sugar content affected by adrenalin,
35:

Blood-pressure in veins, 161, 198.

Brain, irritability during anzmia, 131.

Brown, O. H. The immunity of Fundulus
eggs and embryos to electrical stimula-
tion, IIT.

(Ce chloride, physiological action,
214.

CaNNON, W. B., and H. F. Day.
digestion in the stomach, 396.

Cell substance, method of obtaining, xviii.

Cervical sympathetic, effect of section com-
pared with removal of superior ganglion,
Xviil. :

Chinic acid, influence on elimination of
uric acid, xvi.

Chlorides in urine, affected by diuretics,
etc , 425.

CLowes, G. H. A. The relationship
between the freezing point depression

Salivary

IX.

and specific gravity of urine, under vary-
ing conditions of metabolism, and its
clinical value in the estimation of sugar
and albumin, 319.

Color-sense, frog, 466.

CusHNy, A. On the pharmacological
action of optical isomers, xiv.

ACY. EY
396.
DEAN, A. L. See HENDERSONand DEAN,

386.

Dextrose from pancreatic digest, 380.

Diabetes, phlorhizin, respiration experi-
ments, Xvlil.

Diuretics, effect on chlorides of urine, 425,
xii.

——,, saline solutions, 454.

DoNnaLpson, H. H. Dr. Hatai’s observa-
tions on the effect of lecithin on the
growth of the nervous system of the
white rat, xviii.

See CANNON and Day,

DWARDS, G. H. See HENDERSON
and EDWARDS, 417.
Epinephrin, xvii.
Eye movements, relation to visual stimuli,

122.

petscBes®. M. H. Artificial partheno-
genesis in Nereis, [00.
FouLin, O. The acidity of urine, 265.
Fo.Lin, O. On rigor mortis, 374.
Formaldehyde, action on blood, 187.
Fundulus eggs and embryos immune to
electrical stimulation, IIT.

EOTROPISM of Paramecia, 238.
G1sson, R. B. Observations on the
urine of the muskrat (Fiber zibethicus),
Boer
Gigs, W. J. Peptic proteolysis in acid
solutions of equal conductivity, xvii.
Girs, W. J. On the irritability of the
brain during anemia, 131.

489

490 | Lndex.

Gigs, W. J., and S. J. MELTZER. Studies
on the influence of artificial respiration
upon strychnine spasms and respiratory
movements, I.

Giss, W. J. See TALTAVALL and GIEs,
XV.

Glycogen, formation from proteids, 138.

GuTuRrig£, C. C. The influence of for-
maldehyde on the action of certain
laking agents and on coagulation of
blood, 187.

H AEMOLYSIS, 187.
——, cold) 72.

HANForD, G. A. A study of the physio-
logical action and toxicology of caesium
chloride, 214.

HENDERSON, Y., and A. L. DEAN. On the
question of proteid synthesis in the ani-
mal body, 386,

HENDERSON, Y., and G. H. Epwarbs.
Nuclein metabolism in lymphatic leu-
keemia, 417.

MMUNITY of Fundulus eggs and em-
bryos to electrical stimulation, III.

kK IDNEY, perfusion, 460.

ECITHIN, influences growth of ner-
vous system, xviii.
Leukaemia, metabolism, 417.
LEVENE, P. A. On nucleic acid, xvii.
Light, response of frog, 466.
Lusk, G.,and A.R. MANDEL. Respiration
experiments in phlorhizin diabetes, xviii.
Lusk, G. See STILEs and Lusk, 380.
Lymph-flow, secretin, xv.
Lyon, E. P. Experiments in artificial
parthenogenesis, 308.

ACDOUGALL, R. On the influence
of varying intensities and qualities
of visual stimulation upon the rapidity of
reactions to auditory stimuli, 116.
MacDouGa.1, R. On the relation of eye
movements to limiting visual stimuli, 122.
MANDEL, A. R. See Lusk and MANDEL,
XViii.
Medusa, reactions, 279.
MEETZER, CLARA. See MELTZER and
MELTZER, 57, 147, 252, xViii.
MELTZER, S.J. Some observations on the
effects of agitation upon Arbacia eggs,
245, XViil.

MELTZER, S. J.,and CLARA MELTZER. On
a difference in the effect between the
simple cutting of the cervical sympathetic
nerve and the removal of the superior
ganglion, xviii.

MELTZER, S. J., and CLARA MELTZER. A
study of the vasomotor nerves of the
rabbit’s ear contained in the third cervi-
cal and in the cervical sympathetic
nerves, 57.

MELTZER, S. J., and CLARA. MELTZER.
The share of the central vasomotor in-
nervation in the vasoconstriction caused
by intravenous injection of suprarenal
extract, 147.

MELTZER, S. J.,and CLARA MELTZER. On
the effects of subcutaneous injection of
the extract of the suprarenal capsule
upon the blood-vessels of the rabbit’s
Ear, 252:

MELTZER, S.J. See Gres and MELTZER, I.

MENDEL, L. B., with H.C. THACHER. On
secretin and lymph-flow, xv.

Metabolism, dextrose from pancreatic
digest, 380, xviii.

——,, glycogen, 138.

——,, nuclein, 417.

——,, proteid synthesis, 386.

Moore, A. Some facts concerning geo- °

tropic gatherings of Paramecia, 238.
Muscular contraction, effect on circulation,
161.

N ERVOUS system, growth influenced
by lecithin, xviii.
Nucleic acid, xvii.
Nucleic acid of wheat embryo, specific
rotation, 69.
Nuclein metabolism in leukzmia, 417.
Nucleo-proteid, polarization, 69. ~

PITZ, R. BurRTON-. Muscular contrac-
tion and the venous blood-flow, 16r.
Opirz, R. Burron-. Venous pressures, 198.
Optical isomers, pharmacological action,
xiv.
OsBorNE, T. B. The specific rotation of
the nucleic acid of the wheat embryo, 69.
Oxyhemoglobin, crystallization, 97, xvili.

ARAMECIA, geotropism, 238.
Parthenogenesis, artificial, 100, 308.
Phototropism, 26.
Proceedings of the American Physiological
Society, ix.
Proteid synthesis, 386.

Lndex.

Proteids, a source of glycogen, 138.

Proteolysis, peptic, in acid solutions of
equal conductivity, xvii.

Proteoses, physiological action, 345.

EICHERT, E. T. New crystalline
forms of oxyhzmoglobin artificially
produced, xviii.

REICHERT, E. T. Quick methods for
crystallizing oxyhemoglobin: inhibitory
and accelerator phenomena, etc. : changes
in the form of crystallization, 97.

Respiration, artificial, effect on strychnine
spasms, I.

Respiration in phlorhizin diabetes, xviii.

Respiratory variations in venous pressure,
198.

RicHArpDs, A. N.
RICHARDS, 35.

Rigor mortis, 374.

See VOSBURGH and

Satine solutions, diuretic effect, 454,

xiil.

Salivary digestion in stomach, 396.

Secretin and lymph-flow, xv.

SOLLMANN, JT. The effect of saline injec-
tions, diuretics, and nephritic poisons on
the chloride-content of the urine in the
dog, xii.

SOLLMANN, T. The cause of the greater
diuretic action of hyperisotonic salt-
solutions, xiii.

SOLLMANN, T. The effect of diuretics,
nephritic poisons, and other agencies on
the chlorides of the urine, 425.

SOLLMANN, T. The comparative diuretic
effect of saline solutions, 454.

STEWART, G. N. The influence of cold on
the action of some hamolytic agents, 72.

SSTEWARY, G. N. Differences of potential
between blood and serum and between
normal and Jaked blood, 262.

STILES, P. G.,andG Lusk. On the forma-
tion of dextrose in metabolism from the
end-products of a pancreatic digest of
meat, 380, Xviill.

Stomach, salivary digestion in, 396.

STOOKEY, L. B. On the formation of
glycogen from glycoproteids and other
proteids, 138.

Srorey, T. A. The immediate influence
of exercise upon the irritability of human
voluntary muscle, 52.

491

Strontium, xviii.

Strychnine spasms affected by artificial
respiration, I.

Suprarenal extract, action on blood-vessels,
147, 252.

Sympathetic, effect of removal of superior
ganglion compared with section of cervi-
cal sympathetic, xviii.

ALTAVALL, W. A., and W. J. GIEs.
The influence of chinic acid on the
elimination of uric acid, xvi.
THACHER, H. C. On the excretion of
strontium, xviii.
THACHER, H. C. See
THACHER, XV.
TORELLE, E. The response of the frog to
light, 466.

NDERHILL, F. P. New experi-
ments on the physiological action of
the proteoses, 345.
Uric acid, elimination influenced by chinic
acid, xvi.
Urine, acidity, 265.
—, chlorides affected by diuretics, poi-
sons, etc., 425.
——, depression of freezing point and spe-
cific gravity related to metabolism, 319.
, estimation of sugar and albumin, 319.
Urine of muskrat, 391.

MENDEL and

\/ SsOMOTaR nerves in rabbit’s ear, 57.

. Vasomotors, affected by suprarenal
extract, 147.

VAUGHAN, V.C. The chemistry of bacte-
rial cells, with a demonstration of the
apparatus used in obtaining the cellular
substance in large amount, xviii.

Venous blood-flow during muscular con-
traction, 161.

Venous pressure, 198.

Visual stimuli affect auditory stimuli, 116.

Visual stimuli affect eye movements, 122.

VosBuRGH, C. H., and A. N. RICHARDS.
An experimental study of the sugar con-
tent and extravascular coagulation of the
blood after administration of adrenalin,

35:

/ BEEES: R. M. A study of the re-

actions and reaction time of the
medusa Gonionema Murbachii to photic
stimuli, 297.

Ae

atau

~
yt

~

ims

fac

sicy ao SE

e
Ors Maes

> cig ges Ee LS es

IBINDING SECT. MAR 4 4966

QP American Journal of Physiology

(404

Biological
& Medica]
Serials

| PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

Se eee ee
UNIVERSITY OF TORONTO LIBRARY

‘ o “1
* 1 .
tobe “
oa f ‘
J ’ AN
we ‘ 5
» i
Ni a, ae ae ey ‘ 1
. “4 Z ve) y
wens i : ‘ «4
n i ? ‘ i
. ’
oe ee oan
‘ F j 7
i *
x & ve_-? ‘
,
i TD er =a Ly Ne. ;
- i : P
fp , ve - S. t
4 re = 1
4
* es o '
esr F , ' /
L¢ - a 7 . .
' 2% + *
' va,
' 7 x ;
- J - .
of ‘ vans mer
’ - 7
. F) : _ . r
= 7 ' ; t ;
at os Me +” " .
Ne earners i Re We ¢
Gaul A i , 7 .
a ob : f . : ‘

va hee
watts

sit
Seti ae
«

is’ a’, ity
‘ *
al Natty

he fo at tnter ta aha iecdin tae cee ion as co ae

ae rhe

Po ree

a a er ee ene

TY Ripslts, 2?
dee ie
i

fa
ie
3He

oe OS oe 0 re oe 2 nS

a

<i.

ae

Bg