Oe hw eS ne a, 
ee. 
sn 


sae 
ae. 
See an ree 
‘ aan a : ~ 


ee 
tae eee 


<0 belies ene ene 
~ Sete Se at 


ers 
Via nen Baas 


meee ee 
ne 


2 ret 9 we See te 
— 


6 ree ee eel 
2 ee 


4 pe 
7 petataata at on one = 
<a ss. 7 : 
Nem Se te : 
ns 


Ae Daw 
PGS equ 


THE AMERICAN JOURNAL 


OF 


PHYSIOL OGY. 


EDITED FOR 


The American Physiological Society 


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


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


THE 


AMERICAN JOURNAL 


PHYSIOLOGY 


VoLumME VIII. 


i 


Ve 
0 


%S 
Ary l 
&, Ae 
BOSTON: US. A, An 
GINN AND COMPANY 
1903 


raliv 


oe 


a 


‘ rx 
 OMliby 


way pee 

. 7 y 
, a ra | 
Oy 

“ ‘ 

Px 

¥ 

J 


dosti lM 


 Yy one - Wh Ae . We ; 
3 | 7A) AU Pah) hee 
are 7 : - s i i | : 5 , § 


s A be ‘ . ts aha, 4 


Copyright, 1903 
By GINN AND COMPANY 


> 


Gnidversity Press 


Joun WILSON AND Son, CamBrinGe, U.S.A. 


Coo NOE ENTS. 


No. I, OcroBer 1, 1902. 


ON THE FUNCTIONS OF THE CEREBRUM: I.— THE FRONTAL LOBES IN 
RELATION TO THE PRODUCTION AND RETENTION OF SIMPLE SENSORY- 
Motor Hasits. By Shepherd Ivory Franz 

STUDIES ON REACTIONS TO STIMULI IN UNICELLULAR ORGANISMS. 
IX.—ON THE BEHAVIOR OF FIXED INFUSORIA (STENTOR AND 
VORTICELLA), WITH SPECIAL REFERENCE TO THE MODIFIABILITY OF 
PROTOZOAN REACTIONS. By H.S. Jennings. . . ; 


THE ACTION OF ALCOHOL ON Muscle. By Frederic S. Lee and William 
Salant : 


No. I], NovEMBER I, 1902. 


THE IMPORTANCE OF SODIUM CHLORIDE IN HearT Activity. By 
David J. Lingle 

NOTES ON THE ACTION OF ACIDS AND ACID SALTS ON BLOOD-CORPUSCLES 
AND SOME OTHER CELLS. By S. Peskind . 

THE BEHAVIOR OF NUCLEATED COLORED BLOOD-CORPUSCLES TO CER- 
TAIN H&MOLyTIC AGENTS. By G. NV. Stewart 


No. III, DECEMBER I, 1902. 


THE EFFECT OF DIMINISHED EXCRETION OF SODIUM CHLORIDE ON 
THE CONSTITUENTS OF THE URINE. By R. A. Hatcher and Torald 
Sollmann 


THE MECHANISM OF THE RETENTION OF CHLORIDES: A CONTRIBUTION 
TO THE THEORY OF URINE SECRETION. By Torald Sollmann. 


PAGE 


61 


99 


103 


vl Contents. 


PAGE 
Does PoTAssIuUM CYANIDE PROLONG THE LIFE OF THE UNFERTILIZED 
EGG OF THE SEA-URCHIN? By F. P. Gorham and R. W. Tower. ~ 175 


NOTES ON THE “ PROTAGON” OF THE BRAIN. By W. W. Lesem and 


Witham JiGlesa. toa ee - os a ee 
ON THE GROWTH OF SUCKLING PIGS FED ON A DIET OF SKIMMED 
Cow’s Mitx. By Margaret B. Wilson. . - . « + + + «+ + « IQ7 


MAXIMAL CONTRACTION, “STAIRCASE” CONTRACTION, REFRACTORY 
PERIOD, AND COMPENSATORY PAUSE, OF THE HEART. Ay Rk. S. 
Woodworth <6 8 a a ee! ee te) 


No. IV, JANUARY 1, 1903. 


THE RATE OF NERVOUS IMPULSE IN CERTAIN Mo.uuscs. Sy O. P. 


Jenkins ‘and Ac J. Carlson 3 ie We a) ee ee = 
ON THE INFLUENCE OF CALCIUM AND POTASSIUM SALTS UPON THE 
TONE OF PLAIN MUSCLE. By Percy G. Stiles .-. « %. « =) =e 


ON DIFFERENCES IN THE DIRECTION OF THE ELECTRICAL CONVEC- 
TION OF CERTAIN FREE CELLS AND NvucLeEl. By Ralph S. Lillie 273 

AN EXPERIMENTAL STUDY OF THE CHEMICAL PRODUCTS OF BACILLUS 
Cott COMMUNIS AND BaAcILLUS LACTIS AEROGENES. By Leo F. 


UR Go ate Mb Rerun BP ees 2 > on 
ELECTRICAL POLARITY IN THE Hyproips. Sy Albert P. Mathews. . 294 
THE IMPORTANCE OF MECHANICAL SHOCK IN PROTOPLASMIC ACTIV- 

ity. By A. P. Mathews and BOR. Whitcher . ~. . . » « = ee 
FivE TyPES OF EYE MOVEMENT IN THE HORIZONTAL MERIDIAN PLANE 

OF THE FIELD OF REGARD. By Raymond Dodge... . . - + 307 


ON THE QUANTITATIVE DETERMINATION OF AMMONIA IN URINE. Sy 
Philip. Shaffer -<. 2 sivieg RAD are ey Pea heey a ae ce 


THE INFLUENCE OF FATIGUE UPON THE SPEED OF VOLUNTARY CON- 
TRACTION OF HUMAN Muscle. By Thomas Andrew Storey. . +- 355 


No. V, FEBRUARY 2, 1903. 


ae 


A PHYSIOLOGICAL StuDY OF NucvLeic Acip. By Lafayette B. Mendel, 


Frank P. Underhill, and Benjamin White . . . « « » » «+ « = 909 
THE ACTION OF ACIDS AND AciID SALTS ON BLOOD-CORPUSCLES AND 

OTHER Cetus. By S. Peshind . ss =. © os) 5 1) 
HOW LONG DOES (ARBACIA) SPERM LIVE IN SEA-WATER? By Martin 


HT, Fischer . WO 


VARIATIONS IN THE AMPLITUDE OF THE CONTRACTIONS OF HUMAN 
VOLUNTARY MUSCLE IN RESPONSE TO GRADED VARIATIONS IN THE 
STRENGTH OF THE INDUCED SnHock. By Thomas Andrew Storey . 435 


Contents. Vii 


° PAGE 
THE LAKING OF DRIED RED BLoop-CorpuscLes.. By Charles Claude 
POE en Bie Ee A ee ee ir ie eee er aiee 8 


ON THE NUCLEOPROTEIDS OF THE PANCREAS, THYMUS, AND SUPRARENAL 
GLAND, WITH ESPECIAL REFERENCE TO THEIR OPTICAL ACTIVITY. 
By Arthur Gamgee and Walter Jones. . - - «© + + + + + + + 447 


PROCEEDINGS OF THE AMERICAN PHYSIOLOGICAL SOCIETY . . .- . ix-xliii 


I O58 ee Pat ak oo Os cee ee ck 8. a ABZ 


PROCEEDINGS OF THE AMERICAN PHYSIO- 
Peete -SOCTET Y. 


FIFTEENTH ANNUAL MEETING. 


WASHINGTON, D.C., DECEMBER 30 and 31, Igo2. 


PROCEEDINGS OF THE AMERICAN PHYSIOLOGICAL 
SOCIETY. 


ON GLUCOTHIONIC ACID. 


By PR. A. LEVENE, 


In the preparation of nucleic acid by the picric acid-alcohol method, 
a carbohydrate is precipitated with the acid. The carbohydrate can 
be removed by copper chloride. In the yeast, in bacteria, pancreas, 
and liver the carbohydrate resembled glycogen. In the spleen, it 
had properties similar to those of chondroitin sulphuric acid. It 
contained sulphuric acid in organic combination, and it gave the 
barium test for glycuronic acid. Unlike the chondroitin sulphuric 
acid, its sulphur was equal to 3 per cent and its nitrogen to 5.41 
per cent. Very characteristic of the substance is its behavior toward 
orcin-hydrochloric acid. Chondroitin sulphuric acid, as well as 
pentose, on heating with orcin-hydrochloric acid gave a purple color, 
soon turning into green on the separation of dark green floccules. 
The substance of the spleen forms with the same reagents a bright 
purple color which remains unchanged for days; the floccules have 
the same bright purple color. It is possible that the substance is not 
unlike the glucosaminic acid described by Fischer, combined with 
sulphuric acid. 


ON GLUCOPHOSPHORIC ACID. 
By P. A. LEVENE. 


Tue author analyzed in various seeds the phosphorus-containing 
substance first obtained by Paladin and analyzed by Schulze and 
Winterstein. These authors failed to establish the nature of the 
organic radical of the substance. The author found that about 30 
per cent of the organic part could be split off in form of a carbohy- 


Xl Fifteenth Annual Meeting. 


drate. The carbohydrate gave all the qualitative tests for a pentose. 
The phenylhydrazin compound and the bromphenylhydrazin com- 
pound possessed all the properties of the pentose derivatives, yet 
they were not pure enough to establish their identity. The 
substance contained about 15 per cent of organic phosphorus, 
1.8 per cent of nitrogen, and 50 per cent of ash, which consisted 
chiefly of calcium-magnesium-phosphate. The substance contained 
no glycerine or purin bases. 


ON NUCLEIC ACID. 


BYe IP. CAG UE Vie Nib 


Tue author continued his analysis of some nucleic acids. Attention 
was directed toward the presence or absence in the spleen and 
pancreas of the following substances: glycerine, a carbohydrate, 
pyrimidin derivatives. Glycerine could not be found in any percep- 
tible quantity. A carbohydrate could not be isolated, but five 
different nucleic acids, namely, those of the pancreas, spleen, liver, 
yeast, and tubercle bacilli gave all the furfurol tests on distillation 
with mineral acids. Thymin could be isolated from the nucleic acids 
of both pancreas and spleen. From the same two acids there was 
obtained, in the form of a picrate, a substance which could be trans- 
formed into a disulphate. The composition of the two salts was 
as follows : 
For the picrate, calculated as: (C,xH,N,0)C,H,(NO,),OH, 


Calculated. Found. 
C5) 5 ey OG; 35:56%.- 
es ese eet 2.91%. 
Ng, 2s O2ROSS, yaa! Wt Ey AR 


For the sulphate, calculated: 


Calculated. Found. 
( 29.90% 29.44%. 
H 4.34% 4.01%. 
Ne. ie. ee eGalOnn 26.25%. 
S 9.62% 9.84%. 


In its elementary composition the substance is not unlike cry- 
tosin, as described by Kossel. The formula for episarkin, however 
seems to correspond better with the analytical data. It is, perhaps, 
an amino-oxy-pyrimidin. 


Proceedings of the American Physiological Society. xiii 


FURTHER MUCOID STUDIES. 


By WILLIAM J. GIES. 


I, INVESTIGATIONS into the distribution of osseomucoid indicate 
that glucoproteid is a normal constituent of all bones. It has thus 
far been found in the large bones of wild and domestic mammals 
and birds, and of reptiles. 

II. Connective tissue mucoid shows a tendency to combine with 
other proteids. Thus, for example, an alkaline solution of potassio- 
mucoid and gelatine yields a precipitate with acid more promptly than 
a solution of the equivalent amount of the mucoid salt alone. Fur- 
thermore, the compound precipitate is different physically. In the 
case of the gelatine product the precipitate possesses semi-gelatinous 
qualities. The compound precipitates of mucoid obtained from pro- 
teid solutions weigh more than the control mucoid precipitates. This 
added weight rises, within certain limits, as the proportion of asso- 
ciated proteid in the solution increases. 

III. Acidification of tissue extracts is not sufficient for complete 
precipitation of the mucoid. Even with a fifth alkaline extract of the 
same tendon pieces, the water-clear acid filtrate from the precipitated 
mucoid contains additional glucoproteid. 

IV. Precipitated mucoid shows practically no combining power 
with acids. In the hydration of mucoid by pepsin-acid, however, acid 
combines with the dissolved proteid products formed in the process. 

V. The blood serum of a rabbit, which had been treated with 
several subcutaneous and intraperitoneal injections of neutral solu- 
tion of potassio-mucoid, produced precipitates in neutral and slightly 
acid solutions of the latter proteid substance. 

These researches are still in progress with the co-operation of 
Messrs. E. R. Posner, C. Seifert, and H. G. Baumgard. 


ON THE ORIGIN OF GLYCURONIC ACID. 


By J. A. MANDEL anp H. C. JACKSON. 


MAYER (Verhandlungen des Congresses fiir innere Medicin, 1901, p. 
393), inaseries of inadequate experiments on rabbits, which were fed 
camphor before and after fasting, claims a direct formation of glycu- 
ronic acid from the dextrose of the body. He also considers that 


XIV Fifteenth Annual Meeting. 


elycuronic acid is a normal constituent of the blood. Other investi- 
gators also give this origin for the glycuronic acid which combines 
with foreign substances when introduced into the organism, and 
nearly all have ignored the possibility of its origin from the proteids. 

We fed camphor to fasting dogs for several days, and estimated the 
campho-glycuronic acid and nitrogen excreted, and then fed large 
quantities of dextrose (80 grams) for several days, and observed a 
marked diminution in the elimination of both campho-glycuronic acid 
and nitrogen. On giving the animal chopped meat, the quantities of 
campho-glycuronic acid and nitrogen were correspondingly increased, 
and a constant elimination was obtained as long as the meat and 
camphor were fed, showing a positive proteid origin for the 
glycuronic acid. 

In a series of experiments, making use of two different methods 
for removing proteids, we have failed to detect glycuronic acid in the 
blood. In order to ascertain the seat of formation of the conjugated 
glycuronates, we made several transfusion experiments by passing 
the blood of one dog, which had been previously fed with large doses 
of camphor, through the kidneys of a second dog and found that the 
urine collected during transfusion was lzvo-rotatory and contained 
campho-glycuronic acid. Incidentally we found that the kidneys of 
all our camphor-fed dogs showed, on microscopical examination, a 
pronounced fatty degeneration in the ascending loops of Henle and 
not in the convoluted tubules. It is not improbable that these cells 
are concerned in the synthesis of the glycuronates. 


FURTHER STUDIES OF THE TOXIC AND. ANTITOXIC 
EFFECTS OF IONS. 


By JACQUES LOEB anp WILLIAM J. GIES. 


Tus research was conducted at Wood’s Holl during the past summer. 
It confirmed Loeb’s original observation that each electrolyte in solu- 
tion at a certain concentration is able to prevent the development of 
the Fundulus egg after fertilization, and also to destroy the egg. 
Our experiments further confirmed the fact that this poisonous action 
can, in general, be wholly or partly inhibited by the addition of a 
proper amount of another electrolyte. 

We also obtained results emphasizing the fact first observed by 


Proceedings of the American Physiological Society. xv 


Loeb, and furnishing new evidence to show, that the degree of antitoxic 
influence exerted by the second electrolyte increases with the valency 
of the cation. The antitoxic action of bivalent cations was found to 
be very much greater than that of univalent cations, the antagonistic 
power of trivalent cations is considerably greater than that of the 
bivalent. This rule does not hold with all cations, however; such 
cations as Cu, Hg, and Cd are exceptions. 

Our experiments made it very apparent that the antitoxic action of 
the salts employed was not due to slight amounts of H or OH ions 
in their dissociated solutions, since neither solutions of pure acids 
nor of pure alkalies were able to exert such antagonism. 

It was found, finally, that solutions of non-electrolytes, ¢. g., urea, 
cane-sugar, glycerine, alcohol, have no antitoxic influence except 
under conditions which favor the formation of not easily soluble dis- 
sociable compounds with the electrolyte (such as saccharate), where- 
by the concentration of the toxic ion is considerably reduced. 

Koch’s recent investigations on the influence of ions on lecithin 
solutions emphasize the possibility previously suggested by Loeb, 
that the observed antagonistic effects of ions may be referred, in part 
at least, to changes induced in the physical and perhaps chemical 
conditions of substances such as lecithin in the cell. 


A PROTEID REACTION INVOLVING THE USE OF 
CHROMATE. 


By WILLIAM J. GIES. 


SEVERAL years ago, during a comparative study of the reactions of 
various gelatines, the results of which have not yet been published, it 
was observed by Dr. D. H. M. Gillespie and myself that dilute solu- 
tions of potassium chromate did not precipitate gelatine solutions, but 
that when such proteid-chromate mixtures were further treated with 
acid, a fine yellow flocculent precipitate formed at once. Acids as 
“weak” as acetic, and also the common mineral acids, effected the 
result, the latter acids more promptly, however, even in smaller 
amount. 

At intervals I have returned to this reaction, and lately have made 
a more careful study of it. Solutions of chromates of mono- and 
divalent cations (the only ones thus far employed) cause no precipi- 


XVI fifteenth Annual Meeting. 


tates in neutral or alkaline proteid fluids, but on further treatment 
with small amounts of dilute acids, — strongly dissociable ones par- 
ticularly, — flocculent precipitation of a proteid-chromate compound 
occurs in every case. The reaction is especially striking with such 
bodies as gelatine and proteose (the precipitates with these disap- 
pearing on warming and reappearing on cooling), and it seems to be 
more delicate than the acetic acid and potassium ferrocyanide test. 
Salts containing dichromion or trichromion behave differently. 

Since bichromate is formed from chromate on the addition of acid, 
it might be supposed that such production is responsible for the pre- 
cipitation observed. But bichromate solutions are as inert as the 
chromate. When, however, acid is added to a mixture of proteid 
and bichromate, precipitation occurs, as in the case with chromate. 
Hydroxidion prevents the reaction in all cases. Possibly the precipi- 
tation is due to the formation of dichromic acid, just as in the acetic 
acid and potassium ferrocyanide test it is dependent on the formation 
of hydroferrocyanic acid. 

Further study is expected to determine exactly the character of the 
ions responsible for the reaction. The results thus far point to 
dichromanion in the presence of hydrion. 


NEW EXPERIMENTS ON THE PHYSKOLOGICAL ACTION 
OF THE PROTEOSES. 


By LAFAYETTE B. MENDEL anp FRANK P. UNDERHILL. 


In a recent study of the physiological action of the proteoses, after di- 
rect introduction into the circulation of higher animals, Pick and Spiro 
(Zeitschrift fiir physiologishe Chemie, 1900, xxxi, p. 235) conclude: 
“Es gibt ‘ Peptone’ ohne ‘ Peptonwirkung’ und ‘ Peptonwirkung’ 
ohne‘ Peptone.’” Their experiments gave the following indications: 
(1) It is possible to obtain typical albumoses and peptones after cleav- 
age of proteids (fibrin) by trypsin, autolysis, and alkalies, and by the” 
action of acid on casein and edestin. These products, when intro- 


duced into the blood, fail to show any effect on blood-pressure. - 


furthermore, active products prepared by acid or pepsin-acid lose 
their typical action after purification by a method which apparently 
does not alter their chemical character. (2) It is possible to obtain 
preparations which develop the anti-clotting action in marked degree, 


Proceedings of the American Physiological Society. xvii 


although they may contain only traces of albumoses and peptones, or 
even none at all. From this the authors conclude that the property 
of retarding blood-clotting is attributable not,to the albumoses as 
such, but to an adherent and extremely active substance which is 
present, perhaps, in very small quantity. 

Earlier experiments in our laboratory (This journal, 1899, ii, 
p. 142) have led us to question the general application of these 
conclusions in the case of proteoses purified according to the stand- 
ards current for these substances. Since then a considerable num- 
ber of typical products have been prepared under conditions which 
would exclude as far as possible the presence of the specific contami- 
nating tissue substance (peptozyme) assumed by Pick and Spiro to be 
the active agent. The number includes native proteoses separated 
directly from the wheat embryo, hemp seed, and brazil nut without 
the use of foreign enzymes or acids ; proteoses prepared by the action 
of superheated water or of dilute acids on crystallized proteids; pro- 
teoses prepared by the digestion of edestin and casein with the veg- 
etable enzymes papain and bromelin: products purified by the methods 
of Pick and Spiro. All of these proteose preparations, introduced in 
doses of 0.3 to 0.5 gm. per kilo of body weight into the circulation of 
dogs, have provoked the characteristic symptoms noted in our earlier 
work, namely: a fall in arterial pressure, diminished coagulability of 
the blood, a transitory stage of excitation followed by narcosis, a 
degree of “immunity,” and (in the single experiment thus far tried) 
lymphagogic effects. The investigation is being continued. 


IS THE ACTION OF ALCOHOL ON GASTRIC SECRETION 
SPECIFIC? 


By GEORGE B. WALLACE anp H. C. JACKSON. 


EXPERIMENTS were performed with the object of determining 
whether the flow of gastric juice caused by the introduction of 
alcohol into the intestine is due to a purely reflex action, and further 
whether it is an effect produced by irritant substances other than 
alcohol. 

Dogs were used in the experiments. The stomach was washed 
till free of acid, the pylorus then ligated, and the substance to be 
studied injected directly into the intestine. Three hours after the 


XVIll Fifteenth Annual Meeting. 


injection the dog was killed, the stomach removed and its contents 
collected and analyzed. The total amounts as well as the percentage 
of the various acidities were estimated. In some of the experiments 
the nerve supply to the stomach was cut off, by extirpating the 
cardiac ganglion of the sympathetic system and cutting the vagus 
fibres passing down the walls of the stomach. Oil of peppermint was 
taken for comparison wifh alcohol as an irritant. 

40 c.c. of 25 per cent alcohol, injected into the intestine, produced 
a gastric flow five times as great in amount as when water alone was 
injected. The total and free acid was increased in the same propor- 
tion. 0.3 c.c. of oil of peppermint, in acacia emulsion, produced an 
effect practically identical with that of alcohol. ; 

With the gastric nerve supply cut off, the alcohol and oil of pepper- 
mint injected into the intestine produced no greater effect than when 
water alone was injected. 

Even when alcohol was injected directly into the stomach, the 
gastric nerve supply being cut off, it induced a gastric secretion but 
little greater than when water was injected. 

The conclusions are (1) that alcohol introduced into the intestine 
stimulates the gastric secretions not through its absorption and sub- 
sequent action on the gastric mucosa, but by a purely reflex action; 
(2) that this action is not limited to alcohol, but is possessed by other 
irritant substances, such as oil of peppermint. 


A PRELIMINARY REPORT ON THE PHARMACOLOGICAL AND 
CHEMICAL PROPERTIES OF TRI-BROM-TERTIARY 
BUTYL-ALCOHOL. 


By E. M. HOUGHTON anp T. B. ALDRICH. 


WILLGERODT, in a paper dealing with tri-chlortertiary butyl-alcohol 
or acetone chloroform, mentioned having obtained a similar product 
containing bromine in place of the chlorine, which he called tri-brom- 
tertiary butyl-alcohol, but he did not make a careful examination of 
its chemical and physical properties, and seems not to have considered 
at all its pharmacological properties. This bromine compound is 
produced by the action of caustic alkalies upon mixtures of bromoform 
and acetone. The excess of acetone and bromoform having been 
distilled off, the new bromine compound is removed from the residue 


Proceedings of the American Physiological Society. xix 


by distillation with steam, the product being finally purified by re- 
crystallization from alcohol or other suitable solvent. The purified 
substance is a white crystalline body having a camphoraceous odor 
and taste. The melting point is about 167° C. It is soluble in most 
of the organic solvents, as alcohol, ether, benzoine, etc., slightly 
soluble in cold and more soluble in hot water. 

This compound, since it is a derivative of the fatty acid series, 
when administered in various ways to animals possesses decided 
anzesthetic properties; recovery from small quantities takes place 
without any apparent untoward results. The drug appears to have 
very little influence upon the heart or circulation, as shown by 
myocardiographic and blood-pressure tracings taken from curarized 
animals. 

It is suggested that the name brometome be given to the compound. 


THE ACTION OF ETHYL-ALCOHOL ON CONTRACTILE 
PROTOPLASM. 


By FREDERIC S. LEE. 


THE spontaneous contractions of the bell of the medusa, Gonionema, 
are markedly increased in number by small quantities of ethyl- 
alcohol in sea-water. This increase becomes progressively greater 
with solutions ranging from ;; per cent to } per cent; then pro- 
gressively less, until with about 2 per cent the contractions are 
nearly normal in number; with stronger alcohol they are irregular, 
feeble, and partial. These results agree with those obtained by the 
author and Salant on frog’s striated muscle, and strengthen the 
proof that ethyl-alcohol in small quantity is favorable, in larger 
quantity unfavorable, to the activity of contractile protoplasm. The 
behavior of the bell of Gonionema, with the nerve-ring removed, 
supports this conclusion. 

Concentrated sea-water acts like weak alcohol, but the amount of 
concentration that is required in order to produce a given increase 
in the rate of contraction is so great as to preclude the idea that 
abstraction of water from protoplasm is the important causative 
factor in the favorable action of alcohol. 


XX Fifteenth Annual Meeting. 


EXHIBITION OF NEW FORM OF PLATINUM—MERCURY 
STIMULATOR. 


By W. P. LOMBARD (For W. P. BOWEN). 


Four wooden disks two inches in diameter are mounted on a light 
shaft so as to rotate in separate mercury troughs in the wooden 
base. A platinum wire is placed radially in each disk, and brought 
even with the surface of the edge; the disks are adjusted on the 
shaft so that in each of the two pairs the wires are opposite; then 
the central ends of the wires of each pair are connected, thus form- 
ing two metallic staples whose ends dip in the mercury and con- 
nect the troughs in pairs. One pair of disks is now set so as to 
make and break contact slightly before the other pair. The 
binding posts of one pair of troughs are now connected with the 
primary, and those of the other with the secondary circuits of an 
induction coil. 

Practice in the laboratory has shown this to be a very efficient 
stimulator for rates up to 3 or 4 per second. At higher rates, the 
mercury is thrown, but the use of a cover will permit rates some- 
what higher. The staple of platinum being imbedded in the smooth 
wooden disk prevents waves of mercury being formed as the staple 
passes through, —a fault of many otherwise excellent stimulators in 
common use. 


EXHIBITION OF MERCURY-MERCURY STIMULATOR. 


By W. P. LOMBARD (For W. P. BOWEN). 


PROFESSOR WARREN of Bryn Mawr once said that “ Eternal bright- 
ness is the price of electricity.” Constant brightness at the point of 
contact is one absolute essential of the ideal stimulator. It is secured 
in this instrument by making and breaking a column of mercury by 
means of a celluloid disk perforated with holes of uniform size and 
rotated at a uniform rate. 

The mercury is held in wooden reservoirs; from the base of each a 
glass tube projects horizontally; two of these are placed on opposite 
sides of the disk with the axes of the tubes forming one line and the 
ends of the tubes just far enough apart for the disk to rotate be- 
tween. The disk interrupts the circuit except while a hole is 


Proceedings of the American Physiological Society. xxi 


passing between the ends of the tubes. At such a time there is 
a momentary contact of mercury from the two tubes; then the 
hole moves on, carrying with it the mercury which fills it, includ- 
ing most of the oxide produced by the spark, and this waste 
mercury drops into a beaker beneath the instrument. 

~The frequency of stimulation can be varied either by changing 
the rate of rotation of the disk or the number of holes contained in 
it. This instrument has given good curves with a 2-hole disk at 73 
revolutions per second, which is the fastest rate tried. With a 6-hole 
disk this gave 45 stimuli per second, and the rate can be increased to 
270 per second by making more holes in the disk. The ease with 
which the disks can be made also makes it possible to vary the 
manner of stimulation widely, if desired. 

A second pair of reservoirs, placed on the opposite side of the shaft 
turning the disk and properly adjusted, provides a short circuit for 
cutting out the making shocks. 

Most stimulators, as a result of corrosion due to sparking, fail to 
give regular results with less than a maximal stimulus; but this in- 
strument gives fine curves with almost any strength of current. 


FURTHER OBSERVATIONS ON THE MOVEMENTS OF THE 
STOMACH AND INTESTINES. 


By W. B. CANNON. 


THE cat has been used in the researches I have hitherto reported; it 
is important to know if the observations made on the cat are true 
also for other animals. 

Rhythmic segmentation of the food has been repeatedly observed in 
the small intestine of the dog, at the rate of from 18 to 22 operations 
per minute; and in the white rat, 44 to 48 operations per minute. 
The guinea pig and rabbit have shown no typical segmenting move- 
ments. A to-and-fro swinging of the food in one place has been 
seen in the rabbit; and also segmentation combined with peristalsis, 
as described for the cat. 

Antiperistalsis at the beginning of the large intestine has been 
observed in the dog, at the rate of four waves per niinute. These 


1 CanNnoN: This journal, 1902, vi, p. 260. 


47 


XX fifteenth Annual Meeting. 


antiperistaltic waves in the dog and cat are possibly to be correlated 
with antiperistalsis in the cecum of the rabbit. The appearance of 
fresh food in the caecum of the rabbit is followed by a deep constric- 
tion which sweeps the food swiftly toward the blind end. 

The stomach movements were similar in all animals observed. In 
the dog, rabbit, guinea pig, cat, and rat peristalsis is seen only over 
the pyloric end. . 

The gastric peristalsis can be inhibited in the rabbit, the dog, and 
the guinea pig, just as in the cat, by causing respiratory distress. 

Feeding a cat a fluid food makes it possible to demonstrate a fairly 
rhythmic relaxation of the cardia. At every relaxation the food pours 
through the cardia and passes up the cesophagus to the level of the 
heart or the base of the neck; and it is immediately swept thence 
into the stomach again by a peristaltic wave. This activity may 
recur about every half minute for a half hour or more. 

Observations have also been made on the treatment of different 
foodstuffs in different parts of the alimentary canal. Cats were given 
equal amounts of carbohydrate or proteid foods, having as near 
as possible the same consistency. The carbohydrate food appeared 
in the intestine usually within ten minutes; the proteid food usually 
did not begin to leave the stomach for an hour or an hour and a half 
after it was eaten. Similar observations were made in comparing 
fats and proteids ; the proteids were retained in the stomach nearly 
twice as long as fats of the same amount and consistency. In the 
small intestine the proteids may be seen undergoing segmentation 
at almost any time; the carbohydrates and fats undergo this process 
comparatively rarely. 


SOME MINOR IMPROVEMENTS IN LABORATORY APPARATUS. 
By COLIN C.’ STEWART. 


1. A Moist chamber; 2. An electro-magnetic signal; 3. A frog- 
table; 4. A mercury key in which the movable part of the key is 
hinged in the bottom of one mercury cup to dip into a second, with 
fixed binding posts running to each cup; 5. A contact end for a flex- 
ible wire cable; 6. A cardiopneumatiscope, consisting of a curved 
glass tube sealed to a pin-hole opening at the distal to prevent 
the too rapid escape of the cigarette smoke with which the tube is 


Proceedings of the American Physiological Society. xxiii 


filled ; 7. A tracheal cannula, for artificial respiration, in which the 
troublesome metal collar which regulates the size of the opening 
through which the expiratory blast escapes, is replaced by a short 
piece of rubber tubing. 


ON THE BIOLOGICAL RELATION OF PROTEIDS AND 
PROTEID ASSIMILATION. 


BY P. A. LEVENE anp L. B. STOOKEY. 


IN applying the precipitin test, the authors observed that different 
proteids of a given animal, and perhaps of a given species, possess a 
certain similarity which distinguishes them from all the proteids of 
any other origin. This biological individuality of proteids could serve 
to explain the cause of the fact that proteid material ingested has to 
be broken up by the organism before it is utilized. The molecule of 
foreign proteid material has to be reconstructed into the molecule 
characteristic of the given animal. 

The second aim of the authors was to establish the place of the 
breaking down of ingested proteid. Attempts were made to solve 
also this problem by the application of the precipitin test. This part 
of the work is not yet complete. The authors, however, think that 
foreign proteids normally do. not pass the digestive system (liver 
included) unchanged. 


ON THE DIGESTION OF GELATINE. 


BY Po A. LEVENE. Anp) L. B. STOOKEY. 


Ir is known that gelatine does not yield on digestion with proteo- 
lytic enzymes any perceptible quantity of the usual crystalline 
digestive products. It was, however, established by Chittenden and 
Solley that it undergoes transformation into proteoses and peptone 
like any other proteid. One of the authors, Levene, found that these 
digestive products differ in their content of glycocoll. He also ob- 
served that gelatine subjected to prolonged digestion yields compara- 
tively little of the usual crystalline digestive products. In order to 
gain some light upon the process of transformation of gelatine into 
gelatoses, the authors investigated the formation of free ammonia in 
the course of tryptic digestion of gelatine, and observed that the free 
ammonia increases as the gelatine is transformed into the primary 
and the primary into the secondary albumoses. 


XXIV Fifteenth Annual Meeting. 


THE SECOND MAXIMUM IN THE RESPONSE OF MUSCLE 
TO STIMULATION. 


By COLIN C. STEWART. 


WirH single break induction currents of gradually increasing 
strength the contraction of the frog’s gastrocnemius (and other 
muscles) increases from a minimum to a maximum. Further in- 
crease in the strength of the current elicits no greater response for 
a time, but finally results in a second increase in the height of the 
contraction and a second and final maximum. 

This second maximum is not due to any peculiarity of the induc- 
tion apparatus, nor to the presence in the gastrocnemius of muscle 
fibres of two different lengths. It is obtained with curarized muscle 
stimulated directly, as well as with uncurarized muscle stimulated 
indirectly. The maxima for muscle stimulated indirectly occur 
earlier, that is, with weaker currents, than the maxima which are 
to be obtained with direct stimulation. It follows, therefore, that 
with fresh uncurarized muscle stimulated directly, four maximal 
plateaus are often observed. 

The second maximum is directly ascribable to the presence in these 
muscles of two substances differently affected by changes in tempera- 
ture, by fatigue, and by drugs. Such changes in the whole muscle 
are marked by corresponding changes in the form of the single 
muscle curve, particularly in respect to the presence and relative 
importance of the two apices, which are almost characteristic of the 
gastrocnemius contraction curve. And with each change in the form 
of the muscle curve there is a corresponding change in the relative 
prominence of the two maximal plateaus in the series of contractions 
obtained with gradually increasing induction currents. 


THE ACTION OF THE TWO-JOINT MUSCLES OF THE HIND-—- 
LEG OF THE FROG, WITH SPECIAL REFERENCE TO THe 
SPRING MOVEMENT. 

By WARREN P. LOMBARD. 


Because of the peculiar relation of the two-joint muscles to the 
joints which they cross, any force, whether external or a muscular 
contraction, which flexes or extends hip, knee, or ankle, tends to 


Proceedings of the American Physiological Soctety. xxv 


cause a like movement of the rest. These effects are produced by 
the passive, tendon-like action of the two-joint muscles, and are 
increased by their contractions. It is possible for any of the joints 
to be moved independently, where the two-joint muscles as a whole 
are not contracting. These muscles are attached under very little 
tension, and are sufficiently extensible to yield to even a compara- 
tively slight stretching force. A single joint can therefore be flexed 
or extended by the action of one-joint muscles or by the action of a 
two-joint muscle when the other end of the muscle is prevented by 
one-joint muscles from moving the other joint which it ordinarily 
controls.. A ¢wo-joint muscle may indirectly cause the extension of the 
joint of which it is a flexor. This may occur under the following con- 
ditions: a. The muscle in question, A, must have a better leverage 
at the end by which it extends, than at the end by which it flexes. 
6. There must be on the opposite side of the ieg a two-joint muscle, 
B, which flexes the joint which A extends, and extends the joint which 
A flexes. c. The extensor leverage and strength of A must be suffi- 
cient to enable it to make use of the tendon action of B. When all 
the two-joint muscles are acting at the same time, the energy developed 
by these muscles is transmitted as by an endless chain, having the form 
of a figure 8, with the crossing at the knee, and the effect progresses in 
the direction of the greatest leverage. A preliminary series of typi- 
cal curves of the leverage exerted by each end of all the two-joint 
muscles, and by the one-joint muscles of the knee and ankle, was 
shown. In general, these curves are in harmony with the view 
advanced as to the tendon action of the two-joint muscles in the 
spring movement. Not only is the leverage of the two-joint muscles 
when the leg is flexed in favor of the extending tendon action 
described, but as the leg approaches extension the leverage of many 
muscles changes, extension power lessening and flexion power in- 
creasing. The effect of this is to protect the joints and favor the 
recovery from the spring. Finally, it may be said that the action of 
many of the two-joint muscles, and at least of one of the one-joint 
muscles, may be reversed as the leg passes from flexion to extension. 
A muscle which acts to flex a joint in one position may extend it in 
another, and even flex it again in still another. This fact demands 
caution in the classification of muscles as flexors and extensors. 


XXVI1 fifteenth Annual Meeting. 


THE TONUS OF HEART MUSCLE. 
By W. T. PORTER. 


Tue following facts were demonstrated by graphic records: 

1. As the tonus increases the conductivity of heart muscle 
diminishes. For example, the latent period of an extra-contraction 
produced between two spontaneous tonus contractions was ten-hun- 
dredths second, while the latent period of an extra-contraction pro- 
duced near the summit of a spontaneous tonus contraction was 
sixteen-hundredths second. 

2. The height of the tonus contraction is proportional to the 
strength of the stimulus. The law of “all or none” does not 
apply. 

Induced current in 
Kronecker units 


(successive 
stimulations). 


Height of tonus con- 
| traction in millimetres 
(auricle of tortoise). 


3500 7mm. 
3500 7mm. 


3000 5 mm. 


4000 | 11 mm. 


5000 25 mm. 


Similar results were obtained with the ventricle, but the changes 
were less marked. 

3. The tonus contraction has no refractory period. Extra-contrac- 
tions (both of tonus form and fundamental form) may be produced 
during any phase of tonus contraction. 

4. Tonus contractions may be superposed as contractions of skele- 
tal muscles are superposed. 

5. Summation of sub-minimal stimuli of tonus contractions was not 
secured, though many efforts were made. 

The induced currents employed inhibited the spontaneous funda- 
mental contractions. The spontaneous fundamental contractions 
regained their former height only after several contractions in 
“staircase” form. 


Proceedings of the American Physiological Soctety, xxvii 


THE EFFECT OF EXTIRPATION OF THE GASSERIAN GAN- 
GLION UPON THE SENSE OF TASTE. 


By HARVEY CUSHING. 


As the result of observations upon clinical cases, it has been found 
by many observers that lesions which have involved the fifth nerve 
have almost invariably been associated with a loss of taste over the 
corresponding side of the tongue. 

The writer has given especial attention to the loss or preservation 
of taste in fourteen cases of extirpation of the Gasserian ganglion, 
and the uniformity of the results in this long series of cases seems to 
justify the conclusion that the trigeminal nerve does not serve as a 
path of transmission for these fibres. In one case only a single 
observation was made, namely, six days after the operation; the 
sense of taste was found to be lost. It probably returned subse- 
quently, but this was not proved. In a considerable percentage of 
the other cases a similar loss or diminution in the acuteness of taste 
has been observed at a corresponding time after the operation. It 
has been found, however, that after a period of some days, taste 
always returns, even though a condition of anesthesia to pain, touch, 
and temperature invariably persists after the extirpation. It seems 
not impossible that the degeneration taking place after the ganglion 
extirpation in the fibres of the lingual nerve, which are so intimately 
associated with the fibres of the chorda tympani, may in some way 
interfere for the time being with the normal transmission of taste 
impulses by way of the chorda. It is possible also that this degen- 
eration may produce toxic substances which for the time being act 
directly upon the chorda fibres as a physiological ‘‘ block” to their 
normal activities. However this may be, it has been found, contrary 
to the observations of Gowers, Stewart, Turner, and others, that the 
extirpation of the Gasserian ganglion affects in no way taste upon 
that portion of the tongue presided over by the glossopharyngeal 
nerve, and has only a temporary effect in a small percentage of cases 
in diminishing the perception of taste over the anterior two-thirds or 
chorda territory of the tongue. 


XXVili Fifteenth Annual Meeting. 


SALIVARY DIGESTION IN THE STOMACH. 


BY We -B CANNON Anp) He BDAY. 


EviDENCE was brought forward in 1898 by Dr. Cannon! that Beau- 


mont’s and Brinton’s theories of mixing currents in the stomach were | 


not true. The food in the fundus, over which peristaltic waves do 
not pass, is not mixed with the gastric secretions. Since free hydro- 
chloric acid does not appear in this part of the stomach for an hour or 
more, salivary digestion may continue during that time uninterrupted 
by the acid of the gastric juice. To study further the possible differ- 
ence between carbohydrate digestion in the active pyloric end and 
in the quiet cardiac portion of the stomach, this research was under- 
taken. 

Crackers free from sugar were used as food. In all cases a uniform 
amount of crumbed crackers was mixed with a uniform amount of 
saliva sufficient to make a thick mush. This food was either mixed 
a little at a time, and given at once to a cat to swallow, or mixed and 
introduced by means of a stomach tube. The animals were allowed 
to live one-half, one, or one and a half hours after eating. They 
were then quickly etherized, the abdomen was opened, and a ligature 
tied tightly between the cardiac and pyloric portions of the stomach. 
The pylorus and cardia were next tied, and the stomach then removed. 

The contents of the stomach were received in evaporating dishes. 
The pyloric contents were invariably fluid; the cardiac contents often 
retained their shape so that it was possible to get the internal and 
external food of the region. After enzyme action had been stopped 
by bringing the food to the boiling point, the food was evaporated to 
dryness over steam. The dry mass was then powdered, and the 
sugar content estimated according to Allihn’s method. 

The results show that at the end of an hour the sugar present in 
the cardiac contents averages almost twice that present in the pyloric 
contents, and may be two and a half times as much. When the 
food is liquid, the ratio is diminished, 2. ¢., it is about six to five 
instead of two to one. Also, when small amounts of food are given, 
the sugar content is about the same in both parts of the stomach. 
Two cases in which the stomach was massaged at intervals during 
digestion showed a larger amount of sugar in the pyloric end than in 
the cardiac end. The largest amount of sugar, estimated as maltose, 
which has been found in the stomach contents, is about 49 per cent. 


1 CANNON: This journal, 1898, i, p. 375. 


Proceedings of the American Physiological Society. Xxx 


ON THE ELEMENTARY COMPOSITION OF ADRENALIN. 
By JOHN J. ABEL. 


Tue fact that adrenalin is so easily convertible into the alkaloidal 
modification suggests at once that a simple relationship exists between 
these two modifications of the blood-pressure-raising principle. As 
prepared by the zinc-ammonia process already described,! washed 
entirely free of ammonia and dried in vacuo over sulphuric acid, 
adrenalin is found to be a stable product as long as it is kept perfectly 
dry. As made by this process, redissolved in dilute hydrochloric 
acid, and reprecipitated with ammonia, its composition was found 
to be: 


CESS Miohoy a0); 
Fe =" 6:29) tol” 6177; 
N = 7.38 (KJELDAHL-GUNNING). 


After nine precipitations with ammonia or sodium carbonate, its 
composition changed to: 


C = 58.61 to 58.67, 
H = 6.77 to 6.84, 
N = 7.08 (KJELDAHL-GUNNING). 


These results are such that a formula could be deduced from them, 
viz., C,jH,;NO,.4H,O; but unfortunately it was later discovered that 
the Kjeldahl-Gunning method fails to give all of the nitrogen of 
the compound, and these results are therefore only of value as show- 
ing that repeated precipitation raises the carbon content of the 
compound, 

Commercial adrenalin was now purchased in several 100-grain lots 
and purified by shaking its hydrochloric acid solution with ether, and 
by repeatedly precipitating it from an acid solution with ammonia or 
sodium carbonate. More than thirty analyses have been made, but 
it has been found impossible to secure uniformity of composition 
among the various fractions. 

Setting down all of the variations in respect to C, H, and N in the 
form of a table, we find that the extremes in respect to each element, 
run from — 


56.53 


58.89, 
iv 1 
Uk 


to 
7 to 7. 
59 to 10.65 (DuMAs). 


1 Johns Hopkins Hospital bulletin, November, 1901, and February-March, 1902. 


XXX Fifteenth Annual Meeting. 


A very low hydrogen content, as 4.77, 5.05, 5.46, was less fre- 
quently met with than figures that went over 6 per cent, but these 
low hydrogen percentages were found to occur with high as well as 
with low percentages of carbon, and generally with high percentages 
of nitrogen. The precipitations with alkalies were repeated as many 
as ten times. 

It is very evident then that adrenalin cannot yet be spoken of as 
having a “constant composition” (Takamine), and as being a pure 
chemical individual. 

The writer has found that the substance, purified as far as possible 
by the above processes, is soluble to a very considerable extent in 
warm oxalic ethyl ester, and that it may be precipitated from this 
solvent by the free addition of ether, in the form of a water-soluble 
oxalate. It is hoped that this and other methods of purification will 
soon enable the writer to determine the true elementary composition 
of adrenalin and its exact relationship to the earlier epinephrin 
series. 


ON THE BEHAVIOR OF EXTRACTS OF THE SUPRARENAL 
GLAND TOWARD FEHLING’S SOLUTION. 


By JOHN J. ABEL. 


A PURIFIED extract of both beeves’ and sheep’s suprarenal glands, 
which are rich in the blood-pressure-raising principle, behaves in the 
following manner toward Fehling’s solution. If an aqueous solution 
of the extract is poured into an excess of boiling Fehling’s solution 
(Fehling 1 to water }) and the mixture is kept at the boiling point 
for two minutes and then cooled, no cuprous hydrate or cuprous 
oxide settles out. A flocculent, greenish-white copper compound 
will be found suspended in the fluid, or deposited on the bottom of 
the flask. After boiling from five to six minutes, a considerable 
reduction occurs, and after boiling for fifteen minutes, the reduction 
appears to have reached a maximum, and a heavy deposit of yellow 
cuprous hydrate, with possibly a small admixture of cuprous oxide, is 
obtained. 

After precipitating the adrenalin with sodium carbonate from a 
given portion of extract, the filtrate, contrary to the statements of 
Aldrich, also gives a very abundant precipitate of yellow cuprous 
hydrate when boiled with Fehling’s solution for from six to fifteen 


Proceedings of the American Physiological Society. Xxxi 


minutes. Such a filtrate is estimated to contain from 15 to 20 per 
cent of adrenalin, and the amount of reduction obtained is apparently 
proportional to this unprecipitated adrenalin. 

While extracts of the gland require prolonged boiling in order to 
effect complete reduction, salts of epinephrin or adrenalin are oxidized 
with the greatest ease and rapidity when brought into contact with 
a hot Fehling’s solution; the reaction indeed begins far below the 
boiling point. V. Fiirth’s iron compound has also been found to 
reduce Fehling’s solution on being boiled with it for a sufficient 
length of time. 

It was clearly stated in a former paper! that epinephrin in its 
native, unaltered state, as found in extracts of the gland, and as con- 
tained in v. Fiirth’s basic lead precipitate, failed, in my hands, to 
reduce Fehling’s solution, but that my ow compounds of epinephrin, 
as obtained in various ways, all agreed with adrenalin in their ability 
to reduce this reagent , and that, in this particular, there was no differ- 
ence between them. 


ON THE OXIDATION OF EPINEPHRIN AND ADRENALIN 
WITH NITRIG ACID. 


By JOHN J. ABEL. 


ReEsuLts thus far obtained would indicate that the products obtained 
in the oxidation of epinephrin, C,,H,,NO,, and adrenalin are identi- 
cal. The following example will illustrate the process: 

In very small portions at a time, 10 grams of purified adrenalin are 
dissolved in 60 c.c. of nitric acid, sp. gr. 1.2, in a platinum bowl 
on the water bath. The oxidation goes on with great violence, and 
care must be taken to avoid loss of material due to foaming. When 
the solution is complete, and the evolution of gases has largely sub- 
sided, 10 c.c. of fuming nitric acid is added, and the whole is concen- 
trated on the water bath, water being added from time to time as the 
mass begins to thicken. 

After this treatment has been continued for some hours, the con- 
tents of the dish, on cooling, are found to consist of a solid mass of 
odorless crystals. The larger part of this crystalline material con- 


1 See pages 337-338, and Conclusions 1, 2, and 3 of Summary, on page 343, 
Johns Hopkins Hospital bulletin, rgot, xii; also: zdem, July, 1897. 


XXX1l Fifteenth Annual Meeting. 


sists of oxalic acid. The barium, lead, sodium, and calcium salts of 
this acid were prepared, as also its di-ethyl ester. The calcium salt, 
as prepared from the sodium salt, was found to contain 27.39 per cent 
Ca, the theoretical amount for CaC,O,.H,O as prepared from hot 
concentrated solutions of sodium oxalate by precipitation with calcium 
chloride. The tetragonal crystals, CaC,O, . 3H,O, were also prepared, 


and the acid was shown to behave toward potassium permanganate 


and other reagents as does oxalic acid. 

The other chief product of the above-described oxidation consists of 
a hygroscopic, crystalline salt (oxalate?) of the nitrogenous base to 
which reference has frequently been made and which I have called the 
coniine-piperidine-like body, on account of its peculiar, offensive, and 
penetrating odor. This part of the molecule has apparently remained 
entirely intact. Further treatment with iodine trichloride does not 
destroy it, but enables one to obtain it in the form of slender prisms, 
very soluble in water and alcohol, and little soluble in ether. The 
addition of an alkali to these crystals liberates the peculiar odoriferous 
base, exactly as if it had been added to epinephrin itself. If these 
crystals, or those obtained in the first place from the use of nitric 
acid, are fused with powdered potassium hydrate, an odor like that of 
pyrrolidine is obtained, later this gives place to the fishy odor of 
amines, and later still only pyrrol itself is evolved. In its behavior 
toward other destructive chemicals, as when fused with zine dust, it 
also reminds one of the behavior of certain pyrrol derivatives under 
similar circumstances. 


ON THE INFLUENCE OF CAMPHOR INGESTION UPON THE 
EXCRETION OF DEXTROSE IN PHLORHIZIN DIABETES. 


By H. C. JACKSON. 


Tue work of v. Mering, Minkowski, Cremer, Lusk, etc., has shown 
that in phlorhizin and pancreatic diabetes there exists a definite ratio 
between the amount of dextrose and nitrogen eliminated by the 
kidneys. This D: N ratio is 2.8: 1 in all cases except that of the 
dog in phlorhizin diabetes, where it is 3.75 : 1. The feeding of cam- 
phor to a dog made diabetic with phlorhizin, and upon which the 
relation 3.75 : 1 had been obtained, immediately decreased the 
D: N ratio to 2.8 : 1, or equal to that prevailing in all other animals 
with phlorhizin, and in the dog in pancreatic diabetes. 


aoe 


Proceedings of the American Physiological Society.  XxxXiii 


Microscopical examination of the kidneys of dogs fed with camphor 
alone, showed marked morphological changes in the cells of the 
ascending loops of Henle, the fatty infiltration or degeneration being 
limited to this group of cells and not diffuse, as is the case of phlor- 
hizin poisoning.’ In some cases of phlorhizin poisoning, albuminuria 
sets in, and Lusk has noticed that in such cases also the ratio of 
DN fell to 2.8: 1 from 3.75 : 1. This may be explained by the 
supposition that phlorhizin in extreme cases causes changes in the 
cells of the loops of Henle similar to those induced by camphor and 
with similar results. 

This decrease in ratio amounts to 25 per cent of the absolute 
amount of the sugar eliminated, and it seems probable that this frac- 
tional part of the total carbohydrate excretion is not present in the 
blood in the same form in which it appears in the urine, but suffers 
in some way a change into dextrose as it is excreted by the cells of 
the loops. When these degenerate, as under camphor treatment, 
they refuse to functionate, and the amount of sugar as compared to 
nitrogen appearing in the urine is decreased, and consequently a fall 
in the ratio is noticed. The remaining 75 per cent of carbohydrate 
is evidently eliminated unchanged. 

The evidence seetms favorable to the view of a double origin of the 
dextrose excreted in the urine in phlorhizin diabetes. 


THE TOXICITY OF EPINEPHRIN (ADRENALIN). 
By SAMUEL AMBERG. 


THE toxicity of epinephrin was tested in experiments on dogs with 
intravenous, subcutaneous, and intraperitoneal injections. A dose of 
2.0 mgm. per kg. intravenously proved sufficient to kill an animal. 
One dog with 0.99 mgm. per kg. survived. The injections were made 
quickly. Animals which had been subjected to a chloroform asphyxia 
succumbed during the first few minutes following the injection. 
When the animals were in good condition, a longer time elapsed 
before death ensued. One animal, which had received a dose of 
4.9 mgm. per kg. subcutaneously, survived, while one with 6.0mgm. 
per kg., and others with more, died. The fatal dose by intraperitoneal 
injections lies, according to Herter, between 0.5 and 0.8 mgm. per kg. 

Upon the heart action the drug exercises an influence by an initial 


XXXIV Fifteenth Annual Meeting. 


stimulation of the vagus, followed by a paralysis. Besides that, it has 
a direct injurious effect on the heart, as well as on the respiration. 
The pathological changes produced by the drug consist in hemor- 
rhages, which were observed in the heart, lungs, intestines, perito- 
neum, in and around the pancreas, liver, adrenal glands, and thymus 
glands. 


THE INFLUENCE OF THE H ION IN PEPTIC PROTEOLYSIS. 


By WILLIAM J. GIES. 


Tue fact that pepsin shows digestive power only when acid is present 
implies the dependence of the enzyme upon hydrion for its activity. 
It has frequently been observed that various acids are efficacious in 
this connection, though in different degree. 

In some recent experiments on the influence of acidity, I have used 
purified fibrin, edestin, and elastin as the indicators. Undigested 
residue, neutralization precipitate, and uncoagulable products were 
determined quantitatively in each digestive mixture. Various com- 
mon mineral and organic acids were employed. Varying propor- 
tions of pepsin and acid were taken in uniform volumes (100 C.c.), 
with the same amount of proteid (1 gm.). 

In eguipercentage solutions of acids whose anions have no precipi- 
tative effect on proteid, the relative proteolysis is very different, 
being greatest in “ strong” acids such as HCl, and least in ‘‘ weak” 
acids, such as CH,. COOH.  £guwimolar solutions of the same acids 
gave more concordant results in some respects, although the differ- 
ences between the effects in such acids as HC] and CH,. COOH were 
still very wide. With eguzhydric solutions, the results showed greater 
harmony, though there were still striking divergences. H,PO,, HCl, 
HNO,, HClO;, H,AsO,, and (COOH),, in strengths equivalent to 
decinormal KOH (with 50 mgm. of pepsin preparation, in 100 C.c. 
at 40°C., four hours), showed practically the same ability to assist 
pepsin in the digestion of 1 gm. of fibrin. 

Additional experiments, especially with eguédissociated solutions of 
the acids referred to above, are expected to show the influence 
not only of hydrion, but also of the anions, if the influence of the 
latter in the acids referred to be appreciable. Similar experiments 
are about to be extended to other enzymes. 


_—— 


at Ohad 


Proceedings of the American Phystological Socrety. Xxxv 


SOME OBSERVATIONS ON THE COAGULATION OF MILK. 
By A. S. LOEVENHART. 


THERE is a stage in the rennin coagulation of milk when boiling 
causes a coagulum to separate. The milk acquires the property of 
yielding a heat coagulum before there is any apparent alteration in 
the consistency of the milk. This represents a stage in coagulation 
of milk whether the rennin be of gastric or pancreatic origin. The 
term ‘“ metacasein reaction,” introduced by Roberts, may be con- 
veniently retained to signify this heat coagulation. The interval 
between the time the metacasein reaction may be obtained, and the 
spontaneous coagulation, varies inversely with the amount of rennin 
acting. Thus it may be so transient that it cannot be detected, or it 
may be prolonged indefinitely. The metacasein stage may be pro- 
longed by any agency partially fixing the calcium salts, as by boiling, 
by adding small amounts of ammonium oxalate, etc. 

The addition of larger amounts of ammonium oxalate entirely pre- 
vents any heat coagulation. Hence soluble calcium salts are neces- 
sary for the metacasein reaction. If at the metacasein stage the 
rennin be destroyed by heating at 75° for five minutes, the addition 
of calcium chloride at 40° causes the separation of a coagulum. This 
shows that at the metacasein stage the caseinogen has been largely 
transformed into paracasein. Fresh milk cannot precipitate para- 
casein solutions, nor can it prevent the precipitation of paracasein by 
calcium chloride. From these facts, it would appear that the calcium 
salts in milk are altered in some way during the action of the rennin, 
and by virtue of this become capable of precipitating paracasein. 

It seems most probable that the calcium salts of the milk are very 
loosely combined with some constituent of the milk, and that these 
compounds are dissociated during the action of the rennin. 


NEW INDUCTORIUM, KYMOGRAPH, HEART LEVER, HEAVY 
MUSCLE LEVER, AND SQUARE RHEOCHORD. 


By W. T. PORTER. 
INDUCTORIUM. 


Tuis instrument (Fig. 1) is made entirely of hard rubber and metal. 
The primary coil, wound with double silk-covered wire of 0.82 mm. 
diameter, having a resistance of 0.5 ohms, is supported in a head- 


XXXVI Fifteenth Annual Meeting. 


piece bearing three binding posts and an automatic interrupter. 
The core consists of about ninety pieces of shellacked soft iron 
wire. This core actuates the automatic interrupter. The interrupter 
spring ends below in a collar with a set screw. By loosening the 
screw, the interrupter with its armature may be moved nearer to or 
farther from the magnetic core. Once set, the interrupter will begin 


to vibrate as soon as the primary circuit is made. The outer bind- 


ing posts are used for the tetanizing current. The left-hand outer 
post and the middle post are used when single induction currents 
are desired; they connect directly with the ends of the primary wire, 


FrGure 1.— The inductorium. 


thus excluding the interrupter. These several connections upon the 
head-piece are simply arranged and are all in view; there are no 
concealed wires. 

From the head-piece extend two parallel rods 22 cm. in length, 
between which slides the secondary cot/, containing 5000 turns of 
silk-covered wire 0.2 mm. in diameter. Over each layer of wire 
upon the secondary spool is placed a sheet of insulating paper. 
Each end of the secondary wire is fastened to a brass bar screwed to 
the ends of the hard rubber spool. 

The brass bars bear a trunnion which revolves in a split brass 
block, the friction of which is regulated by a screw. The trunnion 
block is cast in one piece with a tube 3 cm. in length, which slides 
upon the side rods. The secondary spool revolves between the side 
rods in a vertical plane. When the secondary coil has revolved 
through 90”, a pin upon the side bar of the secondary coil strikes 
against the trunnion block and prevents further movement in that 
direction, The right-hand side bar bears a half-circle graduated 


2 =< ee eee 


— = +" 


the clockwork, and is securely bolted to the 


Proceedings of the American Physiological Society. Xxxvil 


upon one side from 0° to go’. An index-pointer is fastened upon 
the trunnion block. One side rod is graduated in centimetres. 

The side rods end in the secondary binding posts, so that moving 
the secondary coil does not drag the electrodes. Next the binding 
posts is placed a substantial ‘“knife-switch”’ short-circuiting key, 
with hard rubber handle. 

The dimensions of the inductorium are as follows: Length, 24 cm., 
breadth, 7 cm., height of head-piece, 9 cm., total weight, 650 gms. 
Excepting the magnetic core, interrupter spring, and armature, the 
metal used is brass, heavily nickelled and polished. 


KYMOGRAPH. 


The kymograph (Figs. 2 and 3) consists of a drum revolved by 
clockwork and also arranged to be “spun” by hand. 

The drum is aluminium, cast in one piece, 
turned true in the lathe to a circumference of 
50cm. The height is 15.5 cm. The weight 
is 600 gms. The drum slides upon a brass 
sleeve in bearings 1.1 cm. deep (to prevent 
“side-lash”), and is held at any desired 
iene by a spring clip’ (Fig. 2). The 
sleeve ends in a friction plate, which rests 
upon a rubber-covered metal disk driven 
by the clockwork. Sleeve and friction plate 
revolve about a steel shaft which passes 
through both the heavy plates containing 


bottom plate. The sleeve bears upon the 
steel shaft only by means of ‘ bushings” at 
the ends of the sleeve, thus securing a 
bearing without “side-lash” and _ with 
little friction. As the sleeve with the 
drum rests upon the friction plate by 
gravity alone, it is easy to turn the drum 
by hand either forward or back, even 
while the clockwork is in action. This 
is a great convenience. At the top of 
the sleeve is a screw ending in a point which, when the screw is 
down, bears upon the end of the steel shaft and lifts the sleeve, 
and with it the drum, until the sleeve no longer bears upon the 


Figure 2.— The kymograph and 
its aluminium drum. 


XXXVI Fifteenth Annual Meeting. 


friction plate. The drum may then be 
‘“spun’”’ by hand about the steel shaft. 


The impulse given by the hand will ; 
produce from seven to ten revolutions 
of the drum. In the middle of the r 
series, the movement will be uniform t 
enough to give a fair record with a f 


tuning fork vibrating 100 times per 
second. : 

The clockwork consists of a stout 
spring about 6 metres in length, driving 

the chain of gears shown in Fig 3. : 
The speed is mainly determined 
by a fan slipped upon an ex- 
tension of the last pinion shaft | 


pi sn 1m ar in the chain. Four fans of 
: au different sizes are provided. 
cn {Thi 7 The speed is regulated by 
tif " ipa 
ea orifice f MM 2 governor consisting of two 


metal wings fastened to the 
same- shaft that carries the 
Ficure 3.— The kymograph, showing clock- fan. When the milled head 

work and steel shaft. shown in both figures to the 


Total 

Length. Height. | hours run Speed in mm. per second at 
Cm. Cm. upon one measured intervals. 

winding. 


None ore aye : ‘Ten-minute intervals : 


50.0 50.0 44.0 36.0 


Ten-minute intervals : 
38.0 38.0 36.0 30.0 


‘Ten-minute intervals : 

17.0 16.0 16.0 16.0 16.0 
16.0 15.0 15.0 15.0 14.0 
3:0" 20 


Half-hour intervals, beginning 
at 10th minute : 
8.0 8.0 $.0 8.0 8.0 
7.0 7.0 6.0 


One-hour intervals, beginning 
at 10th minute : 
5.8 S58 5:7 3g 
5 


Proceedings of the American Physiological Soctety. Xxxix 


right of the steel shaft is up, as in Fig. 3, the gear shown on the 
extreme right no longer engages with the gear driven by the spring, 
but runs “idle,” while the gear attached to the friction plate engages 
with the lower of the two gears shown at the left; the pinion of this 
lower left-hand gear engages with the spring gear. Fast speeds are 
then obtained as indicated in the table on page xxxvili. 

When the milled head is down, as in Fig. 2, the gear attached to 
the friction plate falls below the left-hand gear, while the right-hand 
gear engages with the spring gear and through a pinion drives the 
friction-plate-gear. Slow speeds are then obtained as follows : 


| 


Total 


. | ‘ . 
Length. | Height. | hours run Speed in mm. per second at 
Cm. em: upon one measured intervals. 
winding. 


Ten-minute intervais: 
2.6 Pas 2.3 


Halfhour intervals, beginning 
at 10th minute: 
1.0 1.0 0.9 0.8 


Half-hour intervals, beginning 
at 10th minute: 

0.51 0:50. 0:50 0.49 

0.48 0.47 0.45 0.43 


Half-hour intervals, beginning 
at 10th minute : 


Larger fans may be used. Thus, a fan 14 cm. long and 9 cm. 
high will give one revolution (50 cm.) per hour. 

In both figures, the brake, the winding lever, and stop-pin are 
shown upon the upper plate. 

Each shaft and its pinion is one piece turned from the same solid 
piece of steel. All the parts are highly polished. 


HEART LEVER. 
This very light lever 


(Fig. 4) is used in the 
suspension method of 
recording the contrac- 
tions of the heart, or for Ficure 4.— The heart lever. 


x] Fifteenth Annual Meeting. 


similar purposes. The axle is 7 mm. in length. The axle, with its 
aluminium wire 22 cm. long, and foil writing point 3 cm. long, 
weighs only 0.4 gram. All the parts except the lever are brass, 
heavily nickelled and polished. 


Heavy MuscLeE LEVER. 


It is sometimes necessary to afterload a muscle lever with weights | 


far in excess of those that a light muscle lever will bear without 
“springing” and thus altering the abscissa. Such 
heavy loads are borne by the heavy muscle lever 
illustrated in Fig. 5.. A cast-iron tripod, 27 cm. high 
and 17.5 cm. broad at the base, supports a femur 
clamp and a muscle lever. The latter is a steel tube 
5 cm. long,'pierced by a steel axle 9 mm. long, re- 
volving between heavy brass posts. The lever weighs 
2.5 gms. The aluminium scale-pan weighs 20 gms.; 
it holds one hundred 10-gram weights. The lever 
may be turned completely over in a backward. 
direction and thus be entirely out of the way. 
The steel spring shown upon the left of Fig. 5 
may then be turned to the right to bring its 
wire hook into the opening through which the 
scale-pan is reached. The scale-pan may then 
be attached to this zsometric spring and the 
spring empirically graduated. When the 
graduation scale has been written, the milled 
screw that holds the isometric spring upon 
the left-hand post (Fig. 5) may be loosened, 
the spring turned with the hook up, and the 
screw made fast again. The lower end of 
the muscle may now be attached to the hook 
upon the spring and an isometric curve 
written, 


FiGuURE 5.— The heavy 


The screw clamp holding the muscle 
clamp is insulated. A binding post upon 
the muscle clamp, and another binding post upon the right-hand 
post supporting the axle of the lever, allow direct stimulation of 
the muscle. 


muscle lever. 


ee 


he 


—— 


aS Se 


Proceedings of the American Physiological Society. xii 


SQUARE KHEOCORD. 


The rheocord, or potential divider, illustrated in Fig. 6, is a block 
of hard maple, 12.5 cm. square, upon which is placed a centimetre 
scale beginning at the O-post shown on the left side of the figure and 
ending at the 1-metre post visible in the background to the left. 
Along the scale, between these two posts, is stretched the first 
metre of a continuous German silver wire, 0.26 mm. in diameter and 
20 metres long. The remaining 19 metres of this wire are coiled 
upon a spool, and the free end is 
fastened to the 20-metre post 
shown in the background to the 
right of Fig. 6. The resistance in 
the 20 metres of German silver 
wire is so great (about 184 ohms) 
that the internal resistance of the 
element furnishing the electro- 
motive force, together with the 
resistancé of the large copper 
connecting wires, practically dis- FIGURE 6. — The square rheocord. 
appears for ordinary measurements. 

As the fall of potential is uniform throughout the 20 metres, the differ- 
ence of potential between post o and post 1 will be practically one- 
twentieth the electromotive force of the element. By moving the 
contact block from post 1 toward post o, any desired fraction of this 
one-twentieth may be secured. The under surface of the contact 
block is bevelled so that the metal touches the wire only with one 
edge; the opposite edge is supported by a piece of hard rubber. 

A flexible cable leads from the contact block to the binding post 

shown in the foreground to the right: 


DEMONSTRATION OF THE MOVEMENTS OF THE STOMACH AND INTESTINE OB- 
SERVED BY MEANS OF THE RONTGEN Rays. By W. B. CANNON. 

On THE NUCLEOPROTEIDS OF THE PANCREAS, THYMUS, AND SUPRARENAL 
GLAND, wirH ESPECIAL REFERENCE TO THEIR Optical Activity. By 
ARTHUR GAMGEE AND WALTER JONES. 

This journal, 1903, viii, p. 447. 


xlii Fifteenth Annual Meeting. 


ON THE ORIGIN OF THE RELATION OF THE INORGANIC ELEMENTS TO PRO- 
TOPLASM. By A. B. MAcaALLum. 


OsMoTIC CHANGES IN THE SACRAMENTO SALMON DURING THE RUN FROM 

THE SEA TO THE SPAWNING Beps. By C. W. GREENE. 
Read by title. 

AN IMPROVED APPARATUS FOR ILLUSTRATING THE ACTION OF THE DIAPHRAGM. 

By b. G. WILDER. 
By invitation. 

A LiviNG FROG FROM WHICH THE CEREBRUM WAS REMOVED DECEMBER 4, 

1899. By B. G. WILDER. 
By invitation. 

Quick METHODS FOR THE CRYSTALLIZATION OF THE OXYH&MOGLOBIN OF 
THE BLOOD OF THE DoG AND CERTAIN OTHER SPECIES. By E. T. 
REICHERT. 

INHIBITORY PHENOMENA IN THE CRYSTALLIZATION OF OXYHAMOGLOBIN. 
By E. T. REICHERT. 


THE VasoMOTOR INFLUENCE OF THE THIRD CERVICAL NERVE (AURICULARIS 
MAGNUS) UPON THE CIRCULATION IN THE Rapsir’s Ear. By S. J. 
MELTZER and CLaRA MELTZER. 

ON THE XANTHIN BASES OF THE SUPRARENAL GLAND. By W. JONEs. 

Read by title. 

THE INFLUENCE OF NEPHRECTOMY UPON ABSORPTION. By S. J. MELTZER 
and W. SaALant. 

ON THE Sources OF MuscuLaR ENERGY. By W. O. ATWATER. 


STUDIES ON THE INFLUENCE OF ARTIFICIAL RESPIRATION UPON STRYCHNINE 
SPASMS AND RESPIRATORY MOveMENTS. By S. J. MeL?zer and W. J. 
GIES. 

This journal, 1903, ix, p. I. 

THE Errect OF THE INTRAVENOUS INJECTION OF ADRENALIN UPON THE 
3LOOD-VESSELS OF THE EAR WHEN DEPRIVED OF THEIR VASO-CONSTRIC- 
TorS. By S. J. MELTZER and CLARA MELTZER. 


THE UNMISTAKABLE VASOMOTOR INFLUENCE OF SUBCUTANEOUS INJECTION 
OF ADRENALIN. By S. J. MELTzeR and CLARA MELTZER. 


With demonstration. 
ARTERIAL BLOOD-PRESSURE IN THE SACRAMENTO SALMON, ONCORHYNCHUS 
TSCHAWTSCHA. By C. W. GREENE. 
Read by title. 


Proceedings of the American Physiological Soctety.  x\iii 


* 
THE ACTION Or Hamotytic AGENTS aT 0°C. By G. N. Srewarrt. 
Read by title. 


DIFFERENCES OF POTENTIAL BETWEEN BLOOD AND SERUM AND BETWEEN 
NORMAL AND LakeD Bioop. By G. N. STEWART. 


Read by title. 
THE Laxinc oF Drriep Bioop-corpuscLtes. By G. N. Srewarr, for 


Dr. CLauDE C. GUTHRIE. 
This journal, 1903, viii, p. 441. 


THE 


nerican Journal of Physiology. 


VOL. VIII. OCTOBER 1, _ 1902. NOVEL 


SeetHeE. FUNCTIONS OF -THE CEREBRUM:. I.— THE 
FRONTAL LOBES IN RELATION TO THE PRODUC- 
TION AND RETENTION OF SIMPLE SENSORY- 
MOTOR HABITS.} 


By SHEPHERD IVORY FRANZ. 
[From the Physiological Laboratory of the Harvard Medical School. 


CONTENTS. 

Page 

Introduction RS lt ean: \ oy Beers etl marae A onl Ema. ] 
SOBEL . 2 og eg ERE eee = Sen en el cece nce Oe 5 
Surgical procedure a eles : 2 i eR 5 
Mneieteninination Of HAaDIES, ¢ Suao? soe se ce ees Bee te Se en 6 
Experimental . Bale SW AdiS aida ed FON ee Rae ed 10 
Retention of habits after extirpation of both frontallobes. . . . .. . . 10 
eetention of habwts after other lesions. ©... « « = «secs ms .s « 13 
Formation of associations after removal of both frontallobes .... . . 16 

PE ECIGHALODSCEVALIONSHs shake tare yon ci det mates? Be ela Ser Ae oi) PO) 
IEEE eta) eed uses elle Whar) Byes a Reps! lw mah Sal apele body ae (BE 


I. INTRODUCTION. 


HE present research was undertaken to determine, if possible, 

the relation of the various so-called association areas of the 

cerebrum to simple mental processes. The experiments to be re- 
ported in this paper are concerned mainly with the frontal area.? 


1 A preliminary account of this investigation was read at the Baltimore Meét- 
ing of the American Psychological Association, December, 1900. Psychological 
review, I9QOI, viii, p. 163. 

2 Later, papers will be published upon the parietal and temporal association 
areas and upon the motor and sensory zones. 


2 Shepherd Ivory Franz. 


It is known that the electrical stimulation of certain regions of the 
cerebrum produces no demonstrable results, and that these regions 
may be extirpated usually with little alteration in the sensory endow- 
ments of the animal and in its mental state. Thus the removal of 
the occipito-temporal regions or of the parietal association areas pro- 
duces comparatively no change in the animal’s behavior. The changes 
noted after the extirpation of the frontal lobes vary greatly accord- 
ing to different experimenters. According to some investigators the 
animals deprived of the frontal portions of the cerebrum are apathetic, 
stolid, or idiotic; others have found such animals ugly, irritable 
and restless. Occasionally all these terms have been used to 
describe the effect in one animal. These various descriptive words 
have been taken by some to mean a loss in ability to attend to the 
usual conditions of the mental stream, in other words, the loss of a 
supposed “faculty” of attention. Such a conclusion seems unjusti- 
fied, however, since it would separate the attention process from all 
mental states. In reality the attention is only a characteristic of 
mental conditions. 

For the better understanding of the present state of our knowledge 
regarding the functions of the frontal lobes I may cite the following 
accounts and conclusions. 

Loeb, who has recently written a general account of brain fune- 
tions,! says: “I have repeatedly removed both frontal lobes in dogs. 
It was impossible to notice the slightest difference in the mental 
functions of the dog. There is perhaps no operation which is so harm- 
less for a dog as the removal of the frontal lobes” (p. 275). In a 
discussion of the differences between the frontal and occipital por- 
tions of the cerebrum, he remarks, “ And if the anterior parts of both 
hemispheres be removed, the dog is no longer normal, but idiotic” 
(p. 263). ‘‘ While dogs after the loss of the anterior halves of the 
cerebral hemispheres often become irritable and ugly, dogs which lose 
the occipital halves of both hemispheres invariably become good- 
natured and harmless” (p. 264). 

In a careful review of the physiology of the cerebral cortex, Schafer 
writes as follows: ‘‘It is more easy to produce a condition of semi- 
idiocy in monkeys from extensive bilateral lesions of the temporal 
lobes than from complete severance of the prefrontal region, an oper- 
ation which may indeed be effected without producing any very 
obvious symptoms. . . . I have in several cases completely removed, 


' Loes: Comparative physiology of the brain and comparative psychology. 


On the Functions of the Cerebrum. 3 


in the monkey, the whole of the inexcitable areas of both frontal lobes 
without producing the slightest sign of the mental and intellectual 
dulness and alteration of character which has been regarded as 
pathognomonic of a lesion of this region.” ! 

The results of the early experiments made by Ferrier,? are mostly 
open to the objection that secondary softenings often followed the 
extirpations, and that usually a condition of sepsis was present. 
In addition the animals lived for only a very few days, and possibly 
had not sufficiently recovered from the primary shock effects of 
the operation. Ferrier’s later experiments with Yeo, however, 
are free from these objections, and the results may here be briefly 
mentioned. 

In a baboon the occipital lobes were extirpated. No impairment of 
vision seemed to follow this operation, and the animal to all intents 
was mentally normal. Six months after this operation the frontal 
lobes anterior to the pre-central sulcus were extirpated. ‘In less 
than an hour the animal began to move about, though in a somewhat 
sleepy and listless manner. . . . Only its manner seemed changed 
(7. e., after some days), and this was noted by all who had seen its 
former vivacity. It lost all its fun and trickiness, seemed not to 
know its name, took little or no interest in its companions, and 
was very easily cowed by them. Psychically only it had undergone 
appreciable change and degradation” (p. 481). This animal was 
observed three months before being killed. In a second monkey the 
whole convex surface of both frontal lobes was cauterized.. ‘‘ No 
physiological defect could be discovered, nor could any very defi- 
nite alteration in the animal’s behavior be determined. _ It. seemed 
only less timid of its companions. . . . Till its death by chloroform, 
eleven weeks later, it continued in perfect health and exhibited no 
perceptible deviation from the normal” (p. 525). The bases of both 
frontal lobes in another monkey were cauterized. On the third day a 
dreamy or drowsy condition was maintained, and on the seventh day 
the animal was still dull, taking no interest in anything but its food. 
“Except for general dulness and want of interest in its surround- 
ings, the animal exhibited no perceptible effect of the operation 
and continued in excellent health” (p. 528). Seven weeks after 


1 Text-book of physiology, edited by E. A. SCHAFER. Article on Cerebral 
Cortex by the editor. 

2 FERRIER: Philosophical transactions of the Royal Society, 1875, clxv, p. 433- 

3 FERRIER and YEO: /é7d, 1884, clxxv, p. 479. 


4 Shepherd Lvory Franz. 


this first operation both frontal lobes were wholly excised and the 
same general dulness was noted. ‘‘ Apart from a degree of dulness 
or apathy —and this as time went on not particularly noticeable — 
there was nothing in the animal’s behavior at all remarkable or ap- 
preciably abnormal.” The authors draw the following general con- 
clusions from the results of their experiments: ‘‘ As to the psychical 
effects of the frontal lesions it is difficult to speak at all definitely. 
In some cases there was no marked change, yet in others .. . there 
was a manifest alteration in the character of the animal. On the 
whole there seemed mental deterioration, characterized by general 
apathetic indifference or purposeless unrest, effects which, in com- 
parison with those of other lesions, appear to have relation with 
lesions of the frontal lobes as such” (p. 531). 

Hitzig! and later Horsley and Schafer? have given more definite 
facts regarding the mental condition of animals after frontal lesions. 
The first investigator had a dog which had been accustomed to 
find its food upon a table in the room and to take the food there- 
from. In psychological terms we may say there had been formed in 
the brain of the dog the association ‘“ table — food.”* The sight of 
the table and possibly the smell of both table and food would produce 
a nervous impulse resulting eventually in a complex motor response 
of going to the table, getting up to the food, and taking it. In this 
dog after the frontal lobes had been extirpated the habit was lost. 
Even after all primary effects of shock had had time to pass away the 
animal did not associate (or so it seemed) with the complex of sensa- 
tions which we call table the appropriate muscular movements. In 
fact the association of table — food was lost. A record of a similar 
observation upon a monkey has been recorded for us by Horsley and 
Schafer, but is negative in this respect. These experimenters had a 
monkey which exhibited in its normal condition certain well-defined 
simple habits, or as the authors call them “tricks.” After both 


* Quoted by FERRIER: Functions of the brain, 2d ed., 1886. H1Tz16’s orig- 
inal article is not accessible at the time of writing this account. 

* HorsLey and SCHAFER: Philosophical transactions of the Royal Society, 
1888, clxxix, p. I. 

* The writer does not wish to enter into a discussion of the mental states of 
animals, and the consideration of association of ideas, etc., in animals is left to 
the comparative psychologist. For the most conservative accounts of the mental 
life of animals the reader is referrred to the various writings of LLoyD MORGAN 
and ‘THORNDIKE. 


in i tt a 


On the Functions of the Cerebrum. 5 


frontal lobes were removed, the monkey still retained the habits. 
The tricks were shown as well after as before the operation. 

The important and significant fact that the reader should bear in 
mind is that in these two cases and in the case of Ferrier and Yeo 
mentioned above (case of the baboon which forgot its name, p. 3) 
we have a very definite mental state, in each animal a particular asso- 
ciation which in two cases was lost and in the other was retained after 
removal of the frontal lobes. It will be well to note that in none of 
the cases have we information of how long before the operation the 
associations were formed. The lack of this detail, I believe it will 
be shown later, makes the experiments comparable to only a slight 
extent. In a subsequent portion of this article it will be noted that 
the duration of an association should perhaps be considered of prime 
importance. 

The experiments which form the basis of the present article are 
similar to those of the cases just considered. My endeavor has been 
to determine whether or not simple sensory motor habits were 
affected by removal of the frontal lobes. In certain ways the research 
may be considered an extension of Hitzig’s experiment, although it 
was undertaken and fully planned before Ferrier’s account came to 
my notice. 


METHODs. 


Surgical procedure: — A general anesthetic (ether, ether and alco- 
hol, or the alcohol-chloroform-ether mixture) was used in all opera- 
tions and continued in every case until the animal was ready to be 
taken from the operating table. In some cases, in addition, there 
was given by mouth a quantity of urethane, chloral hydrate, or 
chloretone sufficient to keep the animal quiet after the operation for 
a sufficiently long time to prevent unnecessary movements which 
might cause hemorrhage. 

In all operations the usual aseptic precautions were taken. Before 
the scalp incision was made the hair was cut and the scalp thor- 
oughly washed with a solution of 1: 500 bichloride of mercury or 
with 5 per cent carbolic acid. In only one animal was pus found, 
and this was undoubtedly due to infection after the operation. This 
animal tore off the bandage and scratched the wound. However, 
although pus was found beneath the scalp, it had not gone into the 
brain. 


6 Shepherd Lvory Franz. 


.A-half-inch trephine was used to make the opening through the 
skull and. sometimes the opening was subsequently enlarged by 
means of bone. forceps. Generally, however, the button was taken 
from a point immediately over the portion of the brain which was to. 
be excised. After the button was taken’ out it was kept in warm 
saline and replaced before the scalp was stitched. This procedure 
in addition to a tight bandage kept an equal pressure on fe part of 
the brain exposed and helped to prevent hernia. 

After the button was lifted out the dura was cut to expose the cor- 
tex. Then with a fine cataract knife an incision was made at the 
point selected. When a frontal or an occipital lesion was made the 
knife was pushed down until it struck the base of the skull. With a 
gentle movement it was pulled outward, but toward the side so that 
the particular portion should be cut away from the main part of the 
hemispheres. The portion cut away was allowed to remain within 
thecranium, This method recommended by Schafer tends to prevent 
hernia of the fibrous portion of the cerebrum by keeping the cranial 
capacity normal. 

Bleeding from the diploé and from the brain was checked as s cintdl 
as possible by the application of artist’s wax to the diploé and of 
compresses to the cerebrum. In most 
2 Panonpnit leat cases the head bandage was sewed to- 
ed gether to prevent its being torn away 

by the animal. 

The accompanying diagram of the 
cat’s brain shows the location of the sen- 
sory and motor functions. The frontal 
Ficur«' 1.—A, superior, and B, Tegion, I consider as that portion an- 

lateral aspects of the brain of terior to the crucial sulcus. By pre- 

a cat, Adapted from Wieder frontal would be meant the jonas 

and Gage, and:trom: Berichte gceiie supraorbital fissure. 

The determination of habits. — The method employed for the produc- 
tion of simple associations is that used by Lloyd Morgan! and more 
extensively by Thorndike* and other writers. Briefly described, the 
method is as follows: An animal is placed in a certain environment, 
usually unpleasant or indifferent to the animal. By a simple motor 


VISUAL AREA 
TT 


C. LtoypD MorGAN: Introduction to comparative psychology; Habit and 
instinct; Animal life and intelligence. 
* E. L. THORNDIKE: Psychological review, 1898, Supplement No. 8; dd, 
Igor, Supplement No. 1S 


On the Functions of the Cerebrum. | 


adjustment the cat or the dog or the monkey gets a subsequent 
pleasure. 

At first the motor response is neither accurate nor speedy. At 
the beginning of any series of experiments the animal makes random 
movements according to its nature, scratching, pulling, biting, and 
feeling everywhere. By chance it hits upon the proper movement, 
and after the experiment has been repeated a number of times the 
unnecessary movements are eliminated. Only those movements are 
retained which tend to make the pleasure come soon, and conse- 
quently the time in which the act is performed gradually decreases. 
Usually after twenty or thirty trials the response is received immedi- 
ately after the animal-is placed in the desired environment. 

The activity of the animal is well described by Thorndike: “ When 
put into the box the cat would show evident signs of discomfort and 
of an impulse to escape from confinement. It tries to squeeze 
through any opening; it claws and bites at the bars or wires; it 
thrusts its paws through any opening and claws at everything it 
reaches; it continues its efforts when it strikes anything loose or 
shaky; it may claw at things within the box. It does not pay much 
attention to the food outside, but seems simply to strive instinctively 
to escape from confinement. The vigor with which it struggles is 
extraordinary. For eight or ten minutes it will claw and bite and 
squeeze incessantly. . . . The cat that is clawing all over the box in 
her impulsive struggle will probably claw the string or loop or button 
so as to open the door. And gradually all the other non-successful 
impulses will be stamped out, and the particular impulse leading to 
the successful act will be stamped in by the resulting pleasure, until, 
after many trials, the cat will, when put in the box, immediately claw 
the button or loop in a definite way.” ! 

In the present research only cats were used, and the environments 
were similar to those used by Thorndike, viz.: boxes from which 
the animal could escape by pulling a string, pressing a button, etc. 
Escape from the box always meant “ food.” The boxes were about 
twelve inches high, fifteen inches deep, and twenty inches long. An 
illustration of one of these is shown in Fig. 2, p. 8. The bottom, 
back, and sides of the boxes were solid boards, the front and top 
were made by placing three-quarter inch slats about an inch 
apart except at the door. The door was about six inches wide and 
eight inches high, hinged at the bottom, and so arranged that it 


1 THORNDIKE: Animal intelligence, p. 13. 


8 Shepherd Lvory Franz. 


would fall outward when the button was properly pressed, or the 
cord suitably pulled. In addition to the box here illustrated, two 
other boxes were used. Both of these were manipulated by a cord 
attached toa bolt at the top of the door. In one the cord passed 
upward and over the top of the box to the back where it was 

fastened. In this box at 
een the place where the cord 
passed over the box a 
wire netting was arranged 
about three-quarters of an 
inch above the cord. The 
door would fly open if the 
cord was pulled by the paw 
or if the head was strongly 
brushed against it. The 


FicurE 2.— Box from which animal could escape cord on the other box 
by pressing the button down or up. This box 
is hereafter called dztton. 


passed upward and then 
downward and outward to 
a point about three inches to one side of the door and about two 
inches in front of the slats. The animal could escape from this box 
by strongly pulling or pressing against the cord. In the remainder 
of this article these boxes will be called respectively, Button, String 
above and String front} 

The cats were placed in the boxes when hungry, food was placed 
outside and each cat was watched until it had opened the door 
and had obtained the milk or meat or fish. A record was kept of 
the time of escape at each trial,-—-and the speed together with the 
elimination of random movements determined for me when the habit 
was learned. 

The retention of the association under any set of conditions may 
be learned by placing the animal in the box under such circum- 
stances. If the time remain the same as when the original experi- 
ments were made, we may be warranted in saying that the habit is 
retained, and if the time be somewhat longer or the motor response 
not so accurate we may obtain a fair idea of how well or how ill the 
habit has persisted. 

The accompanying table and figure (Fig. 3) will illustrate the 


' For many other similar arrangements the reader is referred to the mono- 


graphs of THORNDIKE, where a full discussion of results with such environments 
will be found. 


Ss. 


On the Functions of the Cerebrum. 9 


method and at the same time indicate the retentiveness of a cat for 


the simple habit Stvzng front. 


TABLE I. 
String front. Female cat about six months old. Active. 
Feb. 5, 3.00 P.M. Feb. 19, 4.50 P. M. March 25, 3.50 P.M. 
8 secs. 5 secs. 2 secs. 
Borers ey yoke 
aes 3 ae Dae 
6°“ are 7 te 
200 ‘“ 6 iB gs 
LOK * Feb. 22, 3.10 P. M. April 10, 3.10 p.m. 
ES S secs. 2 secs. 
Feb. 7, 1.23 P. M. oe ae 
24 secs. ath a rusk 
25. « 5 « Q « 
LW pe Ki as 2 
an March 3, 3.15 P.M. 
hares 2 secs. 
20 oe 37s 
Ss Bats 
GS AMEE 
her 30 
6 ae 


Applying similar tests in the 
present investigation, if there be no 
alteration in the mental condition 
of an animal subsequent to the 
removal of the frontal lobes, the 
time of escape should be as short 
after the operation as it was before. 
The “memory” curve should be the 
same as in a normal cat. If, on 
the other hand, there has been a 
psychical disturbance, loss. of 
memory may be found, ranging from 
a complete failure to a slight slowing 
in the time of performance. 


Trea! Y\ Ld men | al 
4 We ARC: 


—— ——— 


. tO. 


Figure 3.— Curve of acquirement 
and retention of simple association.! 


1 Numerous similar curves will be found in THORNDIKE’s Animal intelligence, 
pp. 18-26. He has found that in many of his animals — cats, dogs, chicks — there 
is almost perfect memory, or to speak more accurately, perfect retentiveness of 


habits fer periods as long as seventy days. 


10 Shepherd Llvory Franz. 


EXPERIMENTAL. 


Retention of habits after extirpation of both frontal lobes.—In 
Hitzig’s experiment! we have seen that the dog could not find its 
food upon the table after the removal of the frontal areas. In terms 
of physiological psychology we may say that the association of table 
—food with the appropriate motor response had been lost. A simi- 
lar result was obtained from all the cats operated upon by me in the 
same manner. The following cases will, I think, be sufficient to 
establish the fact. 


I. Cat 4. Very active, female, four to five months old. 

Oct. 8. Experiments begun, Button and String above. 

Oct. 15. These habits had been thoroughly learned, so that the cat escaped 
from either box in times of one, two, and three seconds. 

Oct. 17. A.C. E. mixture. Both frontal lobes excised. Later in the day the 
cat seemed in good condition, but on the following day it seemed dazed 
and would not eat. However, it followed me about the animal room 
when called (this habit, it will be noted, was not lost). When returned 
to its cage it appeared restless. 

Oct. 19. Two days after the operation, the animal seemed in much better con- 
dition. It ate well and nothing unusual was noted about it. When tested 
regarding its retentiveness of the habits previously formed, it appeared 
that these were lost. 

Oct. 20-25. During the following six days the animal was tested in both 
boxes, but there appeared no evidence of retentiveness. At this time 
observers said the cat appeared mentally the same as a normal animal. 

Oct. 26. Unfortunately, signs of distemper appeared, and the animal was 
killed to prevent contagion. | 

During the progress of the experiments, after the operation it was noticed that 
when put in the boxes the cat would put its nose or its paws through the 
slats, and it would scratch around just as it had when learning the habit. 
‘To all who observed the cat at this time it seemed that no memory of the 
habits was present. It is interesting to note, however, that on Oct. 23, 
six days after the operation, the cat succeeded in escaping from the String 
above box by crawling out of the top. In all the boxes a top slat was 
left unfastened as a convenient method of introducing the animal into 
the box. In this case the cat accidentally pushed aside the slat, and when 
this was noted immediately crawled out of the box. In the experiments 
before the operation this slat was kept down by a brick or by the experi- 
menter’s foot, so that no memory can be said to have been present of such 


1 See above, p. 4. 


: 


On the Functions of the Cerebrum. IT 


association. Accidentally the slat was pushed aside by the head or back, 
and it was some time before notice was taken of it and escape effected. 

Another observation of interest is in regard to emotional states. In the animal 
room there was a cage of mice. Immediately before the operation every 
cat was held in front of this cage and its reactions noted. In this cat, as 
in the others, the same sort of response was obtained after the operation 
as before it. The cat’s heart beat more rapidly, its eyes followed the 
movements of the running mice, and it repeatedly tried to jump from my 
arms upon the cage in order to obtain its prey. This may also be con- 
sidered as another example of the retention of a long-standing or of an 
inherited habit. 

On post-mortem examination the brain was found normal except where the 
lesion had been made. The place where the cut had been made was 
well within the frontal areas. No hernia of the brain could be noted. 


This case is typical of all the other experiments upon the frontal 
lobes, and it is important that the reader keep in mind the following 
three facts: (a) the recently formed habits of opening the box by 
the string appear to be lost after extirpation of the frontal lobes; 
(b) the impulse (or we may say the inherited habit) to escape from 
an enclosed space remained perfect under these conditions; (c) the 
habit of coming at call—an association of comparatively long dura- 
tion — was retained. 


II. Cat 5. Active female, six to seven months old. 

Oct. 8. Experiments begun, String front and String above. 

Oct. 15. Average time for escape from these boxes, two and one half and 
six seconds respectively. 

Oct. 16. Extirpation of both frontals. Ether + 0.75 gm. chloral hydrate by 
mouth. For three days after the operation the cat was drowsy and apa- 
thetic. When taken from the cage it would walk a few steps and then 
stop. It seemed to have no objection to wetting its paws. 

Oct. 19. On this day it appeared quite like a normal cat, but when put in the 
boxes to determine memory, it showed no signs of retentiveness. Its 
activities were similar to those of a cat that had not been in the boxes 
before. After this date, and until the cat was killed, Oct. 28, nothing 
unusual was noted in the behavior of the animal, except that it did not 
retain the habits. 


1 My friend, Dr. H. N. KinGSForD, has kindly consented to investigate and 
report upon the pathological findings in all the brains, in respect to both gross 
and microscopical changes. In this article, accordingly, only the changes appar- 
ent to the eye will be noted. 


I2 


Shepherd Lvory Franz. 


My note of Oct. 20 in regard to String front is of interest: “ In a general 


Ill. 


way the cat seems to remember the place where its paw should be put to 
get at the cord, but the paw is placed only between the slats, not through 
them.” By “remember” I here meant that there were more impulses at 
or near the appropriate space through which the cat should have put its 
paw. ‘To open the door, it will be remembered, it was necessary to push 
the paw about two inches through the bars. 


Cat 9. Active female, about eight months old. 


Nov. 1. Experiments were begun, Sting front and String above. 


Nov 


. 16. Both habits were fully formed ; the times for escape average three 


and five seconds respectively. Normal memory was tested two weeks 
later, with averages of two and one half seconds for the String front and 
six seconds for the String above. Although the habit was perfect, the 
animal was further practised at intervals of three days until the operation. 


Dec. 10. Operation. Ether + 0.3 gm. chloretone. The animal had a good 


The 


IV. 


Nov. 


Jan. 


Jan. 


Jan. 


recovery after the extirpation and on the following days seemed quite 
normal. It followed me about the room, but when placed in the boxes 
the appropriate responses were not obtained. It moved around within 
the box, scratched in different places, put its paws between the bars, and 
several times had a paw in the proper place to pull the string. In 
these cases, however, it did not complete the motor process. The im- 
pulses to escape were practically the same as in a normal cat, but perhaps 
a trifle less active. Until it was killed, about two weeks after the oper- 
ation, it showed no signs of retentiveness of the habits. 

emotional effects before and after the operation were the same as in the 
preceding case (Case I). 


Cat 11. Active female, about nine or ten months old. 

15- Experiments were begun, Button and String above. These were 
perfect on Nov. 24, but the animal had developed distemper. From this 
it had recovered by Dec. 10, when retentiveness was found to be excellent 
(seventeen days’ interval). Its memory was tested again on Jan. 22 
(thirty-nine days’ interval); the habits were retained. 

22. Operation. Five hours after the operation the animal appeared 
normal except for a little drowsiness, occasioned, undoubtedly, by the 
0.5 gm. chloral which was given after the operation. 

23. When put on the floor it responded to my call, but could not escape 
from the boxes. Some slight inaccuracies of adjustment in walking were 
noted on this day, but these disappeared by the next day. They may 
have been due to the chloral. 

24-29. ‘The cat was placed in the boxes each day, but during this period, 
although the cat was active, no evidence of retention of habits was shown, 


On the Functions of the Cerebrum. 13 
With this cat and the others, if the animal did not escape from the box 
after two minutes the trial was considered a failure and no further attempt 
was made on that day. This was an arbitrary selection by me, but I think 
that it was justified. Such an interval of time enabled the animal to 
escape if the association was retained, but may not have given time for 
the cat to hit the button or string in its random movements. 

After Jan. 29 the cat slowly vedearned the habit. This matter will be more fully 
considered in a later portion of this article. 


At first sight it may seem proper to conclude from these and the 
other similar experiments that the frontal lobes are concerned in 
the production and the retention of simple sensory-motor habits. 
But a very valid objection may be urged against considering the 
matter settled by these experiments alone. To a critic all that the 
results would indicate is that habits are lost when the integrity of 
the brain is destroyed. Whether the loss is due to “‘ surgical shock”’ 
or to the destruction of that part of the brain concerned in the 
mechanism is not decided by the experiment. Such a criticism is 
equally potent against considering as conclusive the experiments of 
Hitzig, of Ferrier and Yeo, and of Horsley and Schafer. 

Retention of habits after other lesions. — Accordingly, the next step 
to be taken was the examination of the effects of “ surgical shock” and 
of the loss of other portions of the hemispheres. Let us first con- 
‘sider one typical case in which the operation was carried to the point 
of destruction of a part of the brain, but stopped there. In this ex- 
periment the animal was etherized as usual, the skull was trephined 
in two places, the dura was cut. The brain was not cut. Then the 
scalp was stitched together and the animal placed under observation. 


IV. Cat14. Active male, about nine months old. 

Dec. 1. Experiments were begun, Strzng front. 

Dec. 5. Habit well acquired. The average time for performance on this day 
was two and two thirds seconds. At 1.45 the operation was performed 
as described above. Fifteen minutes after the operation the cat walked 
about the room with fully codrdinated movements but seemingly a trifle 
“drunk” from the ether. No test of the memory was made on this day. 

Dec. 6. ‘Twenty-four hours after the operation the cat was placed in the box 
and gave the appropriate motor response as readily ason Dec. 5. Average 
time of ten trials, two and one half seconds. No change in the behavior 
of the animal was noted from the time of the operation until it was killed 
five weeks later. The post-mortem examination showed a normal brain, 
uninjured about the points at which the trephine buttons had been taken. 


14 Shepherd Ivory Franz. 


Other similar experiments were performed but in none were the 
habits lost. These are sufficient, I believe, to indicate that the loss 
of the habit when the frontals are excised is not due to the general 
shock effects of the operation, — ether, loss of blood, etc. Moreover, 
the animals were tested upon the day succeeding operation, when the 
effects of shock would be greatest. Even under these adverse condi- 
tions the associations were always retained. It would seem proper 
to conclude that an operation as such has little or no influence upon — 
the retentiveness of these simple habits. 

But, it may be urged, may not the loss of the cerebral ‘ahaa 
produce a greater nervous shock and thus explain the loss of the 
associations? The following four cases answer this question: 


V. Cat 17 and VI. Cat 29. Both active males, respectively about six and 

seven months old. Cat 29 was practised in Button and String above, 
and Cat 17 in String front and String above until the habits were perfect. 
Then both cats were operated upon ; the left parietal was excised. In the 
parietal region the area extirpated was between the motor zone and the 
visual area. 

Jan. 24. Operation on Cat 17. In the parietal region about ? sq. cm. was 
destroyed. For an hour after the operation the animal seemed weak, 
and for the remainder of the day was very restless. 

Jan. 27... Three days after the operation the cat was tested to determine re- 
tentiveness. On this day it escaped from S¢ring front in five, four, nine, 
nine, and eleven seconds, while before the operation it had opened the 
door in five, two, three, two and three seconds. With the String above 
on the day of the operation it escaped in ten, six, four, three and six 
seconds, and on the third day after, the times were three, two, three, 
thirty-five and five seconds. In this case it is of interest to note that 
the habit formed in String above was pulling with the paw. The same 
response was obtained after the operation. Most of the cats had learned 
to arch the back or to raise the head and brush against the cord to raise 
the latch. 

March 12. The whole of the right frontal lobe was excised. On March to 
the cat had escaped in an average of two seconds from String front. 

March 16. Average time of escape, eighteen seconds. 

March 1g. Average time of escape, thirty-two seconds. 

March 10. In String above escape was effected in 8 (H. P.),' 3 (N-), 5 (P-), 
4 (P.), and 20 (N.) seconds. 

March 16. ‘The times were 16 (H.), 22 (H.N.), 15 (N.), 8 (P.), and 16 (H.). 


' These letters represent the portion of the body used by the animal to open 
door: H.= head; P. paw; N.=neck. These were often used in combination. 


On the Functions of the Cerebrum. 15 


March 19. Escape in 6 (N.), 12 (P.), 22 (H. N.), 8 (N.), and 15 (H.). 

This cat, it will be noted, was slowly learning to change the motor adjustment 
from the paw to the head and neck, a procedure requiring less movement. 
The change is very marked after the operation. 

Cat 29 was operated upon ina similar manner. At first only the left parietal 
was excised, and later the right frontal. 

March 11. _ Excision of left parietal. 

March 17 and following days it gave normal times for both boxes, e. g., String 
above: March 11, two, four, five, two, and two seconds ; March 17, five, 
six, three, two, and two seconds. 

April 10. Extirpation of right frontal. After this date there was a lengthen- 
ing of the time similar to that found with Cat 17. 

April 17. Button, escape in thirty-five, seventy-five, thirty-four, six, and thirty 
seconds. String above in eight, six, five, three, and three seconds. 

April 24. Button, two, six, fourteen, ten, three, four, and two seconds. Séring 
above, three, six, four, two, and two seconds. 

During the days succeeding both operations the animals appeared well, no 
motor or sensory defects were apparent, and the cats were neither more 
nor less apathetic or listless than the cats not operated upon. 

A third operation was performed on each cat (left frontal) but both died before 
the effect could be determined. 


VII. Cat 23. Female, ten months old. String front and String above were 
learned and both parietals excised. Both associations were perfect before 
the operation and three days subsequent to it. Nothing of special interest 
was noted in this case. After the death of the cat the lesions were found 
to be about 8 mm. square. 


VIII. Cat 35. Male, five to six months old. This cat was somewhat sluggish 
in its movements, and the time of escape was usually much longer than 
that of other cats. 

By Dec. 15 it had learned String front, and on the following day the right 
frontal and the left occipital were excised. ‘There was a quick recovery, 
and on the day after the operation the times were nearly the same as 

upon the preceding day. 

A month later the left frontal was extirpated. Fifteen minutes after the oper- 
ation the cat walked well, but seemed dazed, probably from the ether. 
Periods of apathy were noted occasionally after the second operation, 
although none were evident after the first. Im general, however, the 
animal seemed intelligent and bright. ‘Tested in the box on Jan. 17 and 
18, the cat failed to open the door in five trials, although its activity 
‘was considerable.: Its movements were very much the same as in a Cat 
that had not been placed in the box before. ‘The movements that were 


16 Shepherd Lvory Franz. 


made were not especially directed to the bars through which the cord 
could be pulled, but were the random movements of a cat trying to 
escape from confinement. Beginning with Jan. 19 the cat slowly regained 
the habit. 


Formation of associations after removal of both frontal lobes. — If the 
frontal lobes are concerned in the production of these simple habits, 


it would be reasonable to suppose that such associations could not be 


formed after the removal of these portions of the hemispheres. The 
following examples indicate, however, that such is not the case. 


IX. Cat 6. Very active female, about four months old. This animal was 
operated upon before any habits had been formed. 

Oct. 2. Operation: extirpation of both frontals. 

Oct. 5. When allowed to roam about the room it appeared normal, and when 
held up in front of the cage of mice it showed well marked emotional 
reflexes. 

The experiments were not begun until two weeks after the operation. 


Oct. 16. Experiments begun, S¢vzzg front. In three days the habit was fully 
formed. ‘The cat was very active after the operation as well as before it, 
and showed no signs of restlessness or of apathy. From the actions of 
the animal it was impossible to determine that anything had been done 
to it. ‘Two weeks later signs of distemper appeared, and the animal was 
killed on Nov. 2. 

Post-mortem examination showed the brain lesion somewhat posterior to the 
crucial sulcus and very close to the motor area. No motor or sensory 
disturbances had been noted in the animal, however. Slight adhesions 
of the dura to the skull and of the dura to the brain were noted. 

The formation of the habit is represented graphically in the accompanying 
curve, Fig. 4. 


X. Cat 7. Active female, nine months old. No associations were formed 
before the operation. 

Oct. 11. Both frontals excised. Nothing unusual noted in the animal's be- 
havior after the operation. The cat was very lively during the succeeding 
days and showed no lack of general intelligence. Little emotional effect 
was noted, either before or after the operation, when the cat was held 
before the cage of mice. Fig. 4 gives a detailed account of the formation 
of the association Pufton. 

The cat was killed Nov. 2. Post-mortem examination showed that the frontals 
had been cut.away just anterior to the crucial sulcus. 


Of more interest are the two following cases in which associations 
formed before an operation were lost after both frontals were taken 


On the Functions of the Cerebrum. 17 


away, and then these same habits. relearned. In one case, Cat 11, 
two successive operations upon the frontal areas caused successive 
loss of habits. After the second operation and loss the animal again 
learned the trick. 


26. 26. 


FIGURE 4.— Curves of the formation of habits. A, Cat 6, String front; B, Cat 1, Button. 


VIII. Cat 35. (For an account of the behavior of this animal, see page 15.) 
The accompanying table and figure give better than a description the 
progress of the cat throughout the experiment. 


TABLE II. 
Cat 36. String front. 
Nov. 12, 9.00 a. M. Nov. 15, 2.00 P. M. Dec. 14, 10 A. M. 
410 secs. 40 secs. 5 secs. 
404 “ 1L5o fest 
168») “6 (eae Ce 
SORE 42) oy es 
Bato 14st San 
Nov. 13, 9.00 A. M. Nov. 16, 6.00 Pp. M. Dec. 14, 10.30 A.M. 
95 secs. 8 secs. Destruction of R. fron- 
1 ea (8 Pes tal and L. occipital. 
1S 5p hla Dec. 16, 4.00 P. M. 
132 * ares 12 secs. 
2 0) SoS Gas 
Nov. 14, 2.30 P.M. Nov. 23, 10.00 A. M. Sse 
130 secs. 3 secs. Sas 
Or a OFS 
Pets Sines Dec. 18, 2.00 P. M. 
Sy pes 4 secs. 
10 “c 4 “ 3 “ 
10°, = Dec. 1, 9.30 A. M. 6 * 
Loy 4 secs. 4 « 
1 Git op ts 
60) a Dec. 18-Jan. 14. 
25 yn oy Rae Practised on habit. 


18 Shepherd Ivory Franz. 


TABLE IL (contsnued) 


Jan. 14, 1.30 P.M. Jan. 26, 3.30 P.M. Jan. 30, 10.30 A.M. 


Extirpation of L. frontal. 34 secs. 5 secs. 
Jan. 19, 3.30 P. M. Secs ons 
Failure. Moye aes one 
Failure. 46 Sie oe 
Failure. 23m (fame. 
Failure. Jan. 27, 1.40 p.m. Feb. 4 11.00 a.m. 
Failure. 8 secs. 3 secs. 
Jan. 22, 2.30 P.M. 2 a Gree 
222 secs. ZS St ee: 
248 “ Bei: 2 ier 
180 16, % PEE 
130 Jan. 28, 3.00 P. M. Feb. 15, 4.15 p. M. 
35 6 secs. 4 secs. 
Jan. 23, 4.50 P. M NE oh 
95 secs. A 2 
128 po aa ets 
63 (jt Ga 
140 Feb. 27 
44 Died 


Ficure 5.— Acquirement and retention of habit by Cat 35. 


On the Functions of the Cerebrum. 19 


5 IV. Cat11. This animal was operated upon Jan. 22, after having learned 
Button. (For remarks, see page 12.) ‘This habit was lacking after the 
operation and was relearned. ‘Then on Feb. 9 a second operation was 

| performed immediately posterior to the first. For the ten days succeeding 
this operation there was no evidence of the persistence of the association. 
Two days after operation, when placed in the boxes, the cat walked up 
and down within, looking out between the bars but making no attempts 
to get out. On Feb. 14 this was continued. At times it stopped in its 

\ walk, looking steadily at some other cats in a near-by cage. It did not 
attempt to get at the button. It did not go passively into the box when 
I tried to put it in, as it used to do, but attempted to prevent my putting 
itin. Smell was present, but it is possible there was a slight anzesthesia 
of the fore-paws. Movements of walking were fully codrdinated. The 
cat did not stand on its hind legs with any degree of stability when food 
was held up for it. Normal response when held in front of the cage of 
mice. When in the box, and fish was placed outside, there was noticed 
a lack of attention to all but the food. On Feb. 22 the cat began to 
relearn the association ; but as it was practised only about once a week 
the process was comparatively slow. A full account is given in Table III 
and Fig. 6. 

Post-mortem examination showed that the first lesions were posterior to the 
supraorbital fissure and the second lesions immediately behind the crucial 
sulcus. 


RSENS R sae neae 
SAUCE oo 


emi ae aisha le IEE | 
Jc Re Neher) hae 
apa es ee ee 


FIGURE 6.— Cat 11 in Auton. Both operations were bilateral lesions in the frontal lobes. 


TABLE III. 
Nov. 15, 3.50 P.M. 12 secs. Nov. 16, 11.30 a. M. 
110 secs. a 6 secs. 
16" 92 « Q « 
Sie oF Mae tak: 1 
| nia hae See ayn 
aiteee ie 1 « 
Sib jie we i Fe 
4 « 26 . 4 « 


bib - 


20 Shepherd Lvory Franz. 


TABLE III (continued). 


Nov. 16, 11.30 A.M. Feb. 5, 2.00 P.M. March 19, 4.10 P.M 
2s 8 secs. 15 secs. 
A hes SMe , iD Ane 
vs apmagy 14s 
Nov. 19, 11 30 A.M. Ow abet, 9 be 
5 secs. Spee Orit 
Dre Feb. 8, 2.00 P. M. March 25, 3.40 P.M. 
ie se 4 secs. 10 secs. 
yee by tt oe 
BAe 14 2 
Nov. 22-Dec. 10. Oe yee 
Prials, nothing unusual. Sate TO} ss 
Dec. 14, 1.50 Pp. M. Feb. 9. April 17, 3.00 p.m. 
2 secs. Second operation. 14 secs. 
Dok Feb. 11-19. She 
Aan No evidence of habits. dh Wee 
1 as Feb. 22, 3.30 P. M. Joge 
Ha 70 secs. AS 
Jan. 10, 3.00 P.M. (ieee Apr. 24, 2.30 P.M. 
2 secs. BO iis 12 secs. 
20 95: 35 el le 
SS tan SS 5 a 
7 March 3, 2.20 P. M. oS Ba 
nRErE 120 secs. 2 eS 
Jan. 22. 12 Apr. 27, 1.50 Pp. M. 
Operation. OF, 20 secs. 
Jan. 24-28. Heys 2 a pode 
No evidence of habits. 3S ye Qo 
Jan. 29, 11.20 a. M. March 10, 11 30 P.M. 4s 
11 secs. 10 secs. 72 ie 
68.“ 140 Sept. 21. 
Cg 90; Died. 
Ointe 4505 
13.4 *6 oe eae: 
Jan. 30, 1.30 P.M. March 15, 3.50 Pp. M. 
22 secs. 75 secs. 
Fi Pe 100“ 
a Sats 
3 Owes 
Ae 60) 


No further discussion of these facts will be made at this time and 
the explanation is reserved for a future paper. 

Additional observations. — Emotional States. In the preceding 
pages I have recorded cases in which the emotional responses were 
the same after as before the removal of the frontal lobes. All the 
cats that showed emotional signs when held in front of the cage of 


On the Functions of the Cerebrum. ay 


mice before the operation were similarly affected after it. Several 
cats showed evidence of displeasure when handled somewhat 
roughly, and there was indication of feeling when unusual noises 
were produced. Al] the animals were as friendly after the operation 
as before it, —some probably more so. A few that tried repeatedly 
to escape from me before the operation, kept close to me afterward. 
Whenever I was among them after the operation, the animals would 
continually brush against my legs and purr as good-natured cats do. 
So far as these observations may be stated in general terms, I should 
say that the emotions remained practically the same after as before 
the operation, but that any change present was a tendency to greater 
friendliness. 

Nutrition. — Almost all the cats used died of distemper. Two had 
been affected with mange for some time before death. On post- 
mortem examination, all animals but one were greatly emaciated, 
although the appetite was usually good. The feces seemed normal, 
but no special chemical examination was made. The change in 
nutrition would lead us to believe that the frontal lobes may be 
concerned in that function. This result, I am aware, is opposed 
to Ferrier’s experience with monkeys.! This investigator found that 
there were greater changes in nutrition after lesions in the occipital 
lobes, and that the frontal lobes seemed to have nothing to do 
with this function. 


SUMMARY. 


1. Cats may be taught simple sensory motor associations in from 
two to five days. 

2. Memory for these habits persists normally for a period of from 
seven to eight weeks. 

3. After a bilateral lesion in the frontal lobes the habits are lost. 

4. This effect cannot be explained as due to “shock,” for other 
brain lesions are not followed by loss of habits. 

5. Unilateral lesions of the frontal areas are usually followed by 
a partial loss, or, rather, a slowing of the association process. 

6. Habits once lost after removal of the frontals may be relearned. 
After a second operation they are again lost, and may be regained a 
second time. 


1 FERRIER: Functions of the Brain, 1886, p. 194. 


22 Shepherd Lvory Franz. 


7. Only newly formed habits are lost after such lesions. Inherited 
and long-standing habits seem to be retained. 
8. The emotional condition of the animal is practically the same 


after as before the operation. . 
g. Emaciation and liability to disease were noted in all animals 


from which the frontal lobes were extirpated. 


STUDIES ON REACTIONS TO STIMULI IN UNICELLU- 
LAR ORGANISMS. IX.—ON THE BEHAVIOR OF 
FIXED INFUSORIA (STENTOR AND VORTICELLA), 
WItH SPECIAL REFERENCE, TO THE MODIFIA- 
BELLY GE PROTOZOAN REACTIONS? 


By H. S. JENNINGS. 


HE only infusorian whose behavior is at present known with 

any degree of fulness is Paramecium. This animal is a type 

of the free-swimming infusoria; while Paramecium does at times 

come to rest, as a rule it is found in rapid movement, — especially 
when under experimental conditions. 

The behavior of an animal which is fixed in a definite position 
will necessarily be of a different character from that shown by such 
an organism as Paramecium. Stentor and Vorticella furnish ex- 
amples of such animals, and, as will appear, their behavior differs 
much from that of Paramecium, — showing indeed a much higher 
development. 

As compared with an organism continually in motion, a fixed 
animal offers many advantages for the experimental study of be- 
havior, for one may keep the same individual continuously under 
observation, or return to it at longer or shorter intervals. It is thus 
possible to observe changes in behavior, and to determine whether 
the reaction to a given stimulus is modified by previous subjection 
to the same or different stimuli. 

The present paper is based upon a study of the behavior of the 
following organisms: Stentor roeselii Ehr., Stentor caeruleus Ehr., 
several species of Vorticella, Epistylis flavicans, var. procumbens, and 
Carchesium polypinum Lin. Stentor is much more favorable for 
such work than any of the Vorticellidze, and Stentor rceselii is in 
many respects the most favorable as well as the most interesting 
of the organisms studied. I shall therefore make an account of the 
behavior of this animal the basis of the paper, comparing the others 
with it. 

' Contributions from the Zodlogical Laboratory of the University of Michigan, 
No. 57. 


22 
ae?) 


24 #1. S. Jennings. 


oO 


—~<---. 
7 ee 


- So \ \\ \ 
ZA ~ \ 
2 Pe A \ 
/ oo Si CEN oe 
vf 7 \ \ \ \ 
eed ra \ 
73 7) \ \\ \ 
/ / NN \ 
/ / \. Ni ; 
ame Vie ah H 
/ / ’ oe \ 
/ } | \ u 
y bal 
: ce \ 
yen : 
fier 
| he 
| 1 


FiGuRE 1.— Stentor reeselii, showing the currents caused by the cilia. At @ a chemical 
is introduced a little to one side of the disc, showing that it does not reach the 
animal. At ¢achemical is introduced above the disc; it is carried directly to the 


mouth. At ¢ particles are seen passing along the ventral surface of the body to 
the edge of the tube. 


On the Behavior of fixed Infusoria. 25 


A. THE BEHAVIOR OF STENTOR RCESELII ERR. 


For understanding the behavior of an organism when subjected 
to stimuli, it is necessary to have well in mind the structure of the 
animal and its normal movements when unstimulated, —a considera- 
tion too often neglected in work on behavior. 

Stentor roeselii Ehr. (Fig. 1) is a colorless or whitish animal 
consisting, when fully extended, of a slender, tapering, stalk-like body, 
bearing at its larger end a broadly expanded disc. The surface of 
the body is covered with longitudinal rows of fine cilia, and bears 
also a considerable number of fine long setz, which disappear at 
times, and are said to be retractile and extensile. The 
disc is surrounded by a circlet of large compound cilia 
or membranella. These make a spiral turn, passing on 
the left side into the large buccal pouch, which leads to 
the mouth. The mouth thus lies nearly in the middle 
of what may be called the oral surface; this surface is 
considered ventral in determining right and left. The | et 
disc is covered with rows of fine cilia which are nearly o¢ Stentor 
parallel with the circle of membranellz. The smaller  reselii, after 
end of the tapering body is known as the foot; here the iat Gees 
internal protoplasm is exposed, ctrl 
sending out fine pseudopodia, by 
which the animal attaches itseif to objects 
(Fig. 2): 

The body contains, next to the surface, 
many fine contractile fibrille, the myonemes, 
through the action of which the animal may 
contract into a short oblong or conical form 
(Fig. 3). 

Be sicntor reselii, Stentor -rocselii is usually attached to a 
contracted into its tube. Water plant or a bit of débris by the foot, and 

the lower half of the body is surrounded 
by the so-called tube. This is a very irregular sheath, formed of 
flocculent material of all sorts, partly held together by a secretion 
from the Stentor. It is frequently nearly transparent, so as to be 
almost invisible. The manner in which the tube is formed will 
be described later. 


26 Hf. S. Jennings. 


Movements of cilia in the unstimulated animal.! — The membranelle 
and cilia of the oral disc are in continual motion in the extended 
animal; the nature of this motion is best seen by adding something 
to the water to make the currents induced by the cilia visible. 
When finely ground sepia or carmine is added to the water, the 
currents caused by the cilia are seen to be as follows: The mouth 


of the animal forms the bottom of a vortex, towards which the water 


above the disc descends from all sides (Fig. 1). Only the particles 
in the water near the axis of the vortex really strike the disc, — those 
a little to one side shoot by the edges without touching. These 
latter curve outward again after reaching a point below the disc, and 
thus a whirlpool is produced,—some of the particles returning 
upward so as to reach again the downward current at a point some 
distance from the Stentor. But most of the particles which thus 
miss the edge pass downward and out of the sphere influenced by 
the Stentor. 

Particles which reach the disc pass to the left, toward the buccal 
pouch, showing that the beat of the membranellaz has a component 
which drives to the left as well as downward, -—the real movement 
of the current being thus a left spiral. The particles thus reaching 
the buccal pouch are whirled about within it a few times, then they 
may take one of two courses. They either pass down into the mouth 
at the bottom of the pouch and thus into the internal protoplasm, 
or they are whirled out over the edge of the pouch, in the mid- 
ventral notch. In the latter case they usually pass backward toward 
the foot of the animal, along the mid-ventral line, as shown at c in 
Fig 1. Apparently the body cilia in this region keep up a backward 
current. The particles reach the edge of the tube, where they may 
cling, thus aiding to build up the tube. 

In determining whether certain given particles shall pass into the 
protoplasm or out over the edge of the disc, there seems to be no 
indication of sorting by the cilia and of choice, — though this would 
not be at all surprising in view of what we know of choice in Ameeba, 
and of corresponding phenomena in inorganic fluids (see Rhumbler, 
1898, or the brief resumé in Jennings, 1902). But in Stentor, as 
long as the disc remains extended, whenever particles of any sort are 
allowed to reach the disc in large numbers, some are taken into 


' It is doubtless to be held that the animal is never really unstimulated; the 
use of this term signifies merely that no special stimulus is acting on the animal, 
beyond what is supplied by the usual conditions of existence. 


On the Behavior of Fixed L[nfusoria. 27 


the protoplasm, while others pass over the edge and away, without 
regard to the nature of the particles (provided they are not too large; 
see p. 30). This is true, for example, of sepia, carmine grains, uni- 
cellular alge and débris of all sorts. When large numbers of 
minute unicellular alga pass into the buccal pouch, apparently the 
proportion taken into the protoplasm is the same as in the case of 
_ sepia or other non-nutritious particles. It seems evident that whether 
a given particle shall or shall not be taken into the internal proto- 
plasm depends upon the mechanical conditions governing the spiral 
currents in the pouch. Many of the particles in the current never 
reach the minute mouth at all, and these are whirled over the 
edge in continuation of their spiral course; those which are so 
situated in the vortex as to be carried directly to the mouth are 
taken in. Of course Stentor does exercise a sort of choice (as will 
appear below), by changing its position, reversing the ciliary current, 
or contracting when injurious substances are present in the water, 
but there is no indication of a sorting and selection of particles 
brought into the pouch by the usual currents. 

When stimulated, Stentor rceselii may contract into its tube (Fig. 
3). Such contractions do not as a rule take place except in response 
to well marked stimuli. This was the rule throughout my observa- 
tions, extending over many days. Undisturbed individuals observed 
without interruption for an hour or more did not contract at all 
during that time. In this respect Stentor differs from Vorticella, 
which contracts at short intervals, even when the conditions are 
apparently quite uniform (see Hodge and Aikins, 1895). 

Reactions to stimuli. I. Mechanical stimuli.— We will first con- 
sider what Stentor does when touched or struck by small objects, — 
its reactions to simple mechanical stimuli. Such stimuli are often 
received in the normal life of Stentor, and there is a surprisingly 
full and complicated set of reactions to them, as compared with the 
simple reactions of Parmecium. 

We will suppose that small solid bodies are brought with the 
water currents to the disc. This may be controlled experimentally 
by drawing a glass tube to a long, excessively fine capillary point, 
filling the tube with water containing finely ground sepia, and bring 
the point near the Stentor. What the animal does depends upon 
a number of different conditions. 

Normal movements continued. — At first the normal currents are 
not changed; the particles pass into the pouch, and some are taken 


28 HT, S. Jennings. 


into the internal protoplasm, while others pass out over the edge of 
the pouch at the mid-ventral notch, as described above. If the 
particles are minute, do not cling together into large masses, are not 
excessive in number, nor mingled with any stimulating chemical, 
the currents continue this normal course indefinitely. 
Bending toward the source of stimulus.—If a small object merely 
touches gently one edge of the disc, the Stentor may bend over 
toward the object 
; (Fig.4).) “Time 
a ——— reaction may be 
seen when a small 
organism comes 
against the disc 
of Stentor, then 
attempts to swim 
away. The Sten- 
tor bends in that 
direction, so as to 
keep in contact 
with the object as 
long as possible. 
In the culture 
dishes containing 
Stentors there 
were many free 
heads of Epis- 
tylis, and these 


FiGURE 4.— Stentor roeselii bending in the direction of a slight frequently swam 
mechanical stimulus. At 1] a bit of débris is allowed to touch ne 
the edge of the disc, and is then pulled to the right. The thus against the 
Stentor follows it, bending into the position shown at 2. Stentors, giving 


rise to the above 
described reaction. The Epistylis heads were not held at all, but 
were merely followed as far as possible by bending. 

This reaction may be produced experimentally by tickling the edge 
of the disc with the tip of a minute glass rod drawn out to the 
finest possible hair.1 The Stentor bends over toward the side 
touched, and if the rod is moved very gently to one side, follows it. 


* These and similar manipulations were carried out under the Braus-Driiner 


binocular microscope, the use of which renders very simple many experiments 
and observations which would otherwise be difficult. 


On the Behavior of Fixed Lnfusoria. 29 


If the rod trembles a little too much, the Stentor will contract 
suddenly, as described below. The most satisfactory way of pro- 
ducing the reaction is to get a bit of soft flocculent débris from the 
bottom of the dish to cling to the rod. This débris may then be 
allowed to come against the disc, and is then gently pulled to one 
side. The Stentor follows it, often bending far over. This experi- 
ment is represented in Fig. 4. The animal may thus bend in any 
direction, — to the right, to the left, or toward the oral or aboral side. 

Bending away.— If the stimulus is a little stronger, as one pro- 
duced by a large hard object, or by the objects becoming too numer- 
ous, as when a dense cloud of sepia reaches the disc, or when the 
objects are accompanied by a weak chemical stimulus, as is the case 
with carmine grains, then another reaction is produced. The animal 
bends away from its present position. This is thus to a certain 
degree the opposite of the reaction last described, but is not so pre- 
cisely localized a reaction as the former one. In this reaction the 
organism shows the influence of its spiral, unsymmetrical structure, 
in that, as in the case of Paramecium, it always turns toward a 
structurally defined side. The reaction in Stentor is as follows: the 
animal twists on its long axis one or two turns, then bends over 
toward the aboral side. It thus bends into a new position, but it 
does not always bend away from the source of stimulus; in some 
cases this reaction carries the animal toward the source. In the 
latter case the reaction is repeated. 

In most cases where the source of stimulus is not large, this 
reaction succeeds in removing the Stentor from its action. Thus, 
if a capillary tube containing sepia is held close to the disc, when 
the animal bends over toward the aboral side the particles of sepia no 
longer reach the disc, and the animal is relieved from the stimulus 
(Fig. 5). Much experimentation shows that this simple reaction 
is more effective in getting rid of stimuli of all sorts than might be 
anticipated. If the first reaction is not successful in accomplishing 
this end, it is repeated. 

Reversal of the ciliary current. — If the turning toward one side 
does not relieve the animal (or in some cases before this is tried), so 
that the particles continue to come in a dense cloud, the ciliary 
current is suddenly stopped and apparently reversed for an instant. 
The particles in the pouch or against the disc are thus thrown off. 
The reversal lasts but an instant, then the current is continued. If 
the particles still continue to come, the reversal is repeated two or 


30 ff. S. Jennings. 


three times in rapid succession. If this fails to relieve the animal 
of the stimulus, the next reaction (contraction) usually supervenes. 

Sometimes this reversal of the current takes place before the 
turning away described above, and it may be followed by that 
reaction, But usually the turning away occurs first. 


Figure 5.— Stentor roeselii bending away when a 
quantity of sepia or of some chemical reaches 
the disc. The animal bends toward the aboral 


side. 


The reversal is produced 
under various circum- 
stances. It occurs when 
a very large number of 
particles reach the disc at 
once, so that there is a 
tendency to clog the pouch, 
or when a large hard ob- 
ject, such as one of the 
loricate Ciliata, gets into 
the pouch. I have seen 
Coleps gotten rid of in this 
way. It occurs also when 
some chemical, as a weak 
salt solution, is mingled 
with the particles, or when 
the chemical alone reaches 
the disc (see the reactions 
to chemical stimuli, below). 

Contraction. — lf the 
animal does not succeed in 
getting rid of the stimulus 
in either of the ways above 
described, or if the stimulus 
is a very powerful one to 
begin with, the Stentor 
suddenly contracts. The 
body becomes short and 
club-shaped or oblong, and 
the Stentor disappears 
within its tube (Fig. 3). 


Here it usually remains twenty to thirty seconds, then rather 
slowly extends, so that from the moment of contraction to the 
moment of complete extension an interval of 40 to 50 seconds 


has usually elapsed, 


On the Behavior of fixed [nfusoria. 31 


) 


When, in extending, the body of the Stentor has become about 
half or two-thirds its original length, the ciliary disc begins to unfold 
and the cilia to act, causing the current to reach the disc as before. 
If with the current the stimulus again acts upon the animal (as when 
the sepia or the chemical is kept near), immediate recontraction 
follows. 

This may be repeated many times. To certain sorts of stimuli, as 
will be seen later, Stentor may get accustomed, so as to unfold and 
behave in the usual manner while the stimulus continues undi- 
minished. We will consider for the present the case where the 
continuance of the stimulus involves continued repetition of the 
reaction. ‘This case is realized when a dense cloud of carmine grains 
is kept where it will strike the Stentor as soon as it expands, or 
when various chemicals are kept in this position. In such a case 
the contractions are repeated, as above described, usually for a period 
of ten to fifteen minutes. Often the animal, after a number of con- 
tractions, remains within its tube a Jonger time than at first. But 
more often there is little change in the time of contraction until 
toward the end of the period of ten or fifteen minutes. If the stimulus 
continues, the next phase of the reaction now sets in, described in the 
following. 

Abandonment of the tube.— After the stimulus has been thus 
repeated at every unfolding of the Stentor for ten to fifteen minutes, 
the animal contracts violently several times, without intervening full 
extension. The short clavate body merely lengthens a little, then 
contracts suddenly and powerfully into a still shorter mass. This 
is repeated until the attachment of the foot of the Stentor at the 
bottom of the tube is broken, and the animal is free. It now leaves 
the tube and swims away. The animal may swim forward out of 
the anterior opening of the tube, but if this takes it into the sphere 
of operation of the stimulus, as will very often be the case, it may 
force its way backward through the substance of the tube, and thus 
gain the outside, swimming backward. It then swims away, to form 
a new tube elsewhere. 

Behavior while free.— While thus swimming through the water, 
after leaving its tube, Stentor takes on the characteristic behavior of 
the free-swimming infusoria, such as Paramecium. In the open water 
stimuli are almost lacking for the guidance of the animal, hence its 
behavior is, paradoxical as this may seem, much less free and varied 
than is that of the fixed infusorian, or the infusorian creeping on the 


32 H, S. Jennings. 


bottom; it becomes quite stereotyped. The writer has previously 
given (Jennings, 1899) an account of the main features in the be- 
havior of Stentor polymorphus when swimming in the open water. 
The behavior of Stentor rceeselii is essentially similar in character. 
It rotates to the left on its long axis as it swims, and at the same 
time it swerves toward one side, — apparently toward the right aboral 
side. Its path thus becomes a spiral, like that of Paramecium 
(for the significance of this spiral swimming, see Jennings, 190T). 
When the Stentor in its course comes into the region of a stimulating 
chemical or other stimulating agent, the animal swims backward 
a little, turns toward the right aboral side, and swims forward 
again. In all these respects its behavior is essentially like that 
of Paramecium, as described in the second of these studies (Jennings, 
1899a), so that it will not be described in detail here. At first after 
leaving the tube the Stentor is strongly contracted, of a very short 
oblong or club-shaped form. Usually as it swims it gradually extends 
a little, taking a long conical form, but remaining much shorter than 
the fixed specimen. The animal thus swims rapidly for some time 
about the vessel in which it is confined. It may be observed that 
the Stentor as it swims secretes over the posterior half of its body a 
transparent mucus or sticky substance of some sort, since carmine 
grains or other small particles in the water often cling to the pos- 
terior half of the body, or are trailed along some distance behind it, 
—the mucus evidently pulling out to form threads. 

On coming against the surface film or the smooth surface of the 
glass, the Stentor behaves in a peculiar way. The (only partly un- 
folded) disc is applied to the surface, and the animal creeps or spins 
rapidly over the surface, often revolving to the left; sometimes not 
revolving, and always progressing in the direction of the right aboral 
side or angle of the disc. 

On coming in contact with a bit of plant tissue or débris (consist- 
ing in the cases observed largely of worm-castings), the Stentor 
usually creeps rapidly over the débris, keeping the ventral surface 
against it. It thus follows all the irregularities of the surface, as 
rapidly and neatly as this would be done by one of the Hypotricha. 
This may continue for some time, the animal seeming to explore the 
object thoroughly; then it may leave the débris and swim about 
freely again for a period. At times the Stentor becomes attached 
to a piece of débris by the secreted mucus. This is drawn out to 
form a thread, often several times the length of the Stentor; by 


On the Behavior of Fixed [nfusoria. 33 


means of this thread the Stentor remains suspended in the water, 
as it were, whirling about on its long axis. It may thus remain 
partially attached for some time; then the thread is broken by a 
sharp contraction of the body, and the animal swims away. 
Formation of a new tube, and attachment of the foot. — Finally (in 
three cases that were timed, after fifteen to twenty minutes) the 
Stentor forms a new tube and attaches itself. This is doneas follows. 
The animal, coming to a small heap of débris, creeps over it with 
ventral surface against it, as above described, exploring it thoroughly. 
It becomes evident that mucus is being secreted over the surface of 
the posterior half of the body, since particles of débris stick to the 
body, or are trailed be- 
hind it. Finally, in a 
certain region, often be- 
tween two masses of 
débris, the animal ¥ 
begins to move _ back- 
ward and forward, - 
through a distance of 
only about three fourths 
of its own length (when 
contracted). This is 
kept up for about two 
minutes, and results in 


the formation of a short 

mucus sheath. from the F!GURe 6.— Illustrating the movement of Stentor in 
, . =] ° F 

forming a new tube. The animal oscillates between 


the positions 1 and 2, giving off mucus, which forms 
surface of the Stentor. the tube. a, mucus forming the tube; 4, debris. 


This process is_ illus- 

trated in Fig. 6. Now the foot is pressed against the débris at the 
posterior end of the sheath, where it adheres, — doubtless by the 
extrusion of pseudopodia, as illustrated in Fig. 2. Now the Stentor 
extends its body to the full length,—and we find it in the usual 
attached condition, with the lower half of the body surrounded by a 
transparent tube of mucus. 

The above account is drawn from observation of the process of 
settling down and formirg a tube in several specimens, and seems 
to be typical. In one case observed, however, the animal attached 
itself to the smooth surface of the glass, and this time the process 
differed. After wandering about for some time, as described above, 


secretion on the outer 


34 F. S. Jennings. 


the specimen applied its disc to the bottom of the vessel, and re- 
volved for some time on its long axis. Then it ceased revolving, and 
slowly bent its body till the foot reached the bottom, — the body 
becoming nearly straight again and tangential to the surface, before 
this was accomplished (Fig. 7). The foot attached itself to the 
bottom, then the disc was lifted up, and the body took a position 


FIGURE 7.— Illustrating the manner in which Stentor roeselii attaches itself to a 
smooth surface. The figures 1-6 represent the successive positions occupied by 
the Stentor. 


perpendicular to the surface. The animal was now attached in the 
usual way, though the beginning of the tube had not been made. 
The tube in such a case is formed later, automatically, as it were, 
by the secretion of mucus on the surface of the body. This becomes 
compacted and confined to the posterior half of the body by the 
contractions of the Stentor in responding to stimuli. 

Usually, however, the tube is formed before the attachment of the 
foot, in the manner first described. 

Having thus described the typical series of reactions when Stentor 
rceselii is subjected to mechanical stimuli, or to a combination of 
mechanical and chemical stimuli, we may return and consider the 
effect of some other stimuli, as well as a number of matters of a 
different character. 

II. Chemical stimuli, — Results essentially the same as those above 
described are obtained in stimulating Stentor roeselii by means of 


Ox the Behavior of Fixed [nfusoria. 35 


chemicals. But there are certain points which are of much impor- 
tance for understanding the method by which such organisms react 
to chemicals; these will be brought out here. 

When a chemical of sufficient strength to act asa stimulus, yet not 
strong enough to be destructive, is allowed to reach the disc of an 
attached Stentor, the same series of reactions is given as has been 
described above, — changing position, reversal of ciliary current, 
contraction, and final abandonment of the tube. These results were 
obtained with a weak solution of methyline blue; with the red fil- 
trate from carmine in water, with ;7, NaCl, with 5% HCl, and with 
#7 cane-sugar. In the latter case the effect was evidently due to the 
osmotic action of the sugar, as will be shown later. Other chemicals 
were not tried. 

After it was found that Stentor would bend directly toward the 
source of a weak mechanical stimulus, as described above, it was 
thought possible that an opportunity might be here presented for 
demonstration of positive or negative chemotropism, — a bending to 
or from the source of diffusion of a chemical. In other infusoria the 
writer has been unable to observe a direct turning toward or away 
from the source of diffusion of any chemical, so that this seemed an 
opportunity not to be missed. The experiments in this direction de- 
veloped certain facts which are of much significance for understanding 
the reactions not only of Stentor, but of other ciliates and flagellates, 
to chemicals. 

The attempt was made to localize very accurately the action of the 
stimulus, by the use of fine capillary tubes, bringing the chemical near 
to one side of the body,—so that it might affect one side alone. 
The Stentor might. then be expected to bend toward or away from 
the side affected. This involves no difficulty in manipulation, but an 
insuperable difficulty is at once met in the course of the currents pro- 
duced by the cilia of Stentor. Chemicals placed at one side do not 
reach the animal at all, as will be seen by an inspection of the course 
of the currents in Fig. 1. The chemical at a@ is carried past the 
animal without touching it. This is rendered evident when some 
colored chemical, such as a solution of methyline blue, is used. If 
the point of the tube is moved farther toward the front of the Stentor, 
the solution is involved in the central vortex and is carried directly 
to the buccal pouch and the mouth (as at 4, Fig. 1). 

Thus unilateral stimulation with a dissolved chemical, elsewhere 
than at the mouth, is practically impossible. This is true also when 


36 Hf. S. Jennings. 


the chemical in solution is advancing with a broad, plane front, as 
illustrated in Fig. 8. In such a case the solution does not reach the 
Stentor uniformly distributed, as determined solely by the move- 
ments of the ions. On the contrary, as soon as the advancing 


Ficure 8. — Illustrating the way in which an advancing chemical is drawn out in the 
alimentary vortex, so as to reach the mouth and disc of Stentor without affecting 
the remainder of the body. 


solution has arrived within a certain distance of the animal, a small 
cone of the substance is drawn out by the vortex, directly toward the 
disc of the animal. The point of this cone reaches the buccal pouch 
and mouth of the Stentor, long before the rest of the chemical has 
affected the animal. This is very clearly seen when colored chemi- 
cals are used. 

The result is that the animal always receives its stimulus from 
a chemical at a certain definite spot, —the mouth or buccal pouch, 
—while the rest of the chemical remains some distance away. It 
is obviously impossible for the animal to orient itself in accordance 
with the natural lines of direction of the diffusing ions. If the 
organism turns away from the side affected by the chemical, it 
will of course turn toward the aboral side, —that opposite the 
mouth, without regard to the original direction of the source of 


yee Ci a poe De ee 


aes 


' On the Behavior of Fixed I[nfusoria. 37 


diffusion of the chemical,—and this is exactly what the animal 
does. 

Parallel conditions exist in the other infusoria. In Paramecium, 
for example, a strong current, corresponding to that which reaches 
the buccal pouch in Stentor, passes along the oral groove to the 
mouth, the current over the rest of the body being slight in com- 
parison. When a colored solution is used, and a nearly or quite quiet 
Paramecium is 
found, it may 
be observed 
that an advanc- 
ing chemical 
behaves in 
much the same 
way as in 
prentor, A 
cone of the so- 
lution is drawn 
out opposite the 
anterior end of 
Dine Para- 
mecium, and 
passes down 
the oral groove 
to the mouth 
(Eis. (9)... The 
Paramecium 
recéives its 
stimulus from 
the chemical, 
therefore, on 
the oral side, 


FIGURE 9.— Paramecium, showing how an advancing chemical 
is drawn out by the alimentary vortex, so as to reach the 
—and responds, oral side without affecting the rest of the body. 


like Stentor, by 
turning toward the aboral side, — usually after swimming backward 
some distance. 

These examples show that we are not justified in expecting the 
ciliate infusoria in which similar conditions occur to orient themselves 
to the lines of direction of diffusing ions, as presupposed by some 
current theories of the reactions of organisms to chemicals. The 


38 Hl. S. Jennings. 


organisms are active, and determine for themselves where the stimu- 
lus shall first affect them. It is not at all surprising, therefore, that 
they have not been found thus to orient themselves. 

In the Flagellata, owing to the minute size of the body, it is 
impracticable to determine by experiment whether the conditions for 
stimulation are or are not the’same as those just described for cili- 
ates. But in view of what is known of the movements of the flagella 
in these organisms, with resultant formation of a vortex having its 
apex at the mouth,! together with the known asymmetry of. most 
flagellates, it can hardly be doubted that the conditions are practi- 
cally identical with those found in the ciliates. In this group, then, 
as in the Ciliata, we should not expect to find the organisms orient- 
ing themselves to the lines of diffusing ions; they do not permit the 
ions to follow alone the laws of diffusion, but actively intervene to 
determine the distribution of the substance in solution. These facts 
certainly deserve consideration in all work on the reactions of Ciliata 
and Flagellata to chemicals. 

Similar considerations apply to the reactions to other stimuli, in 
so far as the distribution of the agents concerned depends upon 
currents in the water. This would be the case, for example, with 
the reactions to heat and cold, in so far as the stimulation is due to 
differences in the temperature of the water in different regions. 
Figures 8 and 9 would serve equally well for the conditions when we 
have an advancing region of water which is warmer or colder than 
that about the infusorian. The warmer (or colder) water would be 
drawn out into a cone and then into a stream, which would affect 
only the oral side of the animal. It is therefore not surprising that 
we do not find a direct orientation produced by heat and cold in these 
animals, the so-called thermotaxis being brought about through the 
mediation of the ‘‘ motor reaction” (backing and turning toward the 
aboral side; see Jennings, 1899a, page 334). 

To radiant heat, to light, and to the electric current, these consid- 
erations, of course, do not apply, as the distribution of the stimulating 
agent in these cases is not affected by currents in the water. 

The fact that Stentor and Paramecium (as well, of course, as many 
other infusoria) are first stimulated by a chemical on the oral side, 
and that they respond by turning toward the opposite (aboral) side, 
seems to indicate that the reaction of these organisms is, primitively 


' See DELAGE et HEROUARD, 1896, pages 306-312, for a full account of the 
moveménts of the flagella and the formation of the alimentary vortex. 


On the Behavior of Fixed I[nfusoria. 39 


at least, truly a localized one. The reason why in reacting they 
always turn toward the same side would be merely because they are 
always stimulated on the same side (the opposite one). If this is true, 
we should expect them, if the stimulus were in some way made to 
affect the other (aboral) side, to turn toward the oral side, contrary 
to their usual habit. This may have been the original condition 
of affairs, and possibly infusoria may exist in which it is realized 
even at the present time. But that it is not true for most of the 
infusoria is shown by the reactions to localized mechanical stimuli, 
as described in the fifth of those studies (Jennings, 1900). It there 
appears that when ciliates are stimulated on the (unaccustomed) aboral 
or right side, they respond by turning toward that side, — exactly as 
when they are stimulated on the opposite side. The unilateral method 
of reaction has become strongly stamped upon the organisms, being 
indicated in the unsymmetrial form. 

III. Osmotic stimuli.— As in the case of Paramecium, sugar seems 
not to affect Stentor through its chemical qualities, but only through 
its osmotic action, so that opportunity is given for determining the 
nature of the reaction to changes in the osmotic pressure of the 
surrounding medium. #5 cane-sugar (about I per cent) caused no 
reaction whatever, though electrolytes of the same osmotic pressure 
caused a marked reaction,— showing the effect to be due to the 
chemical qualities, in the latter case. When Stentor was flooded 
with #4 cane-sugar, there was no reaction for seven or eight minutes. 
By this time the plasmolyzing effect of the solution was very evident; 
the animals had shrunk considerably. Now there was a sudden 
strong contraction, the animal remaining contracted several minutes. 
It then let go its hold and abandoned its tube, forcing its way back- 
ward out of the latter. 

Even with % sugar (about 34 per cent) the response was not 
immediate. The animal conducted itself normally for about twenty 
seconds after it was flooded with the solution. By this time shrink- 
age due to plasmolysis is very evident to the eye; the animal 
contracts and finally leaves the tube. 


B. OTHER FIXED INFUSORIA. 


In giving an account of the behavior of some other fixed infusoria, 
I shall confine myself largely to a comparison with Stentor rceselii, 
bringing out the resemblances and differences, and entering into 
details only in case of important differences or additional features. 


40 Hf. S. Jennings. 


STENTOR C/ERULEUS Emr. 


Stentor czrulus differs from S. roeselii in form and in its blue 
color, and it is usually larger, at least in this region. It does not 
inhabit a tube, and though frequently attached, it is much more 
inclined to a free life than is S. roeselii, so that it is often found 
swimming freely in large numbers. 

In an attached Stentor czruleus the ciliary currents are essentially 
like those of S. roeselii, and parallel statements may be made for 
both species as to the ingestion of food particles. 

Stentor czeruleus is much more sensitive than S. reeselii; otherwise 
its reactions to mechanical and chemical stimuli are of the same 
general character, though with several important points of difference. 

Stentor czeruleus does not usually bend over toward a solid object 
touching one side of the disc, as does S. reeselii. A large number 
of experiments on this point gave uniformly negative results. 

When the particles of solid substance which are brought against 
the disc by the water currents are too large, too numerous, or 
mingled with some chemical, the animal responds, as does S. roeselil, 
by twisting somewhat on its long axis, then bending toward the 


Ficurk 10 — Method by which Stentor czruleus often changes position when stimulated. 
‘The animal occupies first the position 1, then pushes backward into the position 2, 
with stalk bent, then straightens into position 3. 


right aboral side into a new position; by reversing the ciliary 
current for an instant, and repeating this; by contracting the body; 
and finally by breaking from the point of attachment and swimming 
away through the water. In these reactions there is little that is 
essentially different from the corresponding reactions of S. roeselii. 
A favorite method of changing the position when stimulated is shown 
in Fig. 10. The specimen backs strongly until the stalk is doubled 
near the foot. The animal then straightens the body into the position 


On the Behavior of fixed L[nfusorva. Al 


indicated by the part next to the foot,—thus at an angle of 180° 
with its previous one. 

Stentor czeruleus has recourse to the last step in the series of 
reactions, the abandonment of its place of attachment, — much 
more readily than does S. roeselii. In the latter species I was unable 
to force the animal to leave its place by mechanical shocks alone. 
Specimens were stimulated by striking the disc with a glass rod, 
sometimes for an hour continuously, yet the animals did not leave 
their place. Stentor ceruleus, on the other hand, will sometimes 
break its attachment the first time it is struck with the rod. There 
is much variation among different specimens on this point; usually 
it requires many such strokes to produce the result, and all the 
other reactions in the series are tried first, 

After leaving the place of attachment, Stentor czruleus swims 
through the water in a rather wide spiral, revolving to the left. The 
body is usually somewhat curved toward the oral side, and apparently 
as a consequence of this, the animal swerves continually toward the 
oral side. The tendency to deviation thus caused is corrected by 
the revolution on the long axis. When the free-swimming Stentor 
receives a mechanical or chemical stimulus, it swims backward a 
little, then turns. toward the right aboral side. The behavior of 
the free-swimming Stentor is thus essentially similar to that of 
Paramecium. | 

As the Stentor swims about well extended, frequently protoplasmic 
projections may be seen extending from the tip of the foot. These 
are viscid, so that bits of débris stick to them and are dragged about ; 
sometimes other infusoria, such as Paramecium, coming in contact 
with the foot, are thus dragged along. I have seen two Stentors 
become attached to each other through the accidental coming to- 
gether of the two posterior tips. If a small glass rod is placed 
against the tip of the foot, the Stentor may frequently be dragged 
backward by it, owing to adhesion. 

Often Stentor drags its foot over the bottom or over pieces of 
plant material. Sometimes it stops in such a position, and in a few 
seconds the foot is securely fast and the animal is anchored anew. 

Stentor czruleus, unlike S. roeselii, reacts to light. The reactions 
of this and some other ciliates to light will be treated in a separate 
paper. 


42 Hf. S. Jennings. 


VORTICELLA. 


Owing to its minute size, Vorticella is much less favorable for a 
study of behavior than is Stentor. I was especially desirous of in- 
vestigating it however in connection with certain statements made 
in a very interesting paper by Hodge and Aikins (1895). These 
authors investigated chiefly the question of the rhythmical character 
of the activities of Vorticella. They kept a single Vorticella under 
observation without a moment’s intermission for a period of twenty- 
one hours, besides intermittent study for a number of days. The 
observations showed “that a Vorticella works continuously and shows 
in its life no period of inactivity or rest, corresponding to periods of 
rest in higher animals. In other words, a Vorticella never sleeps.” 
During five days the cilia were in continuous motion, and food was 
continuously taken, 

Incidentally, Hodge and Aikins made a number of observations 
on other points. One of these was in regard to the modifiability of 
reactions in Vorticella. An attempt was made to feed the individual 
under observation upon yeast plants, by introducing some of a pure 
culture of these organisms into the preparation. ‘‘ This attempt 
resulted in an interesting demonstration of the educability of Vorti- 
cella. At first they took this, to them, newly discovered food with 
great avidity, filling their bodies to distention with food vacuoles of 
the yeast. In a very few minutes, however, the entire meal was 
ejected with volcanic energy. Not a single torula was allowed to 
remain in the body, and for several hours at least — how long the 
memory lasted was not determined — the individual could not be 
induced to repeat the experiment.” 

It is much to be regretted that further details are not given in 
regard to this interesting experiment. We are not told whether the 
Vorticella continued its normal behavior and took in other food 
during the time in which it refused the yeast. It might be that the 
animal was merely injured by the food, and took nothing more into 
its body until it had recovered. We are not informed in what way 
Vorticella refused later to take the yeast, — whether by contracting, 
by reversing the ciliary current and turning the yeast out of the 
pouch, or in some other way. Yet upon these points depends largely 
the interpretation that shall be given to the observation. 

I have endeavored, and as I judge with some success, to reproduce 


On the Behavior of Fixed Infusoria. 43 


od 


the essential features of this experiment. I have not succeeded with 
the yeast, for the Vorticella at once contracted, in my experiments, 
when the yeast culture was introduced. But a similar result may 
be obtained in a very simple way. In the case of the yeast culture 
we have a fluid containing various chemicals in solution, and holding 
in suspension many small bodies. These conditions may be imitated 
by grinding up ordinary carmine in water. A little of the carmine 
goes into solution, as may be shown by filtering the water, which will 
be found to have become red. This red solution was found to act 
as a slight chemical stimulus, on both Vorticella and Stentor. 

When some of this carmine and water is added to the water about 
the Vorticella, the course of events is about as follows. For a short 
time —ten to fifteen seconds, or sometimes more—the current 
caused by the cilia is kept up in the usual manner, and many of the 
carmine grains are taken into the internal protoplasm, forming red 
food vacuoles. Then there is a sudden contraction of the stalk, 
the ciliary disc closing at the same time. This is repeated several 
times, the ciliary disc however remaining closed while the stalk 
partly extends and recontracts. Then usually the Vorticella extends 
in a new direction. If the carmine continues to be present, the 
contractions are repeated for ten or fifteen minutes or more. Then 
the stalk may remain extended, but the ciliary disc remains closed, 
so that no more carmine is ingested. This condition lasts as long 
as the carmine is present in large quantities. 

Thus in this case, as in that described by Hodge and Aikins, 
Vorticella at first ingests a certain substance which it later refuses. 
This is also true of Stentor, as will be seen by consulting the account 
above given of the behavior of Stentor when much carmine is added 
to the water containing it. In no case was Vorticella observed to 
throw out the granules which it had already ingested, as described 
by Hodge and Aikins, but this is perhaps an unessential difference, 
as of course this has nothing to do with the “educability” of the 
animal. 

In this experiment, I am convinced that the refusal of the Vorti- 
cella to continue to take the substance is due to a too great stimula- 
tion, either in the quantity of material, or, more probably, in the 
strength of the chemical action of the substance, rather than to any 
precise choice in the kind of substance ingested. There is evidence 
for this in the following fact. If the quantity of carmine in the water 
is greatly decreased, so that only scattered grains are left, the 


A4 Hf. S. Jennings. 


Vorticella or Stentor no longer reacts to these, and they are ingested, 
gradually forming red vacuoles in the endosarc. Whether this would 
have been true in the experiment of Hodge and Aikins the data 
they have given do not enable us to judge. 

In regard to a point closely connected with the above, the attitude 
of Hodge and Aikins must, I think, be considered uncritical. This 
relates to the sorting of food by the cilia. Among the “ psycho- 
reflexes” of Vorticella, Hodge and Aikins include ‘3. Sorting of 
particles by the sensory cilia; the driving of food toward the mouth, 
and the driving away of waste particles.” Further ‘‘ When a particle 
is touched by the cilia an act of choice is apparent, and in accordance 
with this choice the particle is carried toward the mouth or whirled 
away.” ‘Particles scarcely visible under the microscope are sorted 
with the greatest apparent precision.” ‘ A prime condition of the 
creature’s life must be ability to distinguish food from that which 
is not food.” 

Beyond these general statements quoted no details are given. 
The authors report no critical observations or experiments as to 
what substances are ingested, what rejected in this sorting process. 
It is well known that investigators that have made such experiments 
have concluded that no such sorting takes place. Thus Verworn 
(1889, page 150) found that Vorticella ingests carmine grains, indigo, 
and chalk crystals, and I have myself observed the same facts. 
These are substances which cannot serve as food, so that Hodge 
and Aikins are certainly mistaken in their belief that “a prime 
condition of the creature’s life must be the ability to distinguish 
food from that which is not food.” Such organisms as Vorticella 
and Paramecium grow and multiply in situations such that the 
substance brought to the mouth by the currents consists largely of 
food, without any sorting; when this condition disappears, the 
organisms quickly die. The ingestion at the same time of some 
substances which do not serve as food is not particularly injurious 
to the organism, as these are simply passed out of the body with the 
waste matter at the time of defecation. 

The impression that a sorting and selection takes place among the 
particles brought to the mouth probably arose from the following 
observations. Vorticella, as well as Stentor, brings to the buccal 
pouch in its alimentary vortex many more particles than are taken 
into the body, When the water contains many particles a continuous 
stream of these may be seen passing out of the buccal pouch. From 


On the Behavior of Fixed L[nfusoria. 45 


this it is most natural to conclude that the material has been sorted, 
the valuable particles ingested, and the particles which are not 
nutritious turned away. But experiment does not support this con- 
clusion. Thus, when all of the particles are of the same sort, either 
nutritious or not, part are taken into the interior, while a large 
portion are turned away. This is true on the one hand of grains of 
sepia, which are quite insoluble and non-nutritious; it is true, on the 
other hand, also, when many nutritious unicellular alge are brought 
to the mouth. In the latter case, as in the former, many more of the 
particles are turned away than are taken in. The explanation of 
these facts is evident when one considers the mechanism of the 
alimentary vortex, as has already been pointed out by Verworn 
(1889). A large, strong current of water is carried toward the disc 
of the Vorticella. Inevitably, much of the water misses the disc 
completely, and the food particles which it contains never touch 
the animal. Another portion of the water strikes the disc, but 
not all of this can enter the relatively small buccal pouch, so that 
many of the food particles which strike the disc are whirled away 
again into the water. In the same way a rapid whirlpool is formed 
in the buccal pouch, but only a small part of the water in this can 
reach the relatively minute mouth, and it is only from this small 
part that the food can be taken. Thus there is a continual stream 
of particles passing out of the pouch, that have never come in contact 
with the mouth. 

These mechanical considerations explain also the following notice- 
able fact. When the water contains but few particles, whether 
nutritious or not, only a few are ingested, while a large proportion 
of them are whirled out of the pouch and away. When the number 
of particles in the water is great a large number are ingested ina 
short time, without regard to whether they are or are not useful 
as food. The number of particles ingested depends primarily upon 
the number which reach the mouth opening, and this is only a 
small proportion of those involved in the general ciliary vortex. 

This question of the sorting power of the cilia is, I take it, merely 
one of fact, and not one involving any important principle. What 
we know of choice in the Rhizopoda, and the parallel phenomena in 
inorganic fluids, to which reference has already been made, shows 
that there would be nothing new in principle if the cilia of Vorticella 
exercised choice in the same way. But the facts seem to indicate 
that they do not. Choice in these animals seems to be shown only 


46 Ff. S. Jennings. 


in such phenomena as the reversal of the ciliary motion, bending 
over into a new position, and contraction, — these being, of course, 
different methods, somewhat crude perhaps, of rejecting certain 
things, and thus of exercising choice. 


C. BECOMING ACCUSTOMED TO STIMULI. 


Are the reactions of such organisms invariable, or does the reaction 


toa given stimulus depend on previous subjection to the same or dif- 
ferent stimuli? The problem of the modifiability of the reactions of 
these lowest organisms is one of great interest, but one on which 
there exists but little precise experimental data. Scattered allusions 
to changeability in the reactions of the lower organisms are to be 
found in the literature, especially with relation to what might 
be called acclimatization to stimuli. Massart (1901, page 8) states 
that it is often to be observed that organisms which have reacted 
several times in succession, at short intervals, to a given stimulus, 
lose, little by little, the power of responding to this stimulus, but 
that this is doubtless to be attributed to fatigue. Loeb (1900, 
page 228) has discussed such a case in the reactions of worms toa 
shadow, as described by Nagel, and has attributed the lack of 
reaction when the stimulus is repeated to ‘‘a simple after effect 
of the stimulus, a case that we often meet with in the physiology 
of both animals and plants.” Davenport (1897, page 108) gives an 
example drawn from the behavior of one of the organisms at present 
under consideration. ‘When an organism has been stimulated by 
contact for some time, it at last becomes changed, so that it no longer 
responds as it did at first. Thus, Dr. W. E. Castle informs me that 
he has seen a colony of Stentors, in an aquarium, being constantly 
struck by Tubifex waving back and forth, yet the Stentors did not 
contract as they usually do when struck.” 

Such contractions in the fixed infusoria furnish a most favorable 
opportunity for an investigation of this matter, and I therefore 
undertook to obtain some precise experimental data upon the sub- 
ject. Experiments were made upon Stentor reeselii, S. czruleus, 
Vorticella, Epistylis, and Carchesium. 

First, the conditions described in the observation by Castle, above 
cited, were imitated, by striking the extended infusorian with a fine 
glass rod or hair, under the Braus-Driiner stereoscopic binocular. 
The chief difficulty in these experiments is to make the successive 


< 


—— 


On the Behavior of Fixed [nfusoria. 47 


strokes approximately equal in force. This can be done but very 
imperfectly; nevertheless the results are clear. 

The first stroke, whether light or heavy, given to an individual that 
has been undisturbed for an hour or more, almost invariably results 
in causing a quick contraction. This is true for all the organisms 
worked with. The animals remain contracted a minute or less, then 
slowly extend. At the instant when extension was complete, another 
stroke was given. This,and several successive strokes usually caused 
the same reaction as the first one. After ten or a dozen reactions, 
however, the organisms usually did not contract as soon as touched ; 
the stroke had to be repeated one or more times before reaction was 
caused. A typical series for Stentor czruleus is given in the follow- 
ing. The figures represent the number of strokes in each case before 
contraction took place, —a contraction occurring thus at each dash: 

I—I—I—I—I—I—I—2—2—I—2—I—2—4~—I—I—I—I—I—2—6—10—I— 
2—9Q—13—3—14—7—3—2—3—3—9—18— (at this point the Stentor 
pulled its foot loose and abandoned its place). 

As is evident from the above, there is much irregularity in the 
number of strokes required to cause contraction. This is due, 
partly at least, to the practical impossibility of giving successive 
strokes of equal force. But the Stentor responded at first to the 
lightest possible touches, while later it required a considerable num- 
ber of smart strokes to cause contraction. 

Sometimes there is a ready response only to the first touch, as in 
the following series (Stentor czeruleus) : 

I—22—25— (breaks away). 

I—1—40— (breaks away). 

In these cases the organism does not remain entirely oblivious to 
the blows, but after it has ceased to react by contracting it continu- 
ally changes its position, by twisting, then turning toward the aboral 
side, as if trying to escape from the blows. The final reaction, in 
Stentor czruleus, is to break away from its attachment and swim 
away. 

In Stentor reeselii, Vorticella, Epistylis, and Carchesium, similar 
results were obtained, save that these organisms never broke away 
from the attachment as a result of such mechanical stimuli. A 
typical series for an individual of Epistylis flavicans, var. procumbens, 
was as follows: 

I—I—I—I—I—I—I—2—33—25—7—1 3—30—20—1 4—I 3—I 3—33—9—30 
—3—31—226. 


48 Hf. S. Jennings. 


In another series the results were as follows: 

I—22—10—3—3—I—I—22—59—125— (continuous blows for 1 min.) 
— (¢ min.) — (td min.) — (45 min.). 

Some series show greater irregularity than the above. As in the 
case of Stentor, during the latter part of the experiment the Epistylis 
continually changed its position, as if trying to escape from the 
blows. 

A typical series for Stentor rceselii, obtained in this case by jarring 
with the rod the leaf to which the Stentor was attached, is as follows: 

I—I—I—I—I—I—I—I—I—I—I—I—3—I—5—I—I—3—I—3—3—48—40 
—2—250—36—36—1 54. 

The results for Vorticellaare similar. In a typical case the animal 
contracted after each of the first nine strokes. Then the contractions 
became less sudden; two or more strokes were required to produce 
them. After about twenty contractions the Vorticella could be 
tapped almost indefinitely without causing further contraction. 

Carchesium polypinum is a tree-like colony composed of many 
Vorticella-like individuals, attached to the branches of a common 
stalk. The stalk muscles of the individuals are not continuous 
throughout the colony, so that it is possible, though not usual, 
for each individual to contract separately. 

Carchesium shows very markedly the acclimatization to a stimulus. 
Observing first the reactions of a single individual that is repeatedly 
stimulated, it is found that its stalk contracts strongly at every stroke. 
But after about five minutes there is a marked change in the readiness 
to respond. Several strokes are required to cause contraction. Still 
later the stalk ceases to contract when the individual is struck, though 
for a time the peristome is folded inward and the ciliary motion ceases 
after every stroke, without contraction of the stalk. When thus con- 
tinuously stimulated, usually the stalk contracts at intervals of two or 
three minutes, — though the strokes come as often as one per second. 

The effect of the stimulation of a single individual on the colony as 
a whole is interesting If a single individual in an otherwise undis- 
turbed colony is struck with the glass rod, usually the entire colony 
contracts at once, forming an almost solid ball. Apparently the 
sharp contraction of a single individual, by jarring the colony, acts 
as a stimulus to cause the contraction of all the other individuals. 
If the stimulus is repeated (on the same individual) as soon as the 
colony has become extended, usually only about half of the colony 
contracts. The third time only the large branch reacts to which the 


- —_ eae 


ei . 


On the Behavior of fixed [nfusoria. 49 


individual stimulated belongs. After this the number of neighboring 
individuals contracting when the single individual reacts is variable, 
ranging usually from half a dozen to thirty or forty. When the 
condition is reached where the individual, continually stimulated, 
reacts but once in two or three minutes, nearly the entire colony 
contracts with it. 


What is the explanation of this failure to react to a stimulus to 
which the organism at first reacts readily? Three possibilities pre- 
sent themselves. (1) The lack of reaction might be due to fatigue 
of the contracting apparatus (corresponding to muscular fatigue in the 
higher animals). (2) It might be due to fatigue of the sensory 
function, so that the organism no longer perceives the stimulus 
(corresponding to fatigue of the sense organs in higher animals). 
(3) It is possible that the phenomenon cannot be explained as fa- 
tigue, so that all we can do is to formulate the facts, calling it an 
“‘after-effect,” or other name which carries no implication as to its 
nature. We should perhaps have parallel phenomena for this also in 
the case of a higher organism, which reacts to a sudden, unexpected 
shock, but does not react a second time, though the stimulus is 
repeated, and is perceived by the organism. 

Some farther data needed for forming an opinion as to which of 
these possibilities represents the truth may be obtained by varying: 
the experiments. Striking the animal with the glass hair is a rather 
brutal method of experimentation; reactions may be produced with 
much slighter stimuli, and the results are much clearer. 

For this purpose weak currents of water may be employed. This 
was done as follows: A tube 28 cm. long and of 5 mm. bore was 
drawn to a very fine capillary point and then filled with water. 
When the capillary end is below, there is of course a slight current 
of water from the tip, due to the pressure of the water in the tube 
above. Now the tip was brought close to an individual of Epistylis, 
so that the current flowed against the latter. At once the animal 
contracts. If the current is continued the Epistylis soon unfolds, 
and continues open and active in spite of the current. If now the 
tube is removed, so that the current no longer acts, then in a few 
seconds is restored, the animal does not react. Moving the tip of 
the tube over to a fresh specimen, this reacts at once. Moving it 
back to the first specimen, this does not contract. With a large 
colony of Epistylis, it was possible thus to test many specimens; 


50 ff. S. Jennings. 


invariably the animal reacted to the stimulus of the current the first 
time, but later did not. In a very few cases a certain individual 
would react also to the second or third or even fourth stimulus, but 
soon ceased, and in a large majority of cases the animals reacted only 


the first time. 
In Stentor roeselii the same results were obtained. The animals 


invariably reacted to the first stimulus of the current, but none of the 


numerous individuals studied reacted to a repetition. Stentor czru-- 


leus behaves ina similar manner. In this species the individuals 
often respond only once by contraction, even to the stimulus of a 
stroke with the glass rod; after the first contraction they react only 
by bending over into a new position. 

A large colony of Carchesium polypinum was situated just beneath 
the surface of the water. Touching the surface film with a needle, 
the colony at once contracted strongly. It was allowed to expand, 
and the surface film touched as before. There was no contraction. 
Repeated touching of the film caused no reaction, except the first 
time. Jarring the branch to which a colony was ablacbe gave rise 
to a parallel series of phenomena. 


From these results it is clear that the lack of reaction cannot be due 
to fatigue of the contractile elements. It is possible, as ] have dem- 
onstrated by experiment, to keep Stentor continuously contracting 
for an hour at atime. Yet the animal responds only once toa weak 
stimulus; it cannot be supposed to have been so fatigued by this 
single contraction that it cannot contract farther. 

It seems evident also that the failure to react after the first time 
cannot be due to fatigue of the sensory or perceptive power. It can 
hardly be supposed that a single stimulus would result in such fatigue 
that further stimuli are no longer perceived. Moreover this suppo- 
sition is directly negatived by the fact that in many cases there is 
other proof that the organism does continue to perceive the stimulus. 
Thus, with Stentor czeruleus, as described above, at the first stimulus 
by tapping with the glass rod the animal contracts suddenly and 
strongly. After this it no longer contracts, but the fact that it per- 
ceives the stimulus is shown by its bending far over first in one 
direction, then in another, as the stimuli are continued, as if trying 
to avoid the blows. The impression made on the observer is very 
much as if the organism were at first trying to escape a danger, and 
later merely trying to avoid an annoyance. Similar phenomena may 
be observed with Epistylis and Vorticella. 


On the Behavior of Fixed [nfusoria. 51 


Thus the third alternative seems the only conclusion to which we 
can reasonably come, in view of the facts. The organism becomes 
changed after stimulation, in such a way that it no longer reacts toa 
stimulus to which it at first reacted. There is a difference in the 
physiological condition of the organism before and after the stimulus. 
One can hardly avoid comparing these phenomena with the fact that 
in a higher organism a sudden unexpected touch or other stimulus 
will cause a reaction or ‘‘ jump,” when the same stimulus, not unex- 
pected, causes no reaction whatever. It seems not improbable that 
the phenomena are similar in fundamental character in the two cases. 

This resemblance is increased by certain further considerations. 
It is only when the stimuli are non-injurious that the unicellular 
organism ceases to respond upon repetition of the stimulus. Ifthe 
stimulus is very powerful or injurious, the reaction is continued in- 
definitely. I attempted to accustom Stentor to the stimulus from 
a very minute quantity of ;%, NaCl, brought close to it with a 
minute capillary tube. Though the stimulus was repeated at very 
short intervals for an hour steadily, the Stentor reacted in every case; 
there was no indication of becoming accustomed to the stimulus. 

The changes to be observed in the character of the reactions to 
a given stimulus when repeated show the same relation to the nature 
of the stimulus. As described in the first part of this paper, when 
the stimulus continues, and is powerful so that the reactions also 
continue, the reaction does not remain the same, but there is a series 
of different reactions. This series is a progression from less effective 
to more effective reactions, culminating in the animal’s abandoning 
its place. On the other hand, as we have seen above, the reaction is 
sometimes changed also in the case of a weak stimulus, as when 
Stentor is tapped with the glass rod. But in this case the progres- 
sion is in the opposite direction, —from a strong, effective reaction 
(contraction) to a weak one (bending over to one side.) The course 
of the reaction series, whether from less intense to more intense, or 
vice versa, depends upon the nature of the effect of the stimulus on 


‘the organism. 


D. ANALYSIS OF THE OBSERVATIONS, WITH DISCUSSION OF THEIR 
BEARINGS ON CERTAIN GENERAL PROBLEMS. 


The examination of the behavior of Stentor shows a striking con- 
trast with the known behavior of Paramecium, in the much greater 


52 ; HT. S. Jennings. 


complexity and adaptability of the former. In Paramecium the be- 
havior seems made up of a few simple reflexes, with little variation 
or adaptability. In Stentor, on the contrary, this is far from being the 
case. This difference is due, I believe, to the different method of 
life. Paramecium is typically a free-swimming organism. As I have 
pointed out elsewhere,! in the open water there are few stimuli to 
guide an organism, the conditions being nearly uniform in al] direc- 


tions. Especially is this true in the case of an organism which, like — 


Paramecium, is not sensitive to light. The result is the development 
of a few simple, almost machine-like devices for governing locomo- 
tion. Such a device is the spiral course, preventing the organism 
from aimless wandering in circles; such a device is the invariable 
turning toward a certain structually marked side when stimulated, 
which is so striking in Paramecium. On the other hand, an organism 
on the bottom is continually receiving stimuli of varied character, and 
it develops in consonance therewith a varied behavior. This differ- 
ence between the behavior of free-swimming organisms and _ that of 
those which live on the bottom is very great, and its importance is 
not usually recognized. Even in the same individual the behavior 
becomes of a very different type on changing from one of these situa- 
tions to the other. Stentor when free-swimming has the same simple 
behavior shown in Paramecium, while in Paramecium and other 
infusoria the behavior is greatly modified by contact with surfaces.” 

Proceeding to an analysis of the behavior of Stentor, it is evident 
in the first place that the same external stimulus is not always 
answered by the same reaction, but that the reaction given depends 
largely on the history of the individual (and thus upon its present 
physiological condition). Thus we find the following to be true :— 

1. After reacting to a given stimulus one or more times, if the 
stimulus is not a harmful one, the organism may cease to react, though 
the stimulus is repeated without change. 

2. After reacting toa given stimulus the first time by a very pro- 
nounced reaction (contraction), the organism may later react, if the 
stimulus turns out to bea non-injurious one, by a very slight reaction, 
as by bending over to one side. 

3. In the case of a stimulus which must in the long run be classed 


1 See JENNINGS, I9OI. 

* On some of the modifications in the behavior of organisms when in contact 
with surfaces, see especially PUTTER, 1900, and JENNINGS, 1897, pages 305-312. 
There is opportunity for further investigation in this matter. 


- 


——* 


tee et SS ee ene eer eel RD Simp Me 


On the Behavior of fixed [nfusoria. 53 


as harmful, as when a dense cloud of carmine is added to the water, 
a series of reactions is to be observed, becoming of more and more 
pronounced character, until by one of them the organism rids itself 
of the stimulus. The course of events in such a case is usually as 
follows : — 

a. Noreaction at first; the organism continues its normal activities 
for a short time. 

b. Then a slight reaction by turning into a new position, a seeming 
attempt to keep up the normal activities and yet get rid of the 
stimulus. 

c. If this is unsuccessful, we have next a slight interruption of the 
normal activities, in a momentary reversal of the ciliary current, — 
tending to get rid of the stimulus. 

d. If the stimulus still persists, the animal breaks off its normal 
activity completely, by contracting strongly, — devoting itself en- 
tirely, as it were, to getting rid of the stimulus, — though retaining 
the possibility of resuming its normal activity in the same place 
at any moment, 

This reaction is repeated many times, the organism extending and 
immediately re-contracting as soon as the stimulus is perceived. In 
this case it is interesting to note that the organism now.responds at 
once to a stimulus (by contracting) to which it at first did not re- 
spond, or to which it responded only by a reaction of different, less 
decided character. In paragraph 1 above we have the opposite case, 
where the organism ceases to respond to a stimulus to which it at 
first did respond. 

e. Finally, if all these reactions remain ineffective in getting rid of 
the stimulus, the animal not only gives up completely its usual activi- 
ties, but puts in operation another set, having a much more radical 
effect in separating the animal from the stimulating agent. It 
abandons its tube, swims away, and forms another one in a situation 
where the stimulus does not act upon it. 

It is to be noted that this series of reactions is not of such a char- 
acter that each step necessarily produces the next one; on the con- 
trary, the bringing into operation of any step depends upon the 
ineffectiveness of the preceding ones in getting rid of the stimulus. 
The series may cease at any point, as soon as the stimulus disappears. 

Further, the succeeding reactions are not mere accentuations of the 
preceding ones, but differ completely in character from them, being 
based upon different methods of getting rid of the stimulus. 


54 HI. S. Jennings. 


Throughout the whole of the series of reactions the stimulating 
agent remains without change. The differences in reaction are due 
then to changes in the organism,—to such changes as in a higher 
organism might be called changes in the ‘“‘state of mind.” Here we 
may perhaps call them changes in the “ state of protoplasm,”-though 
without implying that the two expressions are fundamentally different 
in signification. 


It is clear that it is impossible to bring such behavior under the 


d 


rubric ,“‘ tropisms” or “ taxis,’ or to present it as purely reflex in 
character; we must at the very least take into consideration physin- 
logical states of the protoplasm, as well as reflex factors. To gain a 
really satisfactory insight into the behavior, it is necessary to go 
farther than this, and to take into consideration the ends to be 
attained by the different reactions and changes in reaction, — though 
whether this necessity has its foundation only in the human way of 
looking at things, or is really inherent in the behavior of Stentor, 
is a question on which there may be difference of opinion. In any 
case it will be well to analyze the behavior a little farther from this 
point of view. So far as outward appearances go, Stentor seems ‘to 
react, like a higher organism, not merely to a stimulus now present, 
but to what is to come, —to the results of the action, as well as to the 
present conditions. The changes in the reactions, as the stimulus 
continues, seem to be directed toward the end of getting rid of the 
stimulus, —a different method being tried when one method fails. 
In the method of formation of a new tube, the same apparent reference 
to an end to be attained is forced upon the attention; there is no 
visible stimulus for the backward and forward movement of the 
Stentor, which results in the formation of a new tube; no reason 
that can be seen for this movement, except that it does form a tube. 
We have thus in this unicellular organism the outward signs of 
action directed toward the accomplishment of certain ends, and thus, 
in so far, of intelligent action. There are, of course, a number 
of different ways of interpreting such phenomena. To say that 
the reaction is really directed toward the accomplishment of an end, 
is to say that the animal reacts, not merely to a present external 
stimulus, but also to a non-present result of its reaction. This 
is only possible if the organism has already, at some previous time, 
experienced this result, so that the latter has left a trace; has 
modified the organism,— changed its physiological condition. The 
organism when stimulated reacts in accordance with, or in conse- 


On the Behavior of Fixed [nfusorza. 55 


quence of, this modification, as well as in response to the external 
stimulus; the result is action directed toward an end. 

Thus in action directed toward the accomplishment of an end there 
is an element in the organism, —a “ trace”’ or ‘‘ modification,” corre- 
sponding to the result to be attained, and due to previous experience 
of this result. But a different view is often taken of action which 
appears outwardly to be directed toward the accomplishment of a 
certain result. In many such cases it is maintained that the organ- 
ism really has no trace or modification corresponding to the result 
attained. In the case of Stentor, it would be held that the organism 
has become a sort of mechanism which gives a definite series of re- 
sponses, when energy of such and such a character acts upon it 
under such and such conditions, for such and such a, period of time. 
The result follows just as a precipitate is produced in a chemical 
reaction. The difficult problem according to this view is how 
reactions happen to be produced that are adapted to the accom- 
plishment of certain ends. This is explained (usually) by natural 
selection. © 

In many of the instincts of higher organisms, such a view as that 
last set forth seems forced upon us by the fact that the organism has 
had no opportunity to get impressed upon it any trace or modification 
corresponding to the result to be produced. The animal responds 
before it has ever experienced the result. Cases of this sort will 
occur to every one. 

In Stentor, however, this difficulty perhaps hardly exists, since it 
is not possible to separate sharply the given Stentor from its ances- 
tors that may have experienced the results of any given reaction. 
Since each Stentor arises by simple division of a previous Stentor, 
there is here no special difficulty in the inheritance of acquired 
characters. If a given Stentor. has become modified by certain 
éxperiences, there is no evident reason why the two Stentors de- 
rived from it by division should not retain this modification. . Hence 
we have no absolute ground for maintaining on this basis that in 
Stentor the apparent reaction with reference to the result to be 
attained is not really a reaction with such reference. In other 
respects, we seem to have the same problem in attempting to 
explain the behavior of Stentor that we have in the instincts of 
higher animals. 

It may not be out of place, finally, to indicate the bearing of the 
behavior of Stentor on the problem of consciousness in the lower 


56 Ll. S. Jennings. 


organisms, a matter which has been much discussed of late. I do 
not see that there can be any objective criterion of consciousness, 
hence this question in strictness does not fall within the field of an 
investigation directed to the end of determining what observation 
and experiment can tell us of the behavior of an organism. But it 
may be of interest to point out the relation of the phenomena de- 
scribed to certain questions that have been raised. In former papers 
(Jennings, 18992, page 339; 18990, page 13), I expressed the opinion 
that in Paramecium the behavior was comparable to that of an isolated 
muscle, and that ‘‘ we are not compelled to assume consciousness or 
intelligence in any form to explain its activities.” This statement is, 
of course, well within the facts, as far as objective investigation can 
give them to us, yet it is perhaps of little significance, since it could 
probably be made for any organism, outside of the self. The behavior 
of Paramecium is of a character to emphasize strongly the possible 
machine-like character of the activities of the lower organisms, In 
Stentor we have a very different case, showing that the behavior of 
Paramecium cannot be considered a type for that of all infusoria. 
Paramecium has become adapted in its behavior to a very simple set 
of conditions, and its behavior is of corresponding simplicity. In the 
behavior of Stentor, we find all the outward indications of action 
directed toward the accomplishment of certain ends. We have then 
the same ground for attributing consciousness to Stentor as to higher 
animals which show behavior of a similar character, —no more, no 
less. 

In a recent paper Minot (1902) has expressed the opinion that 
‘the function of consciousness is to dislocate in time the reactions 
from the sensations,” — to inhibit the direct reactions at certain times ; 
to cause reactions at certain times to stimuli that have occurred pre- 
viously, ‘This disarrangement ... seems to me the most funda- 
mental and essential characteristic of consciousness that we know, — ” 
‘“‘__and so far as we know, it belongs exclusively to consciousness.” 

Judged by this criterion (substituting “stimulus” for ‘sensation ” 
as used by Minot) we should clearly have to attribute consciousness 
to Stentor. As shown above, this organism at certain times inhibits 
the reactions to stimuli, to which at other times it reacts strongly. 
Moreover, the nature of its reactions to a given stimulus depends upon 
stimuli previously received, and this I think is all we can mean when 
we say that an animal reacts at a certain time to a stimulus previously 
received. I confess that Dr. Minot’s criterion seems to me by no 


On the Behavior of Fixed L[nfusoria. 57 


means an absolute one; and that unconscious mechanisms could be 
constructed, and indeed do exist, in which there is a dislocation in 
time between the action of an outer agent upon the machine and the 
reaction of the machine, similar to that which we find in organisms. 
It is nevertheless interesting to find the behavior of a unicellular 
organism falling within the category that would be considered con- 
scious by Minot. 

If we consider now the criterion held by Loeb and Bethe, — that 
consciousness depends upon associative memory, upon the power 
of learning,— it is perhaps not so easy to decide where Stentor 
stands. The changes in reaction when the stimulus is long continued ; 
first no reaction, then bending into new position, then reversal 
of the ciliary motion, then contraction many times repeated, and 
final leaving of the tube, could perhaps be considered cumulative 
effects of the stimulus, and hence as not giving evidence of associa- 
tive memory. But this of course leaves quite out of consideration 
the fact these different reactions are all adapted, by different methods, 
to getting rid of the stimulus, and it is exactly this adaptation to an 
end that furnishes the real problem. How does the organism happen 
to give these particular reactions, thus adapted to the accomplishment 
ofan end? If it gives these particular reactions as a result of experi- 
ence, it has learned, hence, on the hypothesis we are considering, it 
has consciousness. If it has not learned to give these purposive reac- 
tions, the only alternative hypothesis as to how this has come about 
is, so far as I am aware, through the action of natural selection upon 
chance movements. 

All together, it must be clearly recognized, I think, that objective 
study can give us nothing final on the problem of whether conscious- 
ness does or does not exist in the lower organisms, We can have 
indeed no absolute proof of the existence of consciousness outside 
of ourselves. Whether one holds that Stentor and Paramecium have 
or have not consciousness will depend chiefly upon his general 
system of philosophy, which is of course not mainly determined by 
observation and experiment. 


E. SumMMARY. 


The foregoing paper comprises a study of the behavior of Stentor 
roeselii, Stentor czeruleus, and Vorticella. 

Taking Stentor rceeselii as the type, the following are the most 
important points brought out: 


58 FH. S. Jennings. 


I. In the unstimulated Stentor the ciliary motion causes a vortex 
whirling to the left and descending to a point on the left oral side. 
Only a small part of the water or suspended material in the vortex 
reaches the mouth. The unstimulated Stentor does not contract. 

II: Under the influence of slight mechanical stimuli, as when 
carmine grains or other small objects are carried to the disc by the 
vortex, the behavior is as follows : — 

A. For a time the normal behavior may be continued, some of the 
particles being ingested. 

B. In some cases the Stentor may bend toward an object touching 
one side of the disc (‘“ positive thigmotaxis”’). 

C. With repeated weak stimuli, the Stentor may react the first 
time by contraction; then cease to react farther, though the stimulus 
is continued. Or the animal may react the first time by contraction; 
later by merely turning to one side, as described in paragraph 1, 
below. ou 

D. When the objects striking the Stentor are very numerous or 
large or are combined with a chemical stimulus, or are otherwise 
unfavorable, there is a series of reactions, as follows: — 

1. The Stentor first bends into a.new position, by twisting on the 
long axis, then bending toward its aboral side. It may thus rid 
itself of the stimulus; if not, this reaction is usually repeated a 
number of times. 

2. If the reaction described above does not rid the animal of the 
stimulus, it next reverses the ciliary current for an instant. This 
reaction may be repeated a number of times. 

3. If the stimulus still continues, the animal next contracts into 
its tube. This reaction is repeated many times, if the stimulus 
continues, and usually the period during which the animal remains 
contracted becomes longer. 

4. Finally, if the stimulus continues, the animal lets go its hold 
and abandons its tube. It swims away through the water, its be- 
havior while free being similar to that of Paramecium. After a time 
it forms a new tube, by a peculiar process, in another place, where 
it is not affected by the stimulus, and remains there. 

III. Under chemical stimuli (1) the reactions are essentially the 
same as above described for mechanical stimuli, the series 1-4 
described above taking place in a similar manner. (2) The distri- 
bution of a diffusing chemical approaching the Stentor is determined 
by the ciliary vortex of the Stentor. The result is that the chemical 


eo 9 ry eat 


> he. 


On the Behavior of Fixed [nfusoria. 59 


arrives at the mouth and oral surface of the Stentor before it touches 
any other part of the body; the latter may remain for a long time 
quite unaffected. Hence the conditions necessary for the orientation 
of the body in lines of diffusing ions are not present, and such 
orientation cannot occur. This is true also for other ciliate infu- 
soria; probably also for flagellates. 

Similar considerations apply also to the reactions to temperature 
variations in the water. 

IV. To osmotic stimuli Stentor responds only after plasmolysis is 
far advanced. 

V. In Stentor ceruleus the behavior is essentially similar to that 
of S. roeselii, — the same series of reactions being given to continued 
stimuli. The following differences are to be noted : — 

I. Stentor czruleus does not bend toward a weak mechanical 
stimulus at one side of the disc, as does S. rceselii. 

2. Stentor czruleus has no tube and abandons its place of attach- 
ment much more readily than does S. rceselii. 

VI. In general the reactions of Vorticella are similar to those of 
Stentor, though it was not observed to abandon its place as a response 
to stimuli. 

VII. Stentor, Vorticella,; Epistylis, and Carchestum were found 
to become accustomed to repeated mechanical stimuli, so that they 
cease to respond by contracting when the stimulus is repeated. 
This is not due to fatigue, since they frequently respond only to the 
first stimulus. It is likewise not due to lack of perception of the 
stimulus, since after ceasing to contract they often give other evi- 
dence that the stimulus is perceived. 

VIII. On the whole the behavior of Stentor is complicated, as 
compared with that of Paramecium, and shows considerable power 
of adaptation. Whether the animal reacts to a given stimulus or 
not, and how it reacts, depends upon previous subjection to this 
stimulus, and upon the previous method of reacting to it. Ifa 
stimulus continues, the animal gives a series of reactions which are 
not invariable in order or length of continuance; each reaction of 
this series is adapted, by a different method from the others, 
to getting rid of the stimulus. These reactions, together with the 
method of forming a new tube, have the appearance of being directed 
toward the accomplishment of definite ends. 


60 ff. S. Jennings. 


LITERATURE CITED. 


DAVENPORT, C. B. 
1897. Experimental morphology, i. 


DELAGE et HEROUARD, 
1896. Traité de zoologie concrete, i. 


HonpcE, C. F., and AIkIns, H. A. 

1895. American journal of psychology, vi, pp. 524-533- 
JENNINGS, H. S. 

1897. Journal of physiology, xxi, pp. 258-322. 

1899a. This journal, ii, pp. 311-341. 

18994. American naturalist, xxxill, pp. 373-389. 

1goo. This journal, iii, pp. 229-260. 

1go1. American naturalist, xxxv, pp. 369-378. 

1902. Journal of applied microscopy, Jan., pp. 1597-1602. 
JOHNSON, H. P. 

1893. Journal of morphology, viii, pp. 467-556. 
LoEB, J. 

1900. Comparative physiology of the brain and comparative psychology. 


MASSART, J. 
1got, Annales de l'Institut Pasteur, Aug. 25. 


MINoT, CHARLES S. 

1902. Science, xvi, no. 392, pp. I-12. 
PUTTER, AUG. 

1900. Archiv fiir Physiologie, supplement-Band, pp. 243-302. 
RHUMBLER, L. 

1898. Archiv fiir Entwickelungsmechanik, vii, pp. 103-350. 
VERWORN, M. 

1889. Psycho-physiologische Protistenstudien, 218 pp. 


THE ACTION OF ALCOHOL ON MUSCLE. 


By FREDERIC S. LEE ann WILLIAM SALANT. 


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


yes recent years the subject of the physiological action of 

alcohol has received in the laboratories much attention, but 
experimenters have strangely neglected the action of this drug on 
muscle. 

Twenty years ago Kobert,! experimenting on frog’s muscle, main- 
tained that alcohol, when given in large doses, greatly diminished 
contraction, but that this did not occur with small or medium quantities. 
Later Lombard, Frey, Destrée, Kraepelin and his pupils, and Scheffer, 
obtained with the ergograph interesting and valuable results with the 
drug in the human being; but such an organism is too complex to 
afford a solution of the problem at hand. It was only two years ago 
that the correct method, that of isolating the muscle tissue, was again 
employed, and in this manner Scheffer? discovered that the frog’s 
gastrocnemius under the influence of alcohol was able to perform at 
first a markedly increased amount of work, but after the period 
of three or four hours this changed to a decrease; after the elimi- 
nation of the intramuscular nerve-endings by curare, alcohol had 
no effect, and its former action was hence ascribed to the peripheral 
nervous system. The work of the present authors has also been 
performed on isolated muscle and has also yielded positive results, 
which, partly agreeing and partly disagreeing with those of Scheffer 
in his own field, have covered a wider range. 

It is perhaps only natural that those who chance to read the 
present article may be tempted to apply to the human being the 
facts here presented. This, the authors, who have made no experi- 
ments on man, have deliberately refrained from doing. ‘While it is 
not at all improbable that the general facts discovered with frog’s 


1 KosperT: Archiv fir experimentelle Pathologie und Pharmakologie, 1882, xv, 
P. 73: 
2 SCHEFFER: L6id, 1900, xliv, p. 24. 
OL 


62 Frederic S. Lee and William Salant. 


muscle may be true also of human muscle, we would warn all readers 
against assuming that the same quantitative relations prevail in the 
two species. Experimental pharmacology has revealed innumerable 
instances of a drug acting upon two species with qualitative ima 
but with quantitative inequality. 


METHOD. 


Our experiments have been made upon healthy, active frogs, the 
common leopard frog, Rana virescens, having been used in all except 
one series, in which young bullfrogs were employed. In all except a 
few of the earlier experiments, where commercial ethyl alcohol was 
used, we have employed Squibb’s absolute ethyl alcohol, diluted in 
various known proportions with distilled water. So far we have 
made no experiments with any variety other than ethyl alcohol. In 
the majority of experiments alcohol alone was given, but in one series 
curare was also administered. 3 

Our usual procedure has been to ligate one leg and then to inject 
the solution either into the dorsal lymph sac or into the stomach. 
Thus the drug was able to enter the circulation and reach all parts of the 
body except one leg. The non-alcoholized leg was at once amputated, 
and its gastrocnemius muscle was excised and prepared in the usual 
way for stimulation. In all cases the stimulation was direct, never 
through the nerve. In most of the experiments sixty opening in- 
duction shocks per minute were given and continued until the 
muscle was exhausted, the contractions being recorded on a very 
slow drum by the isotonic method. In twenty to seventy-five 
minutes (in the majority of the experiments about forty-five minutes) 
after the injection of the alcohol the frog was killed, and the alcohol- 
ized gastrocnemius was removed and studied similarly. Thus from 
each animal records were obtained from a non-alcoholized and an 
alcoholized muscle. From the records a comparison could readily 
be made of the number and extent of the contractions, of the total 
amount of work that each muscle was capable of doing, of the oncom- 
ing and course of fatigue, and of various other phenomena. For 
certain specific purposes this method of study was at times altered 
in certain ways. Thus, in studying the total amount of work per- 
formed by the muscle, instead of recording the contractions graphi- 
cally a work-adder was sometimes found convenient. Further, a 
series of experiments was performed in which, instead of the stimu- 
lation occurring at regular intervals of time, the muscle at the end 


The Action of Alcohol on Muscle. 63 


of each relaxation closed a circuit and thus stimulated itself anew. 
This enabled a record to be made of the total number of the possible 
contractions and relaxations in a given time. In another series a 
single contraction curve of the alcoholized muscle was superimposed 
on that of the non-alcoholized, and thus a comparison of the detailed 
features of the two contractions was made possible. 

There appear to be wide variations among frogs in the rate of 
absorption of alcohol, a phenomenon which has been observed like- 
wise of other animals with the same drug. Ordinarily symptoms of 
intoxication, such as sluggishness, appear within a few minutes after 
the injection, but not infrequently all symptoms are greatly delayed. 
In our earlier experiments we were careful to administer the alcohol 
in known quantities of the desired solution per gram of frog. But 
the great variation in absorption made such care superfluous, and 
later we merely filled the dorsal lymph sac or the stomach fairly full 
of the solution and waited the desired time before preparing the 
muscle. At first sight it might seem that we employed with each 
animal somewhat large quantities of the drug, but the quantity which 
we introduced into the body is only broadly suggestive of the relative 
quantity which the muscle actually received. Of greater importance 
in this respect is the strength of the solution, a stronger solution 
having an effect differing in kind from that of a weaker one. Yet 
with any one percentage employed our results show a considerable 
range of quantitative variation, which seems to be due in part to the 
varying rate of absorption. It would be interesting to investigate 
the cause of this latter phenomenon. In all our experiments care 
was taken to make the record with the muscle at once after removing 
the latter from the circulation. 

We have been especially careful to take into consideration and 
eliminate the various possible causes of error. It is a well-known 
fact that under normal circumstances corresponding muscles on 
the two sides of the body are capable of performing very different 
amounts of work. A few preliminary experiments made it prob- 
able that any error arising from this source would not usually 
exceed 10 per cent of the whole amount of work performed. Our 
results obtained with alcohol are far in excess of this. Some of our 
earlier experiments suggested the idea that in the frog the right 
gastrocnemius is normally stronger than the left. Enough control 
experiments were made to show that this is not true, but that the 
average capacity for work is approximately equal on the two sides. 


64 Frederic S. Lee and Willhiam Salant. 


The number of experiments that we have performed is approxi- 
mately two hundred. It is thought that with this large number any 
possible errors due to individuable variation have been eliminated. 

The work was carried on during the months of January to June, 
1901, inclusive, and December, 1901, to May, 1902, inclusive. 


RESULTS. 


We have found that the action of alcohol on muscle varies with 
the relative quantity of the drug employed. Our experiments may 
be arranged in three groups, according as small, medium, or large 
quantities were used. In brief it may be stated that alcohol in small 
quantity has no appreciable action; in medium quantity it is favor- 
able to activity; in large quantity it is unfavorable to activity. 

I. Action of alcohol in small quantity. — A few experiments were 
tried with 0.03 c.c. of 10 per cent alcohol per gram of frog, admin- 
istered for forty-five minutes. The results were by no means uniform, 
showing in some cases a greater, in some a less working power in 
the alcoholized muscle. But all these variations were within the 
limits of error, as exhibited by a comparison of normal muscles under 
equivalent conditions, and hence the results are not to be ascribed 
to a possible action of alcohol. In other words, alcohol in the small 
proportion used does not appear to exert any action whatever on 
muscle tissue. Scheffer claims to have demonstrated in Rana escu- 
/enta a favorable action in the proportion of 1 part by weight of pure 
alcohol to 1000 parts of body weight. Taking this as literally 1 gram 
and not I c.c. and computing our results on the same basis, the 
density of ethyl alcohol being 0.79, we find ourselves unable to 
demonstrate a genuinely favorable action in doses of 2.37 parts by 
weight of pure alcohol to 1000 parts of body weight. This slight 
difference between Scheffer’s results and our own we are not able to 
explain. 

2. Action of alcohol in medium quantity. — Because of its novelty 
and the general ignorance regarding it, the favorable action of 
alcohol has proven of great interest to us-:and has engaged our 
especially careful attention. 

Let us consider the action of a 10 per cent solution in the propor- 
tion of 0.08 c.c. per gram of frog, or, in other words, 40 parts by 
weight of pure alcohol to 1000 parts of body weight, injected about 
thirty to forty-five minutes before the testing of the muscle. Fig. 1 
shows the graphic record of a typical experiment of this kind. 


000A meme rere r 


The Action of Alcohol on Muscle. 65 


Record of contractions of corresponding gastrocnemii; two-fifths the original size; upper, normal; lower, 


FicurE 1.— Experiment 56. 


Total number of con- 


Total amount of work in 


moderately alcoholized; 0.08 c.c. of 10 per cent alcohol to 1 gm. of body weight for 45 minutes; stimuli ] per second. 


(normal) 363, (alcoholized) 819; percentage of increase = 125.6. 


4614, (alcoholized) 8260; percentage of increase = 79. 


tractions or total working time in seconds, 


gram-millimetres = (normal) 


The initial contraction of such an alco- 
holized muscle is often, although not 
always, slightly less in extent than that 
of the normal. The phenomenon of the 
progressive decrease in extent of the 
four or five so-called “ introductory con- 
tractions”’ is more common than in the 
normal muscle. Following these the con- 
tractions show a progressive increase in 
extent, which is more gradual than is 
normally the case, the maximum being 
reached after the muscle has performed 
often one-fourth, and in some cases one- 
third, of its total number of possible con- 
tractions. As to the actual extent of the 
maximal contractions there seems to be no 
constant difference between the muscle 
with and that without the medium quan- 
tity of alcohol, sometimes the former 
surpassing, at other times the latter, 
while at still other times no difference is 
appreciable. 

The most striking feature of the record, 
however, and one of the most prominent 
phenomena of the experiment is the larger 
number of contractions of which the alco- 
holized muscle is capable. In the experi- 
ment now presented, the normal muscle 
made 363 contractions, while that which 
was under the influence of alcohol made 
819, an increase of more than 125 per cent, 
before an equal amount of exhaustion set 
in. Other examples are as follows, the 
average percentage of increase in the 
twenty-six experiments being 59.5. 


66 Frederic S. Lee and William Salant. 


No. of 
experiment. / 


PAB Ai: 


No. of contractions 
of normal 
muscle. 


| No. of contractions 


of alcoholized 
muscle. 


411 


Percentage of 
increase in number 
of contractions. 


The Action of Alcohol on Muscle. 67 


This table indicates the great range of quantitative variation which 
is seen throughout the study of the physiological action of alcohol. 
It can readily be observed that differences in the rate of absorption 
exist in different individuals, and to this is doubtless due in con- 
siderable part the variation in results. But in how far this phenom- 
enon is responsible, and in how far there are other and unknown 
causes, it is difficult to say. In this feature our experience is paral- 
leled by that of many other experimenters with the same drug. 

Associated with this larger number of contractions, and a phenom- 
enon which is often so marked as to be plainly evident from the 
record, is the greater amount of work which the alcoholized muscle 
is capable of performing before exhaustion sets in. In Experiment 
56 the total quantities of work, expressed in gram-millimetres, 
performed by the non-alcoholized and the alcoholized muscles 
respectively, were 4614 and 8260, an increase of 79 per cent in 
favor of the alcohol. Other examples are found in Table II, the 
average increase in all the experiments quoted being 40.4 per cent. 

This increase in work is one of the very interesting features of the 
present research. It is commonly believed that a human being 
under the influence of alcohol is able to perform, for a certain time 
at least, a larger amount of muscular labor, and the investigations 
of Lombard, Frey, Destrée, and Kraepelin, although not agreeing in 
details, have demonstrated with scientific exactness this main fact. 
But until Scheffer’s work of two years ago it had not been conclu- 
sively shown that this increase is a function of the muscle and not 
of the central nervous system. In this feature our results agree 
essentially with those of Scheffer, a difference being that we have 
found it necessary to use larger quantities of alcohol than he, a fact 
of minor importance. It is a gratification to be able to confirm 
so fully the findings of his experiments. It can hence be accepted 
-without question that with the help of a moderate quantity of alcohol 
a muscle itself working continuously and until exhaustion sets in, 
is capable, quite apart from the central nervous system, of doing 
a greater amount of work than a similar muscle without alcohol. 

Associated further with the larger number of possible contractions 
is the delay in the oncoming of fatigue and exhaustion. In Experi- 
ment 56 the actual time during which the normal muscle was able 
to perform contractions when stimulated once every second was 
363 seconds, which is to be compared with 819 seconds for the 
alcoholized mnscle, the increase being 125.6 per cent. The table 


68 


No. of 
experiment. 


TABLE II. 


Total amount of 
work, in gram-milli- 
metres, performed 

by normal 
muscle. 


2613 


2716 
2821 
5049 
4576 
3186 
4043 
6222 


Total amount of 
work, in gram-milli- 
metres, performed 

by alcoholized 
muscle. 


3759 
MS 
5034 
7812 


6468 
6829 
2808 
10310 


Frederic S. Lee and Wilham Salant. 


Percentage of 
increase in total 
amount of work. 


Lhe Action of Alcohol on Muscle. 69 


on page 66 will indicate the actual working time of the muscles in 
question, if in the second, third, and fourth columns the words, 
“number of contractions” be replaced by ‘total working time in 
seconds.” In the twenty-six experiments there quoted the average 
prolongation of the working time, due to the alcohol absorbed, was 
59.5 per cent of the normal working time, a result as remarkable and 
as interesting as the increased amount of work. In the experiments 
in the two tables above given, of which we have the records of the 
increase both of working time and of amount of work performed, we 
find the average of the former to be 56.5 per cent, of the latter 42.4 
per cent; that is, the working time seems to be increased to a some- 
what greater degree than is the total amount of work accomplished. 
In considering this prolongation of working time, or delay of fatigue, 
the conditions of the experiment should be kept in mind. The two 
muscles, the one normal and the other alcoholized, are stimulated 
at the same rate. The necessity of adhering to this condition will 
be appreciated later, when we come to study the matter in another 
way (page 70). 

A final feature recognizable from the graphic record now under 
discussion is the more pronounced contracture in the normal muscle. 
This is characteristic of nearly all 
our curves. It begins fairly early 
and increases at first slowly, but 
it is especially pronounced and 
increases rapidly toward the close 


of the experiment. It is obviously 
due to the inability of the muscle, 


during the brief interval of one Ficure 2.— Experiment 189. 


second between two stimulations, 
to complete its activity and relax 
to its previous state, and the 
lesser contracture of the alcohol- 


Curves of 
single contraction of corresponding 
gastrocnemii; original size; slower, 


normal; quicker, moderately alcohol- 
ized; about 0.08 cc. of 10 per cent 
alcoho] to 1 gm. of body weight for 25 


: minutes. 
ized muscle suggested to us that 


the drug may quicken the whole process of muscular activity. 
We proceeded to investigate this by superposing the muscle 
curves, and we obtained such a result as is shown in a clear-cut 
manner in Fig. 2. This reveals the interesting fact that by the 
agency of the aicohol both the phases of the process, both con- 
traction and relaxation, are actually shortened in time. It is well- 
known that, of the two phases, relaxation is in general more readily 


70 Frederic S. Lee and Witham Salant. 


influenced by external agents than contraction, and we find this 
to be the case with alcohol: 7.¢., the internal processes expressing 
themselves outwardly by contraction are quickened to a certain 
degree, those expressing themselves by relaxation are quickened to 
a greater degree. In the experiment represented in Fig. 2 the ratio 
of quickening of the two phases is 1:1.5, but in many cases the 
effect on relaxation is relatively greater than this. 


||] 


| ll | Mi i 


ii 


WAU qu iy it ANU wi ui HAA HUGH itt cli 


Hi 


AW j 


{th 


i hil 


FicureE 3.— Experiment 138. Record of contractions of corresponding gastrocnemii ; 
original size; upper, normal; lower, moderately alcoholized ; 0.12 c.c. of 10 per cent 
alcohol to 1 gm. of body weight for 20 minutes; stimuli succeeding on complete 
relaxation. Below is line of seconds. Total number of contractions in 2 minutes 
= (normal) 53, (alcoholized) 167; percentage of iucrease = 215. 


This phenomenon. of quickening led us to a further method of 
investigation. With quicker action a muscle is able to make in 
a given time a larger number of complete contractions, and unless, 
as is rarely the case, these are pronouncedly weak, the muscle is 
able to perform in the given time a greater amount of work. This 
can be tested by allowing the writing-lever of the muscle, on com- 
plete relaxation, to make an electrical contact, complete a circuit, 
and thus be the means of stimulating the muscle anew. This 
we have repeated many times with a result illustrated in Fig. 3. 
Here in the period of two minutes the normal muscle made 53 
complete contractions, the alcoholized muscle 167, an increase of 
215 per cent. Even with the diminished extent of the contrac- 
tions occurring in this experiment, a feature not universally present, 
the alcoholized muscle performed in the same time an amount 
of work 2.6 times greater than that performed by the corresponding 
muscle without alcohol. It is obvious that under such circumstances 


et 


Cw 


The Action of Alcohol on Muscle. 71 


a muscle may, and our experiments have shown that it does, often 
become fatigued or exhausted actually sooner than a normal muscle. 
This is no contradiction to the conclusion reached on page 67, regard- 
ing the delayed fatigue with equal rates of stimulation. 

The favorable action of alcohol in moderate quantities seems abun- 
dantly proved. This may be summarized as quicker contraction, 
quicker relaxation, larger number of contractions and increase of 
work in a given time, larger number of contractions and greater 
total amount of work before exhaustion sets in, and delay of 
fatigue. 

Is this action exerted upon the nervous tissue within the muscle or 
upon the protoplasm of the muscle cells? Scheffer was unable to 
demonstrate any influence of alcohol on the total work of the muscle 
after the elimination of the nerves by the use of curare, and he 
concludes that alcohol does not act dynamogenically on the muscular 
apparatus, but is a true excitant of the peripheral nervous system. 
Because of its intrinsic interest, and because of the fact that Scheffer 
expresses himself so decidedly, we have examined this matter with 
especial care and by many experiments. We curarized the frog 
by means of an injection of a few drops of a solution of curare 
into the dorsal lymph sac and, after complete paralysis had resulted, 
we followed the usual plan regarding the injection of alcohol, 
testing by stimulating directly at approximately sixty times per min- 
ute and recording the contractions until the muscle became exhausted. 
In each case indirect stimulation was first tried, to convince ourselves 
that the curare had acted properly, and that no contraction was possi- 
ble through the mediation of the nerve; and unless this state was 
reached, the experiment was abandoned. The results were the same 
as in the experiments already discussed, in which no curare was used. 
With the curarized and alcoholized muscle there is the same larger 
number of contractions before exhaustion sets in, the delay of fatigue, 
the increased amount of work, and the lesser contracture, which indi- 
cates quicker action. Fig. 4 is typical of this series of experiments. 
Here the increase in the number of contractions and the total work- 
ing time is 80.9 per cent, the increase in the total work performed 
is 129.2 per cent. From these experiments, the conclusion is un- 
avoidable that the favorable action of alcohol is exerted directly on 
the protoplasm of the muscle cells and not on the intra-muscular 
nervous tissue. 

A word may be said as to the quantity of alcohol used in obtaining 


72 Frederic S. Lee and William Salant. 


the favorable effect and the time during which it was allowed to act. 
Pronounced favorable action was obtained with quantities ranging 
from 0.06 c.c. to 0.12 c.c. of a 10 per cent 
solution per gram of frog; or, expressed 
in another form, with quantities ranging 
from 4.74 to 158 parts by weight of pure 
alcohol to 1000 parts of body weight. 
Larger quantities of the 10 per cent solu- 
tion were not tried. It must be borne in 
mind, however, that doubtless owing chiefly 
to the very great variation in the rate and 
quantity of absorption in different indi- 
viduals, these figures are only broadly 
significant. 

The same may be said of the time re- 
quired for the action to take place. No 
action could be assured in a shorter time 
than twenty minutes. After thirty min- 
utes the stage of increased activity was 
usually well marked. In most of our ex- 
periments we allowed the drug to act 
forty-five minutes. -From one hour on, 
our observations led us to expect that 
the favorable action might be replaced by 
an unfavorable one. But we have not yet 
made a careful study of the later effects; 
nor have we yet investigated the power 
and course of recuperation after the in- 
creased work following the injection of 
alcohol —a subject of evident and con- 
siderable interest. 

3. Action of alcohol in large quantity, — 
A few experiments were performed with 
commercial ethyl alcohol diluted to 19 per 
cent and allowed to act for the usual time. 
In this series the results varied, showing 
in three of the seven experiments an in- 
crease of work and in four a decrease, the 
percentage of decrease being the greater. It thus seems probable 
that with this strength of solution we are approximately on the neutral 


80.9 per cent; 


AA A a 


Ada 


Lh 


| 


Increase in number of contractions or total working time in seconds 


Record of contractions of corresponding curarized gastrocnemii; one-half the original size; 
in total work performed = 129.2 per cent. 


stimuli 1 per second. 


Experiment 162. 


4. 


upper, normal; lower, moderately alcoholized; about 0.08 c.c. of 10 per cent alcohol to 1 gm. of body weight for 30 


minutes ; 


increase 


FIGURE 


: 
" 
'. 


Record of contractions of corresponding gastrocnemii; two-fifths the original size; upper, normal; lower 
’ 


; 0.2 c.c. of 33.3 per cent alcohol to 1 gm. of body weight}for 45 minutes. 


Figure 5,.— Experiment 29. 


The Action of Alcohol on Muscle. 73 


Decrease in total number of contractions 


strongly alcoholized 


cent. 


5 per 


73.4 per cent; decrease in total amount of work performed = 92 


or total working time in seconds 


zone between quantities causing a distinctly 
favorable effect and those causing a distinctly 
unfavorable one. Our observations make it 
also probable that, owing to individual peculi- 
arities, this zone of neutrality varies greatly 
with different individuals. A strength of solu- 
tion favorable to one may easily be unfavorable 
to another. . 

With a percentage of 334 the action in 
nearly all cases is distinctly unfavorable. In 
the seventeen experiments in this series the 
following data were obtained: In four cases’ 
the alcoholized muscle either failed altogether 
to contract with the customary stimulus, or 
gave to the lever a barely perceptible motion ; 
in three the contractions were very feeble and 
few in number, as is illustrated in Fig. 5; in 
eight there was a marked diminution in the 
amount of work performed, the number and 
extent of the contractions were much less, and 
exhaustion ensued earlier than in the normal 
muscle; in one the detrimental effect of the 
alcohol was slight and within the limits of indi- 
vidual error; and in one there was a favorable 
action, characterized by the usual increase of 
work, greater number of contractions, and de- 
layed fatigue. Thus, of the seventeen ex- 
periments, in fifteen there were decidedly 
unfavorable effects, in one a doubtful, and in 
one a favorable effect. The action on the indi- 
vidual contractions and relaxations was not 
studied, but as regards the number of contrac- 
tions, the working time, and the total amount 
of work accomplished, the alcohol in the strong 
solution produced effects exactly the reverse of 
those observed with the medium solutions, while 
at the same time the extent of the contractions 
was less. In all these respects the stronger 
alcohol is distinctly toxic to the protoplasm and 
unfavorable to its activity. 


7A Frederic S. Lee and William Salant. 


CAUSES. 


We have aimed to set forth merely the facts which experiment and 
observation have revealed. A number of possible causes of the phe- 
nomena in question may readily be imagined. But it will be of more 
value to reserve a discussion of these until further experiments, which 
are now under way, bearing more particularly on the subject of cause, 
are completed. 

One feature in the present research which has proved of special 
interest to the authors is the fact that it has been possible in the 
case of a specific narcotic to establish in so clear-cut a manner the 
fact of a stage in its action which is altogether favorable to the activ- 
ity of a specific kind of protoplasm. The results suggest that the 
action of alcohol on other kinds of living substance ought to be very 
carefully examined, and it is not at all improbable that studies of 
other narcotics along similar lines might be advantageously followed. 
The attention of investigators hitherto seems to have been confined 
too exclusively to the specifically narcotic action of these interesting 
substances. 


SUMMARY. 


1. In small quantity ethyl alcohol does not appear to exert any 
action on frog’s muscle. 

2. In medium quantity it exerts a favorable action, which is char- 
acterized by a quickening of the contraction; a quickening of the 
relaxation; the power of making a larger number of contractions and 
of performing a larger amount of work in a given time; an increase 
in the working time, or, in other words, a delay of fatigue; and the 
power of making a larger number of contractions and of doing a 
larger amount of work before exhaustion sets in. This action is 
exerted directly on the muscle protoplasm itself, not on the intra- 
muscular nerve tissue. 

3. In large quantity ethyl alcohol exerts on frog’s muscle an 
unfavorable action, which is, in general, the reverse of that caused 
by medium quantities of the drug, and is characterized by a decrease 
in the extent of the contractions; a decrease in the working time, or, 
in other words, a hastening of fatigue; and the power of making a 
smaller number of contractions, and of doing a smaller amount of 
work before exhaustion sets in, 


THE IMPORTANCE OF SODIUM CHLORIDE IN 
HEART. ACTIVILY. 


BY Dav LD |, LENGE: 


CONTENTS. 
Page 
| Thslaieyaliteyn tay ites cack couiltohab al eisai ince (lon hagerh Me ein age ae Ray 5 
tie Dhe-actionof caffein =.= . - : Part tora be tee 7 (6 
III. The action of oxygen gas in eaabiiaton a mae deleaane a. As PaO 
a. The action of hydrogen peroxide on heart strips. . ..... =. 79 
b. Action of oxygen gas on strips ina moist chamber . . . . . . . 82 
IV. Theories as to the action of salts in causing beats in heart strips . . . . 85 
Memen acts thatido notagree with the calcium theory. . . . . . .. . << . 86 
VI. Facts that support the sodium chloride theory . . Seale at EDO, 
VII. Answers to some of the objections to the sodium pularide eaty sites: i696 
SMS OHEIUISIONS © SMe ee Vs, te Pn 4 a ees UA) gna eae ROT 


I. INTRODUCTION. 


RELATIONSHIP has. been demonstrated between blood salts 
and heart activity. The meaning of this is at present un- 
known, and investigators are not agreed as to the réle of the various 
salts involved. Loeb! worked with tissues of the rhythmically con- 
tracting jelly fish, Gonionemus, and came to the conclusion that 
sodium chloride is of primary importance in rhythmic phenomena, 
while Howell,? Ringer, and others seem inclined to give calcium 
a preponderating influence in the reactions originating beats in heart 
strips. 

In a former paper® the author applied Loeb’s ideas to the heart 
of the turtle, and was led to believe with him that sodium chloride 
is fundamentally necessary for the origination of rhythmic activity in 
isolated strips of cardiac tissue. 

This article will present some new facts confirming this conclusion 
and indicating clearly that sodium chloride plays a special rdle in the 


1 Loes: Archiv fiir die gesammte Physiologie, 1900, Ixxx, p. 229. 
2? HOWELL: This journal, igor, vi, p. 181. 
8 LINGLE: This journal, 1900, iv, p. 265. 

75 


76 David J. Lingle. 


physiology of heart muscle. An attempt will be made to prove so- 
dium chloride a peculiar factor in the complicated chemical and physi- 
cal interactions that star¢ rhythmic action, and at the same time some 
results will be described that do not agree with Howell’s! latest theory 
as to the use of salts in causing the cardiac rhythm. 

The results recorded here are true for narrow strips cut from the 
ventricle of the turtle, that is, pieces of normal tissue that have ceased 
to beat as a result of violence inflicted in preparing them for experi- 
mentation ; and an answer has been sought for the question, When a 
strip in this condition begins to beat again in a salt solution, which 
element is it that renews or permits the renewal of the rhythm? The 
answer seems to be that it is sodium chloride. The power of this 
salt in this respect appears to be unique. 


Il. THe ActTIon oF CAFFEIN. 


Previous work with heart strips in salt solutions convinced me 
that unless these contained sodium chloride they could not origi- 
nate beats. It was found that solutions of lithium chloride, cane- 


FIGURE 1.—A shows the character of beats before caffein was given. #. After the 
drug had acted. Both tracings were made by the same heart strip. 


sugar, dextrose, and glycerine were individually unable to re-establish 
rhythms, nor did they succeed in starting beats when combined with 
salts of calcium and potassium, using the latter in the proportions 
considered most favorable for the development of beats. Was this 


1 HOWELL: Loc. cit. 


Importance of Sodium Chloride in Heart Activity. 77 


failure to establish rhythms due to the absence of sodium chloride ; 
or was it due to something else, possibly a slight physical or chemical 
obstruction ? An attempt was made in the following way to secure 
an answer to this question. If the absence of sodium chloride from 
such solutions was not an insuperable barrier to the development of 
beats, then it might be possible to produce beats in strips without it 
by increasing the rhythmic power of the tissue. This can be accom 
plished by using a heart stimulant, and a series of experiments was 
made to find a stimulant with a marked action on strips from the 
turtle’s ventricle (Fig. 1). Among others caffein was used. It exerts 
a decided augmenting influence, and has the additional advantage of 
being a crystalline compound. A solution made by adding from 5 to 
10c.c. of a saturated solution of the caffein to enough sodium chloride 
solution to make one hundred cubic centimetres was found effec- 
tive. In such a mixture strips beat with greater force, 7. e. the 
rhythmic power is intensified. But when caffein in these propor- 
tions was used in combination with sugar solutions and lithium chlo- 
ride solutions no beats ever appeared in the heart strips, nor did they 
appear when the most favorable proportions of calcium and potassium 
salts were added. The solutions used in this series of experiments 
were the following : — 


Experiment 1.— go c.c. cane-sugar 7, 
ro c.c. saturated caffein solution. 


Experiment I1.— 90 c.c. cane-sugar 4, 
10 c.c. saturated caffein solution. 


Experiment I1[.— 8o c.c. cane-sugar %, 
Io c.c. saturated caffein solution, 
to c.c. CaCl, 0.26 per cent. 


Experiment IV.— Same, except LiCl solution took the place of sugar. 


Experiment V.— 8g c.c. cane-sugar 4, 
Io c.c. saturated caffein solution, 
1 c.c. KCl solution 0.3 per cent. 
Experiment V/.— Same, except LiCl replaced sugar. 
Experiment VII.— 79 c.c. cane-sugar 4, 
Io c.c. saturated caffein solution, 


to c.c. CaCl, solution 0.26 per cent. 
1 c.c. KCl solution 0.3 per cent.. 


Experiment VIII. — Same, with LiCl in place of sugar. 


78 David J. Lingle. 


No beats were caused by any of these solutions. The controls, 
however, with sodium chloride developed good beats. It seems then 
that solutions containing a powerful stimulant like caffein are unable to 
start beats unless sodium chloride is present. The experiments also 
indicate that the series of beats seen in a sodium chloride solution 
belong to a rhythm originated in the strips, and are not simply micro- 
scopic or invisible beats intensified. Gaskell! has called attention to 


the fact that it is not always easy to determine when a heart has’ 


really ceased beating. And the same may well be true of strips of the 
ventricle. Microscopic examinations were also made of the strips, 
and failed to reveal any beats in quiet strips such as were used in 
these experiments. 


III. THe ACTION OF OxYGEN GAS IN COMBINATION WITH 
SALT SOLUTIONS. 


It has long been known that the heart is extremely sensitive to 
oxygen and carbon dioxide. These gases also have a remarkable 
action on heart strips.2_ Under certain conditions oxygen can origi- 
nate beats, and in normal beating strips carbon dioxide abolishes 
them. The action of the two gases in this latter case resembles that 
of muscarin and atropin (Fig. 2). But in studying the action of 
salt solutions on heart strips, the important powers possessed by these 
gases have been ignored, and it is possible some erroneous ideas have 
arisen in consequence. 

Howell ® in his work has paid little attention to the action of oxygen 
because he was convinced that salt solutions contained the necessary 
oxygen required by strips taken from the hearts of cold-blooded 
animals; but this idea is open to objection, for a salt solution evidently 
cannot supply oxygen at the same rate as the blood, and all hearts 
respond very quickly to even moderate variations in the oxygen 
supply of the latter, and beside it will be shown that the addition of 
oxygen modifies the action of heart strips when they beat in sodium 
chloride solution and also in mixed salt solutions. Indeed, we may be 
certain that ordinary salt solutions do not furnish enough oxygen for 
the moderate activity in heart strips. The investigation of the action 
of oxygen was suggested by observing that strips active in pure 

' SCHAFER’S Text-book of physiology, 900, ii, p. 185. 
2 PoRTER: This journal, 1898, i, p. 517. 
® HOWELL: Loc. cit. 


Importance of Sodium Chloride in Heart Activity. 79 


sodium chloride solutions usually beat stronger in the air while a 
transfer was being made from one solution to another, an effect 
which a few experiments proved conclusively to be caused by the 
oxygen of the air. It was found too that sodium chloride and 


c b Og @ x 
Oy CO, CO, NaCl 
FIGURE 2.— Shows action of oxygen gas on beating strips. At (#) beats began in NaCl, 


at (a) the strip was surrounded by CO, and beats ceased; Og was then added and 
they reappeared, to be stopped apain by CO, (4), and again started by O,(c). The 
drum moved at rate of 18 inches in twenty-four hours. 


oxygen are interestingly related. For when strips are treated with 
sodium chloride and oxygen the beats are stronger and last in favor- 
able cases as long as those in a solution of mixed salts. <A study of 
oxygen effects also throws light on the nature of the so-called sodium 
chloride arrest. The influence of oxygen can be demonstrated either 
by placing the heart strips in pure oxygen gas in a moist chamber 
or by adding hydrogen peroxide directly to the bathing solution. 

a. The action of hydrogen peroxide on heart strips. When two 
strips are taken from the same heart and one is put into a pure sodium 
chloride solution and the other into go c.c. of sodium chloride solution 
plus 10 c.c. of a 3 per cent hydrogen peroxide solution, beats begin in 
both cases. The one with the hydrogen peroxide becomes quickly 
frosted over with small bubbles of oxygen gas. Some of these break 
loose and rising form a frothy mass on the surface. This strip always 
beats stronger and longer than its companion (Fig. 3). 

During the very hot weather of the summer of 1891 the strips with 


80 David J. Lingle. 


hydrogen peroxide usually beat about three times as long as controls 
without it. It may be questioned whether this is simply an oxygen 
effect, for the addition of hydrogen peroxide to a salt solution modi- 
fies it in three ways: (1) The salt solution is diluted. This is not 
the factor which is responsible for the prolongation of the rhythm in 
hydrogen peroxide, for the same amount of distilled water has no such 
influence when added to the salt solution. (2) Traces of salts and 
acid which are always contained in hydrogen peroxide are added to 
the solution. The salts and free acids were removed by careful dis- 
tillation, and it prolonged the rhythm the same as the commercial 
hydrogen peroxide, showing that the salts and acids contained in the 
commercial hydrogen peroxide are not the cause of the prolongation 
of the active period. (3) Oxygen gas is given to the strip and the 
solution. The facts that will presently be given prove that this is the 
real agent in the experiment. 

Not only can hydrogen peroxide improve beats; it can do more. 
When beats stop in a pure sodium chloride solution, the addition of 
10 c.c. of 3 per cent hydrogen peroxide to the 90 c.c. of sodium 


FIGURE 3.— One-half the original size. Shows the effect of H,O, on strips in NaCl. 
These strips were made from the same heart, and the utmost care was taken to have 
them anatomically and physically alike. The upper strip was in pure NaCl, the 
lower in 93 c.c. NaCl 4- 7 c.c. of 3 per cent H,O,. The beats in 4 lasted less than 
one hour, those in #& almost nine hours. 


chloride solution will restore the beats and cause them to continue for 
many hours. 


Howell’ has made the following summary of the various ways in 


1 HOWELL: Loc. cit. 


Lmportance of Sodium Chloride in Heart Activity. 81 


which beats can be restored in strips when they have stopped in 
pure sodium chloride : — 

1. Immersion of the strip in Ringer’s solution. 

2. Immersion in a mixture of sodium chloride and calcium 
chloride. 

3. Immersion in dextrose or cane-sugar solution. 

4. Immersion in lithium chloride solution. 

To these should be added the method with hydrogen peroxide. 
The hydrogen peroxide is added directly to the solution in which 


a 
FIGURE 4.— One-half the original size. Shows recovery of beats in a solution of NaCl 
by adding 7 c.c. of HO, (redistilled) to the salt solution. (a) shows beats in NaCl 


lasting from 10.32 to 12.02; (4) when H,O, was added. The drum moved at rate 
of 18 inches in twenty-four hours. 


beats have stopped. Last summer this method was compared with 
the others, and it was repeatedly noticed that hydrogen peroxide was 
effective in restoring beats when other solutions failed. In this re- 
covery there is a latent period lasting from thirty minutes to two 
hours (Fig. 4). Then feeble beats appear. These slowly increase 
to a maximum which, at its greatest, is always weaker than that 
in sodium chloride, and then comes a very gradual decline in force. 
The duration of this second series of beats is from three to fifteen 
times as long as that seen in sodium chloride alone. 

In all these experiments sodium chloride is indispensable. If it is 
lacking the hydrogen peroxide has no power to originate beats, as is 
shown by experiments with strips in the following solutions and 
combinations : — 


Experiment I, — 93 c.c. LiCland 7\e-c. H,O,. 
Experiment L[. — 93 C.c. cane-sugar # and 7 c.c. H,O.. 


§ 10\¢-¢. 11,0, and 
Experiment IIT, — toc.c. CaCl, o. 26 per cent. 


( 90°C.G solution ‘ E.G:G; KC ] 0.3 per cent. 


89 C.c. cane-sugar 7%. 


82 David J. Lingle. 


to c.c. CaCl.o.26 percent, 
85 c.c. solution ~) 1 CC: KCl 0.3 per cent. 
( 89 c.c. LiCl solution. 


15 c.c. H,O, and 
Experiment LV. — 


( ro cic. H,O, and . 
Experiment V.— “ to c.c. CaCl, 26 per cent. 


hee c.c. solution ~ 1 ¢-¢. KCl 0.3 per cent. 
( 8g c.c. LiCl solution. 


Experiment VI, — Same as Experiment V, except that cane-sugar 
was used in place of the LiCl solution. 


Experiments ViZand VI//.—Same as Experiment V/. LiCl and cane-sugar 
solutions, but with only 5 c.c. of H,Q.. 


In none of the above solutions did strips originate beats, though 
hydrogen peroxide was present and in some cases calcium and po- 
tassium also in the proportions considered most favorable to their 
appearance. Here again it is seen that without sodium chloride a 
combination of favorable conditions is without power to start beats. 

b. Action of oxygen gas on strips in a moist chamber. — If- heart 
strips are subjected to an atmosphere of pure oxygen in a moist 
chamber they show certain peculiarities that strikingly reveal the 
fundamental importance of sodium chloride in heart beat phenomena. 
If a strip is placed in sodium chloride solution until beats begin, and 
then, while beating, is removed to a moist chamber containing oxy- 
gen, the force of its beats is strengthened and the rhythm sustained. 
In some instances beats under these conditions lasted seventy-two 
hours, and apparently stopped then because putrefaction destroyed 
the strip. These experiments were made during excessively hot 
weather, and if putrefaction could have been prevented the strips 
probably would have contracted longer, for in some cases the lower 
fourth of a strip was yellow and putrid twenty-four hours before the 
upper part ceased beating. 

Under these circumstances we have a series of rhythmic beats ori- 
ginated under the influence of a bath containing a single salt in solu- 
tion, and this is sustained in pure oxygen gas as long as rhythms 
usually are in a solution with a mixture of salts. Such an experiment 
shows the exceptional position held by sodium chloride among agents 
that originate heart beats, for treatment with no other solution of a 
single salt accomplishes the same result. It also may be considered 
as partially confirming one side of Loeb’s theory as to the action of 


EE 


¢ 


Importance of Sodium Chloride in Heart Activity. 83 


sodium chloride. In the experiment described beats are only started 
in the sodium chloride solution. Had the heart strips remained in 
this they would have stopped after a short time. But the withdrawal 
from the solution while beats were beginning prevented the develop- 
ment of the unfavorable stage produced by the diffusion of an excessive 
amount of sodium chloride into the strips, and when this is avoided in 
the presence of oxygen there is no sodium chloride standstill. Air has 
the same power as pure oxygen, but its influence is not so marked. 
In the experiments made with strips taken from a salt solution and 
exposed to moist air, the contractions were always much smaller than 
in pure oxygen. Furthermore, oxygen gas, like hydrogen peroxide, 
can restore beats in strips when they have run down in a sodium 
chloride solution. If a strip in this condition is removed from the 
sodium chloride solution, and transferred to a moist chamber full of 
oxygen gas, a latent period follows, then feeble beats reappear which 
gradually grow stronger until a maximum is reached, and this is sus- 
tained for a long time. The whole series continues twenty-four hours 
or, in some instances, longer (Fig. 5). In this case recovery occurs 
without any diffusion of salts, which indicates clearly that the ordinary 
sodium chloride arrest is largely due to a lack of oxygen. 


ba 


FIGURE 5.— One-fourth the original size. Shows that oxygen gas will restore beats 


where they have ceased in NaCl solution. Beats at (2) when in NaCl; at (4) strip 
was put in moist chamber and surrounded with oxygen gas. The duration of the 


restored beats was about twenty hours. 


These results with hydrogen peroxide and pure oxygen occur only 
when the strips are under the influence of the salt, sodium chloride. 
Oxygen of itself is unable to start beats or to sustain those not origi- 
nated under the influence of sodium chloride. When inactive strips 
are put into a sugar solution, or any salt solution with hydrogen per- 
oxide, but without sodium chloride, beats never develop. Nor will 
quiet strips start beating in an atmosphere of pure oxygen without 
having been treated previously with sodium chloride. An apparent 


84 David J. Lingle. 


exception should be mentioned here : sometimes a non-beating strip 
when put into pure oxygen undergoes regular rhythmic variations in 
tone. These show slow elongations and contractions, and on a drum 
making only one revolution in twenty-four hours they look like heart 
beats. But that the resemblance is only superficial is apparent when 
it is noted that each contraction and relaxation on the curve repre- 
sents a period of about fifteen minutes. These are evidently not 
beats but rhythmic variations in tone, and are not necessarily an 
oxygen effect. For they were seen in strips in an atmosphere of 
pure hydrogen, and in the latter case were even better developed. It 
is possible they belong to the same class of phenomena noted by 
McWilliam! in mammalian hearts during rigor, or to those discovered 
by Fano? and Botazzi® in the auricle of the frog. Though, as might 
be expected, in the turtle’s heart they last much longer, in some cases 
over twenty-four hours. 

Apparently, then, oxygen gas cannot of itself start rhythms such 
as are set up when sodium chloride is present, nor will it restore or 
sustain beats that have originated independently of sodium chloride. 
The opportunities of testing this latter point occur but rarely. Some- 
times, in exceptional cases, when a ventricle is cut into strips these 
do not stop beating, and when tied to a lever, records can be made of 
their contractions. Beats of this kind do not last long, and, of course, 
originate without the strips coming in contact with a sodium chloride 
solution. When strips in this condition are put while beating into 
oxygen gas the rhythm is not sustained. In a typical case it ran 
down in thirty minutes. Nor would the continued action of oxygen 
restore the beats again, as it does with those that have stopped under 
the influence of sodium chloride. In one case out of five instances 
of this kind observed during the past year and a half, a strip that was 
beating with unusual vigor was seen after it had run down to make 
microscopic irregularities in its curve that suggested a rudimentary 
recovery in the oxygen gas. But these in no way resembled the 
beats seen in strips treated with sodium chloride. 

To sum up the action of oxygen we may say: It cannot start beats 
when sodium chloride is lacking. Combined with sodium chloride, it 
increases the force of the beats, and lengthens the duration of the 


' MACWILLIAM: Journal of physiology, 1901, xxvii, p. 338. 

* FANO: Beitrage zur Physiologie, Carl Ludwig gewidmet, Leipzig, 1887, 
p. 287. 

8’ BorTazzi: Journal of physiology, 1897, xxi, p. I. 


Eee 


Importance of Sodium Chloride tn Heart Activity. 85 


rhythm. It also restores beats in strips that have ceased to beat in a 
sodium chloride solution. Oxygen and sodium chloride together 
can maintain beats as long as a mixture of salts, provided the sodium 
chloride does not act for too long a time. Oxygen will also improve 
the beats of strips in a solution with a mixture of salts. These facts 
throw a little light on the réle of salts in causing strips to beat, and 
they modify, in some respects, the present theories explaining the 
role of salts in such cases. 


IV. THEORIES AS TO THE ACTION OF SALTS IN CAUSING 
BEATS IN HEART STRIPS. 


Two theories have lately been put forward to explain why quiet 
heart strips renew beats in salt solutions. Loeb! has advanced the 
idea that rhythmic activity is primarily the outcome of oxidase action 
which sets free energy. Salts are, however, necessary to make physi- 
cal conditions such that the energy can assume the form of rhythmic 
action, and sodium chloride produces physical conditions peculiarly 
favorable to the development of such activity. But only when it acts 
a certain time; if this is exceeded it produces conditions that are 
unfavorable. This excessive action of sodium chloride produces the 
sodium chloride standstill. And he believes the unfavorable condi- 
tion can be neutralized or overcome for a time by calcium which acts 
like an antitoxin. 

Howell? explains the rhythm-producing power of salt solutions 
differently. In the first place he believes the heart strips do not beat 
primarily because there is enough potassium salts normally present in 
them to inhibit such activity, and he thinks beats begin in a solution 
of sodium chloride because there the strip losing potassium faster 
than calcium, after a time enough potassium leaves the strip to 
weaken its inhibitory power, and beats begin. These result from the 
interaction of the calcium and the sodium chloride in the strip. He, 
with Ringer and others, apparently attributes to calcium a preponder- 
ating influence in the process; for he thinks that beats continue only 
so long as a supply of calcium remains in the strip, and that they cease 
when a certain amount of it has diffused out. The standstill in 
sodium chloride, therefore, is not due to the diffusion of too much 
sodium chloride into the strip, but to the diffusing of calcium out of 

1 Loes: Festschrift fiir Fick, Braunschweig, 1899, p. 99. 


> 


2 HOWELL: Loc. cit. 


86 David J. Lingle. 


the strip. The phenomena of the whole series of beats in sodium 
chloride is explained, according to this idea, by salt losses sustained 
in the sodium chloride solution, and not by the special action of 
this salt in the bathing solution. 

Possibly neither of these theories is entirely correct, and a brief 
discussion of them may help to clear the way for a better understand- 
ing of salt action in relation to rhythmic activity. Howell,! in his 
latest paper has presented some objections to Loeb’s conception of 
the effects of sodium chloride. These can be answered more or less 
successfully. At the same time some facts can be given that do not 
at all agree with Howell’s idea. The point of view of the two theories, 
it should be noted, is not the same. To Howell certain salts that 
start a strip beating and keep it beating are the causes of the beat; 
while to Loeb the processes originating rhythmic activity and those 
sustaining it are in a sense distinct and more or less antagonistic in 
nature, and he thinks the inevitable excessive action of the originat- 
ing agency must be neutralized by some means if beats are to per- 
sist. The point we wish to make is that the beginning of rhythmic 
beats in the strip is associated with the action of sodium chloride, 
and this is its special rédle. Though if such beats are to persist the 
sodium chloride action must be sooner or later balanced by calcium 
or some other salt. Without sodium chloride rhythms do not begin. 


V. FAcTS THAT DO NOT AGREE WITH THE CALCIUM THEORY. 


1. As to beats beginning because potassium diffuses from the 
strips. 

If the diffusion of potassium chloride from the strip were the 
cause of the beats they should occur in solutions of dextrose, cane- 
sugar, and lithium chloride, where the diffusion of potassium is not 
interfered with; but they do not. It is not because these are injuri- 
ous, for sodium chloride in combination with them causes strips to 
beat well. Again, the prevention of the diffusion of potassium from 
the strip should, according to this idea, prevent the origination of 
beats. When this is done by adding a salt of potassium in varying 
amounts to the solution about the strip, it is found that heart strips 
beat in solutions of sodium chloride which contain so much potassium 
that a loss of this element is absolutely impossible. Beats, for exam- 
ple, have been repeatedly seen to develop in sodium chloride solutions 


1 HOWELL: Loc. ct. 


Importance of Sodium Chloride tn Heart Activity. 87 


containing from 0.03 per cent to 0.4 per cent of potassium chloride. 
In such solutions potassium does not leave the strip, but rather the 
strip gains potassium, as is shown by the short duration of the series of 
beats. With 0.4 per cent of potassium chloride it is not easy to get 
beats, and when they do occur they last but a few minutes. But the 
point is, beats are here originated under conditions that are prohibi- 
tive according to Howell’s theory, and sometimes, though not often, 
with a large amount of potassium chloride present, they last a consid- 
erable time (Fig. 6). 

2. If the latent period, as Howell states, represents the time re- 
quired for a sufficient amount of potassium to diffuse out of the strip, 
it should be longer in solutions containing more potassium salt than 
in solutions with less; but this is not so. A large number of experi- 
ments were made comparing the action of sodium chloride solution 


Ml 


FIGURE 6.— Two-thirds the original size. Shows that beats will originate in NaCl with 
0.2 per cent of KCl dissolved in it. In this case they lasted almost four hours. The 
lever was raised at (2). 


alone with solutions of sodium chloride containing potassium chloride 
in varying proportions, and the results show no constant perceptible 
difference in the length of the latent periods. It sometimes happens, 
too, that strips begin to beat within a few minutes after they are 
placed in a solution of sodium chloride. Here any diffusion of potas- 
sium is impossible, while the whole exposed surface comes into instant 
contact with sodium chloride and may be influenced by it. 

3. The second part of the theory states that beats in strips begin 
when potassium has been removed because then calcium and sodium 
directly interact in the tissues. As to this point results are not con- 
clusive. It has been impossible so far to eliminate calcium from a strip 


88 David J. Lingle. 


and have it beat. For no experiment has been devised that simply 
accomplishes this; other complications are always introduced. So 
in dealing with the relative importance of calcium and sodium chloride 
in the origination of beats we must use indirect or suggestive evidence. 
We can experimentally differentiate the action of these two elements 
to some extent. If strips are put into a solution made up of 70 c.c. 
lithium chloride % + 30 c.c of an % sodium oxalate solution, they begin 
to beat and continue active in some cases as long as three hours. 
These beats are always feeble, and are apt to show periodic variations 
in strength. But the point is, here beats are developed in a solution 
containing more than enough oxalate to precipitate all the calcium in 
the strip. They develop when the oxalate must be continually throw- 
ing down calcium and removing it from any physiological rdle in the 
tissue, and yet under these conditions beats begin and last for some 
time. In sucha solution no sodium chloride, as such, is given, but 
there are present in solution both sodium and chlorine ions. That 
both of these are needed is seen when a solution of lithium chloride 
and lithium oxalate or of sugar and sodium oxalate is used. With 
these liquids in each of which one element, either sodium or chloride, 
is lacking, no beats originate, — at least they did not when the follow- 
ing combinations were used : — 


Experiment I.— 70 ¢.c. cane-sugar 7, 
30 c.c. LigC,Q,. 


Lxperiment I[.— 70 c.c. cane-sugar 4, 
201c:c. Nase. 

Experiment L11. — 85 c.c. cane-sugar 7, 
Perens lr OHO} 
7 IGG. Nay alas 

Experiment ITV. — 70 c.c. cane-sugar 7, 
15 c.c. LigC,Ox, 


15 c.c. NasC.QO,. 


If a solution which precipitates calcium permits beats to develop 
when sodium chloride is present, and will not start them if so- 
dium chloride is lacking, it indicates that sodium chloride does 
more than calcium in originating beats. Even under extreme con- 
ditions of this kind it is possible to get beats started, for example, 
in a solution containing 50 c.c. sodium chloride and 50 c.c. of oxalate 
solution, and sometimes, though it is rare, these oxalate solutions can 
start beats when potassium chloride also is present to the extent of 


— 


Importance of Sodium Chloride in Heart Activity. 89 


0.06 per cent in the solution. Such a liquid constitutes an exceed- 
ingly unfavorable medium for the development of beats, according to 
Howell’s theory, and yet beats occur if sodium chloride is present, 
and not otherwise. 

4. In his paper Howell! attributes the arrest in sodium chloride 
to a lack of calcium in the strip. It occurs, he says, when the strip 
has lost by outward diffusion a certain percentage of its calcium. 
The experiments previously described, with oxygen and hydrogen 
peroxide, make this explanation without modification unsatisfactory. 
If a strip that has ceased beating in 93 c.c. of a sodium chloride solu- 
tion be given about 7 c.c. of hydrogen peroxide it begins to beat 
again and continues active for hours, notwithstanding the previous 
calcium loss and the fact that calcium is under these conditions, 
according to the theory, leaving the strip continually. If the arrest 
in the first instance was due to a loss of calcium, how could the addi- 
tion of a little hydrogen peroxide restore this? From the beginning 
of such experiments to the end calcium diffusion had gone on, we may 
suppose, uniformly and continuously, and yet during this time beats 
in the strip had begun, grown to a maximum, declined and disap- 
peared, reappeared, increased to a second maximum, and slowly dis- 
appeared a second time. The variations in activity in such a case do 
not correspond with the history of calcium in the strip. If calcium 
is indispensable for the origination of beats the strip when it stopped 
in sodium chloride must have had enough calcium left in it for the 
second series of beats, as no new calcium was given to it in any form. 
Hence it is difficult to see how the strip can be so sensitive to a lack 
of calcium that this could be the cause of the arrest as ordinarily 
described. 

A study of the action of oxygen shows that the onset of the 
arrest as seen in sodium chloride is greatly accelerated by a lack of 
oxygen in the salt solution. If oxygen is furnished the arrest is post- 
poned for hours in some cases. And it is possible that the arrest as 
usually observed with sodium chloride solution is intensified or caused 
by something related to mild asphyxia. That the ordinary arrest can- 
not be due toa lack of calcium and that calcium is not so fundamen- 
tal, was further demonstrated by some experiments made with solu- 
tions of sugar and sodium chloride. When a small amount of sodium 
chloride, 16 to 18 c.c. % solution, was added to 100 c.c. of a sugar 
solution, strips sometimes remained quiet in this mixture for twenty- 


1 HOWELL: Loc, cit., p. 206. 


90 David J. Lingle. 


four hours and then began to beat. A similar result was seen when 
sodium citrate was used in place of sodium chloride. If a strip stops 
in a solution of sodium chloride after two or three hours’ immersion 
because too much calcium has left it, how is it possible for a strip to 
begin beating in a sugar solution after calcium diffusion has gone on 
for nearly twenty-four hours? In the sugar solution calcium diffu- 
sion, so far as we know, occurs just as rapidly as in a solution of 


sodium chloride. Either the calcium does not diffuse rapidly or it | 


is not fundamental for the origination of beats, unless in the presence 
of oxygen, and in sugar solutions containing a small amount of sodium 
chloride, less of it can do the work. This idea, that calcium is not so 
essential as sodium chloride for the origination of beats, is supported 
by some work of Loeb’s* in which he has shown that calcium actually 
inhibits rhythmic beats in striped muscle and in Gonionemus tissue. 
That it also acts as an inhibitory factor in nerve tissue seems not 
impossible, because when this is treated with reagents like the oxa- 
lates its irritability is greatly increased. So that stimuli, ordi- 
narily inactive, become afterward intensely so. It is indeed prob- 
able that calcium owes its well-known value as a sustainer of beats to 
this very restraining power. Possibly it is this very characteristic 
that enables it to preserve tissue, and it is by holding in check, rather 
than by causing activity, that calcium is enabled to prolong the active 
period of heart muscle. If calcium and sodium chloride both caused 
activity, their interaction should shorten, not prolong, the working 
time. That this is the normal réle of calcium is suggested by a 
study of its action in excess, which, as is well known, always slows 
beats in heart strips. This slowing tendency may become so great at 
times that beats entirely stop, —a very clear case of an inhibitory 
rather than an inciting power.” 


VI. Facts THAT SUPPORT THE SODIUM CHLORIDE THEORY. 


1. The fact is well established that sodium chloride solutions are 
the best of all media for originating beats in heart strips. But is 
there evidence that in these solutions it is sodium chloride that is the 


1 LoeB: Loc cit. 

* That the beats ina calcium solution are always stronger than in sodium 
chloride does not disprove this idea. The beats in calcium solutions usually have 
a lower rhythm; hence the energy expended in two or more beats in sodium 


chloride is in calcium solutions combined in a single beat. 


Importance of Sodium Chloride in Heart Activity. 91 


effective agent ? The following facts may be considered as support- 
ing this idea: — 

a. There is the fact that the efficiency of a solution in producing 
beats in strips is to some extent dependent on the amount of sodium 
chloride in it. If sodium chloride is the active agent, we should ex- 
pect this result; but if the beats in sodium chloride are started by 
the diffusion of potassium, when this is otherwise provided for and 
physical conditions are maintained, the amount of sodium chloride 
should be immaterial. But the following experiments indicate clearly 
that this is not the case. 


Experiment I, — 98 c.c. cane-sugar 7, No beats. 
2 c.c. NaCl §. 


Experiment IT, — 98 c.c. LiCl %, No beats. 
2 c.c. NaCl 4. 


Experiment I[[.— 96 c.c. cane-sugar ¥%, No beats. 
4 c.c. NaCl %. 


Experiment 1V.— 96 c.c. LiCl %, No beats. 
4 c.c. NaCl 2. 
Experiment V.— 94 c.c. cane-sugar 7, Seven beats after a latent period 
6 c.c. NaCl %. of about four hours. 
Experiment VI. — 94 c.c. LiCl %, No beats. 


6 c.c. NaCl %. 


Experiment VII.— 92 ¢.c. cane-sugar 7, No beats. 
8 c.c. NaCl %. 


Experiment VIII.— 92 c.c. LiCl %, No beats. 
8 c.c. NaCl §. 
Experiment IX. — go c.c. cane-sugar 7, Good beats lasting nearly one 
Fo.c.c: NaCl 7. and one-half hours. 
Experiment X.— go c.c. LiCl %, Latent period about four hours. 
10 c.c. NaCl 4. Irregular beats one hour. 
Experiment XI, — 88 c.c. cane-sugar 7, No beats. 


12 c.c. NaCl %. 


Experiment XII.— 88 c.c. LiCl %, Long latent period, three and 
12 c.c. NaCl %. one-half hours. Beats. last- 
ing over two hours. 


92 David J. Lingle. 


Experiment XITI.— 86 c.c. cane-sugar #7, A very long latent period, of 
rice; NaCl 2. five hours ; then beats lasting 
two and one-half hours ; then 
pause of about seven hours, 
and then again for five hours. 
The last of these occurred over 
twenty-six hours after strip was 
put into the salt solution. 


Experiment XIV.— 86 c.c. LiCl %, jatent period about four hours. 
14 c.c. NaCl %. Good beats lasting one hour. 

Experiment XV.—- 84 c.c. cane-sugar ¥7, A latent period lasting twenty- 
16 c.c. NaCl %. two hours; then very strong 


beats that were going vigor- 
ously twenty-five hours after 
the strip was put into the 


solution. 
Experiment XVI.— 84 c.c. LiCl %, Latent period of six hours. 
16 c.c. NaCl %. Beats lasting about one hour ; 


not so strong as those in sugar. 


Experiment XVII. — 82 c.c. cane-sugar #%, Latent period about one and 
To 7c.c., Natit: one-half hours; very power- 
ful beats lasting twenty-four 

hours. (In this there were 

two periods of inaction, each 

about one hour long, and 

a third one of about four 

hours. The beats at the end 

of the twenty-four hours were 

almost invisible. ‘The rhythm 

showed periodic groupings 


throughout.) 
Experiment X VIII, — 82 c.c. LiCl %, Latent period three hours. Very 
18 c.c. NaCl §. powerful beats, lasting about 
two hours. A typical NaCl 
series. 


This series of experiments was made on strips taken from nine differ- 
ent hearts. If they could have all come from the same heart there 
would doubtless have been greater uniformity, as it is well known 
that there is considerable difference in the reaction of different hearts 


Luportance of Sodium Chloride in Heart Activity. 93 


to salt solutions. Nevertheless the results show that the origination 
of beats, the duration and the character of beats are determined by 
the amount of sodium chloride present. If sodium chloride is the 
important factor in causing beats these results agree with expecta- 
tions. But if Howell’s ideas are correct they are incomprehensible. 
In these cases the loss of potassium and calcium was not interfered 
with. The only factor changed was the amount of sodium chloride 
present. In experiments fifteen and seventeen, both in cane-sugar, 
latent periods of almost twenty-four hours’ duration occurred, and 
then vigorous beats began. Then, too, such beats occur and last for 
over twenty-six hours in some cases, where sodium chloride is dilute, 
a case of retarded development of the toxic stage. When it is strong, 
one-eighth normal, they never last so long. The beats in sugar solu- 
tion are the remarkable ones, and this is in accord with the fact that 
heart strips sometimes beat for a time in this solution alone, —a fact 
as yet unexplained. 

(b) The experiments with oxygen gas prove that sodium chloride 
is a remarkable factor in originating beats. When strips are placed 
in sodium chloride long enough to start beats, and are then trans- 
ferred to an atmosphere of oxygen, they continue active as long as 
they ordinarily do in mixed salt solutions. In this case cal- 
cium is not needed in the bath because the short stay in sodium 
chloride does not cause injurious effects. It may be objected that the 
long duration of beats in this case results because calcium could not 
diffuse out of the strip when it was in the oxygen, and so calcium as 
much as sodium chloride is responsible for the beats. From the 
nature of the case here we cannot prove the contrary. But the 
indirect evidence previously given is against this, and in addition to 
that we may state the following points to indicate that calcium is not 
the more important factor in the case under discussion : 

(1) The fact that in solutions with hydrogen peroxide present 
where calcium diffusion goes on, beats are very slowly abolished. 

(2) The fact that strips can remain in sugar solution twenty-four 
hours and then begin beating and remain active much longer, indi- 
cates that beats are possible in sodium chloride after a very long out- 
ward diffusion of calcium, even after sufficient time has elapsed to 
permit the loss of most of the diffusible calcium. If in these cases 
the loss of calcium does not prevent beats, how can we consider cal- 
cium to be the cause of the beats in strips influenced by sodium 
chloride and oxygen? 


94 David J. Lingle. 


(c) The fact that in sodium chloride solutions beats begin feebly 
and then increase to a maximum, shows that they originate and 
develop in the same ratio as sodium chloride enters. The intensity 
of the phenomena increases as inward sodium chloride diffusion 
progresses. 

(d) Lastly, we have the facts demonstrated by Loeb on ordinary 
striped muscle. This is nota rhythmic tissue, and calcium, if it be 
the important factor in rhythmic activity, is certainly not present here 
in the proportion to develop beats, and yet sodium chloride develops 
rhythmic activity in this tissue. It has been stated that solutions of 
thoroughly purified sodium chloride cannot originate beats, but this 
is not correct. To prove this solutions were made with sodium 
chloride that had been so thoroughly purified by precipitation with 
hydrochloric acid gas that no trace of impurity could be detected even 
with the most careful tests. There was no doubt as to the salt being 
of the highest purity, for it had been repeatedly tested in the chemical 
laboratory of the University. This solution started beats just the 
same as the salt ordinarily used. 

That sodium chloride plays a remarkable réle in rhythmic phenom- 
ena is beyond doubt; whether it is unique is a question, and one 
that cannot be answered until the whole field of inorganic chemistry 
has been worked over. So far no salt has been found that is able to 
replace it. We may safely say that when strips of heart tissue beat 
well and normally they are, or have been, in a solution containing 
sodium chloride. 

If we adopt the physical theories as to the nature of muscular con- 
traction, and regard salts as agents producing. those conditions that 
result in what we call the contraction, then there is no reason why 
other salts should not be found capable of producing the same changes 
as sodium chloride. So far in this work but one case of this kind has 
been noted, and it is by no means saeeacaeeh When strips were 
placed in a solution consisting of 30 c.c. of an % Li,C,O, solution and 
70 c.c. of an % LiCl solution, a series of Babies beats was originated 
which lasted abot one and one-half hours. These beats were very 
irregular and showed a marked tendency to group themselves. The 
series did not begin gradually and develop to a maximum; the first 
beat was as powerful as any. The decline, however, was gradual, 
resembling that seen in sodium chloride. In this case sodium ions in 
the bathing solution were lacking, but lithium and chloride ions were 
present, and Dr. Loeb has shown that lithium ions are capable of 


I 


Importance of Sodium Chloride in Heart Activity. 95 


replacing sodium ions in striped muscle. But I know of no other 
instance of this interchangeability in heart muscle. The exception in 
this case occurs with an ion closely related physiologically and not 
unlike sodium chemically. Besides, it must not be forgotten that 
there is another factor here. These beats occurred in an oxalate 
solution, one that precipitated the tissue calcium, and so may have 
greatly increased the irritability of the muscle and caused a predomi- 
nance of sodium chloride in the tissue. In heart tissue in this condi- 
tion the lithium solution was able to start beats, which it certainly is 
not able to do in normal tissue. But at best this doubtful ability to 
do the work of sodium chloride cannot be compared with that of 
sodium chloride in causing rhythms. 

Ammonium sulphate will sometimes cause rhythmic activity in 
striped muscle. But it will not do so in heart muscle when used in 
the following proportions : — 


Normal solution. 


Lal 
. 


2. % solution. 
2. ” solution. 
4. % solution. 
a ” solution dil. 2. 
6. % solution dil. 3. 


Nor in the following combinations : — 


7° 89 CG. 4 (NH,)2SO,, 
10 c.c. CaCl, 0.26 per cent, 
2.¢.c: KiGl}s per cent. 


8. 50 c.c. ¥ ammonium sulphate, 


39 c.c. % cane-sugar, 
to ec: CaCl, 0.26 per cent, 
i €.c, Kel per cent. 


9. 25 c.c. 4 ammonium sulphate, 


to c.c. CaCl, 0.26 per cent, 
Ecc. RGR per cent, 
64 c.c. # Cane-sugar. 


In the controls from the same hearts with sodium chloride in place of 
ammonium sulphate good beats developed in every case. 


96 David J. Lingle. 


VII. ANSWERS TO SOME OF THE OBJECTIONS TO THE SODIUM 
CHLORIDE THEORY. 


Howell! makes the statement that sodium chloride without the 
presence of calcium cannot start beats. This may be true, but the 
experiment the author gives to prove it is not satisfactory. He says 
that when calcium is removed from a strip by means of oxalate, 
treatment with sodium chloride is powerless to start beats; hence it 
cannot do so without calcium. Howell’s experiment is exactly what 
we should expect according to Loeb’s theory, namely, that when cal- 
cium is removed from the tissues in this way, the sodium chloride in 
it not being neutralized is present in sufficient amount to be injurious. 
In other words, the strip is then suffering from an excess of sodium 
chloride. If so, it is clear why the addition of more sodium chloride 
is ineffective; it simply makes matters worse instead of better. This 
idea is based on the supposition that the normal strip contains sodium 
and calcium in proportions that nearly, but not quite, balance each 
other. Calcium is in slight excess and inhibits the strip. It seems 
to be, as Loeb? has shown, calcium, and not potassium, that prevents 
strips from beating in Ringer’s solution and in blood serum. If beats 
are to begin in these this calcium must be overcome by sodium 
chloride of the solution in which it is immersed, or calcium must be 
diminished in the strip. We can always start beats by the former 
method, but in heart muscle the latter is not as yet practicable ® 
because the oxalate, the usual means of getting rid of calcium, seems 
to be actively injurious in other ways. 

In a former paper the series of beats occurring in a strip when it 
is transferred to a sugar or a lithium chloride solution, after beats 
have ceased in a solution of sodium chloride, was explained as due 


to the sugar and lithium chloride permitting the escape of the — 


excess of the sodium chloride that had stopped the beats in the first 
case, so that favorable conditions for beats again arose. Howell? 
brings forward the following objections to this idea: (1) It cannot 
be a sodium chloride effect because the strip still has a supply of cal- 


l HOWELL > G0c, ctz., Pp, 100, 2 LoEB: Loc. cit. 

8 The experiments described on page gt with strips in sugar solutions contain- 
ing a small amount of sodium chloride where beats begin after twenty-four hours’ 
immersion, show that sodium chloride can start beats when, if salts do diffuse out 
at all readily, most of the diffusible calcium has been removed. 

4 Howe. Loe. cif., p. 197. 


semen 


Liportance of Sodium Chloride in Heart Activity. 97 


cium which can be detected by the ammonium oxalate test. But 
according to Howell’s ideas such strips had stopped previously in 
sodium chloride solution because there was not enough calcium left 
in them to keep them going! Now he states that the beats pro- 
duced in these strips in lithium chloride and sugar solutions result 
because there is calcium in them. The same amount of calcium ina 
strip in one case causes a standstill, in another beats. In the sugar 
and lithium chloride solutions calcium diffusion continues, and the 
strip should be passing continuously into less and less favorable con- 
ditions for beats if calcium is essential. Yet under these circum- 
stances beats begin and continue for some time. Here, again, the 
calcium history of a strip and its activity do not run parallel courses. 
The sodium chloride explanation seems simpler and more consistent. 

2. Sometimes strips that have made a long series of beats in 
sodium chloride make but a short series in sugar. The idea is, that 
if it took a long time for the sodium chloride to enter and cause beats 
it should take a long time to escape and so cause a long series of 
beats, but it does not. This fact strongly supports the sodium 
chloride idea, and it agrees with that theory perfectly. A strip which 
has been a long time in sodium chloride will be more or less injured. 
The prolonged action causes permanent injury; hence the return 
series of beats is naturally shorter in such cases. 

3. When a strip is treated with a solution of sodium chloride 
and sodium oxalate for a period equal to the ordinary activity of a 
strip in sodium chloride, a transfer to sugar will not cause beats. 
Possibly this failure results from two things: (a) When treated with 
oxalate in this way, as has been said, the strip is poisoned, or injured, 
by an excess of unneutralized sodium chloride; (b) It is also injured 
by the oxalate. These two causes could together depress irritability 
so that the strip would not beat in sugar. 

4. A strip that has run down in sodium chloride, and has been 
restored in sugar, does not always beat when sodium chloride is 
given it again. The answer to this is the same; such strips have 
been under the influence of sodium chloride a long time, and have 
sustained permanent injury; if they beat at all they should do so 
feebly, as is the case. 


CONCLUSIONS. 


1. Sodium chloride is absolutely necessary for the origination of 
rhythmic activity in heart strips. 


98 David J. Lingle. 


2. Agencies like caffein that can intensify rhythmic activity can- 
not originate it. 

3. What has been described as the sodium chloride arrest is prob- 
ably due to a lack of oxygen in the salt solutions. The presence of 
oxygen in these postpones its development, and starts the rhythms 
again. 

4. Ordinary salt solutions do not contain enough oxygen for nor- 
mal activity of heart strips. 

5. Oxygen gas and sodium chloride, if properly used, will keep 
strips beating as long as a mixture of salt solution. 

6. Oxygen gas has a powerful influence on rhythmic power, but is 
of itself powerless to originate rhythms. 

7. Oxalate solution that precipitates calcium will permit beats to 
begin if sodium chloride is present. 


NOTES ON THE ACTION OF ACIDS AND ACID 
SALTS ON BLOOD-CORPUSCLES AND SOME 
OTHER CELLS? 


By S. PESKIND- 


[From the Physiological Laboratory, Western Reserve University.] 


N connection with a research on the action of numerous laking 
agents, carried on in conjunction with Dr. G. N. Stewart, large 
quantities of corpuscles free from serum were required. Efforts 
were made to obtain them by adding to blood minute traces of 
hydrochloric acid or ferric chloride. These reagents cause the cor- 
puscles rapidly to sink to the bottom of the vessel. It was found 
that, during the subsequent washing with salt solution and centri- 
fugalization, they lake very easily. With the view of discovering 
some way of preventing the laking of the precipitated corpuscles, a 
systematic investigation of the reaction was made. The results 
are presented in the following brief notes. 

1. If to defibrinated human blood, dog’s blood, cow’s blood, or the 
blood of a chicken, there is added a very small quantity of most acids 
or the acid salts of iron, aluminum, zinc, copper, mercury, tin, silver, 
gold, uranium, and molybdenum, an immediate agglutination and 
precipitation of the blood-corpuscles takes place. Neutral salts 
are unable to produce this reaction. 

2. The serum constituents play no part in this reaction, since 
serum-free corpuscles are likewise agglutinated and precipitated by 
the same reagents. 

3. The reaction is due to an effect of the reagents on the stromata 
of the corpuscles, since, if the serum-free corpuscles are laked by any 
hemolytic agent, —¢. g., sapotoxin, —and to this solution of blood- 
corpuscles a trace of any of the precipitants is added, the ghosts 
are agglutinated and precipitated, settling very rapidly. 


1 These experiments were performed during my tenure of the H. M. Hanna 
Fellowship. A more detailed report will be made in a subsequent paper. 


99 


100 S. Peshind. 


4. The constituents of the stromata are, according to competent 
analyses, lecithin, cholesterin, and a globulin which, in all probability, 
exists as alkali-globulin. 

If blood-corpuscles are laked by ether, the lecithin and cholesterin 
are dissolved out of the stromata, yet the addition of any of the 
reagents still causes precipitation of the ghosts. Therefore neither 
the lecithin nor the cholesterin is concerned in the production of 
this reaction. We must, therefore, look to the alkali-globulin of the 
stromata for an explanation of the phenomenon. 

5. The reagents which precipitate the stromata are also precipi- 
tants of alkali-albumin and alkali-globulin, and it seems a warranted 
conclusion that the reaction is due to a combination of the precipi- 
tant with the alkali-globulin of the stromata. 

6. A slight excess of the reagent causes precipitation of the cor- 
puscles, but lakes them very rapidly. 

7. Laking of the precipitated corpuscles may be prevented by 
washing several times with ice-cold saline solution and centrifugal- 
izing after each washing. Keeping precipitated blood at o° C. (z.e., 
without washing) retards laking for a long time. 

8. More than a s/zght excess of reagent prevents agglutination and 
precipitation of the corpuscles. 

g. If corpuscles precipitated by HCl or FeCl, are washed five or 
six times with ice-cold saline solution (centrifugalizing after each 
washing), the corpuscles will then, on being shaken with saline 
solution, resume a practically normal suspension, from which they 
can be reprecipitated by the ordinary reagents. 

10. All acids and acid salts cause precipitation of alkali globu- 
lin, but certain acids and acid salts do not precipitate stromata and 
intact blood-corpuscles. Intact corpuscles were found not to be pre- 
cipitated by osmic acid, boric acid, uric acid, hydrogen sulphide, car- 
bonic acid, arsenious acid. 

The only acid salts which would precipitate intact corpuscles were 
those mentioned in paragraph 1 including the acid sulphates of 
sodium and potassium. 

Why certain acids and acid salts do not precipitate the stromata 
and the intact corpuscles, I cannot at present say. It will perhaps 
be found that the penetrability of the corpuscles for a given acid salt 
determines whether or not this reagent will cause agglutination and 
precipitation of the corpuscles. If this be the case, then we have a 
ready means of determining whether any given metallic base pene- 


Action of Actds on Blood-Corpuscles. 101 


trates the corpuscles, — 7. ¢., by seeing whether the acid salts of -this 
base cause precipitation of the corpuscles. There is reason to believe 
that in this reaction the reagents do not penetrate the corpuscles to 
any great distance, but combine with the envelope and remain close 
to the surface. The envelope is made sticky, which accounts for 
the agglutination. 

11. Leucocytes,! thoroughly washed, are strikingly agglutinated 
and precipitated by the same reagents which act on red _blood- 
corpuscles. The blood separated into two layers,—an upper milky 
layer of serum containing a larger number of leucocytes and abso- 
lutely no red cells, and a lower layer containing the red cells and 
some leucocytes. Some of the upper leucocytic layer was diluted 
with saline solution, centrifugalized, and washed. The sediment, 
consisting purely of leucocytes, was used in my experiments. 

Halliburton has shown that leucocytes contain the same globulin 
(called by him 8 globulin) as the stromata of the red blood-cor- 
puscles. (This globulin is found to have fibrinoplastic properties.) 
The reaction in the case of leucocytes (7. ¢., the agglutination and 
precipitation) is therefore, in all probability, due to a combination 
of this 8 globulin with the reagents. 

12. This 8 globulin is a constituent of all typical cells, —e. 2, 
liver cells, kidney cells, —and reasoning by analogy from the chem- 
ical constitution of the stromata (which form the framework of the 
red blood-corpuscles), I believe that this globulin forms part of the 
framework, or at least the envelope, of all these typical cells. If this 
be so, then the reaction (7. ¢., agglutination and precipitation) must 
show itself to be of a general nature and not confined to blood- 
corpuscles. On trial, this was found to be actually the case. Sper- 
matozoa, yeast cells, bacilli, the mycelia and spores of a fungus, and 
ciliated epithelium, were all found to give the reaction. 

Spermatozoa, washed free from spermatic fluid, are agglutinated 
and precipitated by ferric chloride, although the reaction was not 
obtained with copper sulphate or HCl. 

Yeast cells show marked agglutination and precipitation, if they 
are suspended in saline solution and then treated with a little ferric 
chloride. A solution of peptone in 0.9 per cent saline was inoculated 
with a motile bacillus. After several days the bacteria were centrif- 
ugalized and washed several times with saline. The bacilli, sus- 


1 The leucocytes for my experiments were obtained from leukaemic blood 
kindly given to me by Dr. W. T. Howard, Jr. 


102 S. Peskind. 


pended in saline, were immediately agglutinated and precipitated by 
small amounts of either ferric chloride or hydrochloric acid solution. 

A fungus rubbed up in saline solution gave a suspension of mycelia 
and spores which under the microscope showed practically no clump- 
ing. On the addition of ferric chloride, a flocculent precipitate, con- 
sisting of agglutinated masses of mycelia and spores, came down. 

Ciliated epithelial cells from the larynx of a rabbit, suspended in 
saline solution, are agglutinated and precipitated by ferric chloride. 

13. The specimen of leukzemic blood investigated contained in its 
serum an isolysin, —2z.¢., a substance which caused laking of human 
colored corpuscles. This blood-serum also caused agglutination of 
human blood-corpuscles, as well as agglutination, precipitation, and 
laking of dog’s corpuscles. 


THE BEHAVIOR OF NUCLEATED COLORED BLOOD- 
CORPUSCLES TO CERTAIN HASMOLYTIC AGENTS: 


Ge UN STEWART- 


[From the Physiological Laboratory, Western Reserve University.| 


— some preliminary experiments, as mentioned in 
a former paper,? had shown that nucleated and non-nucleated 
colored corpuscles exhibit the same general behavior towards NaCl, 
NH, Cl, saponin and water, it seemed of consequence to determine with 
as much exactness as possible how close the resemblance really is, for 
the following, among other, reasons: (1) If the agreement turned out 
to be complete it would justify the use of mammalian blood, which 
has hitherto been alone employed in observations on the permea- 
bility of the corpuscles,’ in an investigation designed to throw light 
on the relation between the life and certain of the selective powers of 
cells, since nobody, whatever view he may hold in regard to the mor- 
phology of the non-nucleated corpuscle, can doubt that the nucleated 
corpuscle is a cell. (2) The comparison would not be without inter- 
est in connection with the morphological relationship of the two kinds 
of corpuscles. It would be interesting, ¢. g., to determine whether the 
behavior of the nucleated corpuscles of the blood of the mammalian 
embryo was the same as that of the nucleated hamoglobin-contain- 
ing elements of the red marrow of the adult mammal and the same as 
that of the nucleated corpuscles of the adult bird, and whether all the 
kinds of nucleated corpuscles resembled in their behavior the non- 


1 The major portion of the facts mentioned in this paper was communicated to 
the American Physiological Society at the Chicago meeting, December, 1Igol. 
Additional results were embodied in papers read before the American Association 
of Pathologists and Bacteriologists at Cleveland in March, 1902, and the British 
Medical Association (Section of Anatomy and Physiology) at Manchester, in July, 
1902. 

2 Journal of physiology, 1899, xxiv, p. 211. 

3 The experiments of G. Manca and G. CATTERINA (Archivio di farmacologia 
sperimentale, 1902, i, fasc. ii), on the behavior of the osmotic “resistance” of 
nucleated corpuscles kept for several days after withdrawal from the body, had not 
been published when this was written. 

103 


104 G. N. Stewart. 


nucleated corpuscles of the mammal. Such investigations seemed 
not incapable of throwing light on the genetic relationship between 
them. Since, as I have shown, the selection of NH,Cl in preference 
to NaCl by the corpuscles depends on their physico-chemical structure, 
and not on what we call their life, and since some other cells, as pus 
corpuscles, do not possess this property, the question arises whether 
it is present all through the development of colored corpuscles, or is 
acquired at some definite point. It is obvious, I think, that by in- 
vestigating questions of this kind, not only in regard to the perme- 
ability of NH,Cl, but of urea and all the other substances which easily 
penetrate the fully formed corpuscles, we may notably supplement 
our morphological work and throw light on the ancestry of the 
colored corpuscles, and the nature of the process by which they 
develop from their colorless predecessors. (3) It seemed not unlikely 
that such a comparison might bring to light differences in the physio- 
logical properties of the nucleus and cytoplasm. 

In this paper I shall first describe the behavior of the nucleated 
corpuscles of fowl’s blood to the substances (NH,Cl, NaCl, saponin 
and water) already studied in relation to mammalian blood. The 
behavior of the nucleated corpuscles of embryonic blood and of red 
marrow in regard to these and other substances will be next investi- 
gated. In the third section of the paper, the effects of certain of the 
laking agents on the corpuscles of Vectwrus, which, on account of their 
great size, are especially suitable for microscopic observations, will be 
described. The fourth section will include one or two miscellaneous 
experiments cognate to the general subject. Detailed discussion of 
the modus operandi of the various hemolytic agents is reserved for a 
future communication. 


I. THe BEHAVIOR OF FowL’s BLoop To NH,Ci, NaC, SAPONIN 
AND WATER. 


In Experiment I, page 127, it is shown that the conductivity of 
fowl’s defibrinated blood is diminished by dilution with water to 
a much smaller extent than would be the case if a simple solution of 
electrolytes with the same conductivity as the blood were diluted to 
the same degree. This is still more conspicuous in the case of a 
sediment of blood rich in corpuscles. This agrees entirely with the 
behavior of mammalian blood.! 


1 Cf. STEWART: Journal of physiology, 1899, xxiv, p. 211. 


Behavior of Nucleated Colored Blood-Corpuscles. 105 


In Experiment II the behavior of fowl’s defibrinated blood to 
NH,Cl, NaCl, saponin and water, is compared with that of the defibri- 
nated blood to which formaldehyde had been added. Two specimens 
of formaldehyde blood, one containing 50 per cent more formaldehyde 
than the other, were used in order to determine whether the degree 
of hardening had any influence on the result. In the table, as in all 
the others, “+ NH,Cl” means that an equal volume of NH,Cl 
solution was added to the blood. ‘‘+ NaCl” is the corresponding 
abbreviation for the NaCl solution; ‘‘+ saponin” always means, 
when no quantity is mentioned, that the saponin solution (unless other- 
wise noted, a 3 per cent solution of crude quillaia saponin in NaCl solu- 
tion) was added to the blood in the proportion of 0.4 c.c. of saponin 
solution to 5 c.c. of blood; ‘‘+ water” signifies that the blood had 
twice its volume of water added to it. 

The experiment shows that the nucleated corpuscles, like the non- 
nucleated, exercise a marked preference for NH,Cl as compared with 
NaCl. This is the case also in the formaldehyde blood, although the 
difference is less, especially when the conductivity is measured imme- 
diately after mixture of the salt solutions with the blood. When the 
mixtures of formaldehyde blood with NH,Cl and NaC] are allowed to 
stand, the difference increases continually with the time. There is no 
notable difference between the two specimens of formaldehyde blood 
as regards their behavior to NH,Cl and NaCl, although in both the 
preference for NH,Cl, as compared with NaCl, becomes less pro- 
nounced as the hardening proceeds. On the other hand, the increase 
of conductivity produced by saponin is as great in both specimens of 
formaldehyde blood as in the defibrinated blood, and the relative 
increase of conductivity produced by water is, if anything, greater 
in the formaldehyde blood than in the original defibrinated blood. 


In Experiment III fowl’s blood which had been treated with for- 
maldehyde was followed for twelvedays. At the end of that time the 
blood still behaved very much in the same way as at an earlier period 
in the experiment, the chief difference being, as in the case of mam- 
malian blood, that the preference of the corpuscles for NH,Cl, as 
compared with NaCl, manifested itself more slowly in the later obser- 
vations, presumably owing to the slower penetration of the NH,Cl. 
The preference became very distinct when the mixtures of the salt 
solutions with the blood were allowed to stand a considerable time. 


106 


This 


The 


The 


G. N. Stewart. 


experiment also contains a series of observations in which the conduc- 
tivity of specimens of the defibrinated blood, treated first with formalde- 
hyde, and then after a while with NH,Cl, NaCl, saponin or water, was 
compared with that of mixtures containing the same quantities of blood, 
formaldehyde solution, NH,Cl, NaCl, saponin or water, but to which the 
formaldehyde was added after the other reagents had produced their 
characteristic effects. Certain differences, which, of course, depend on 
the presence of the corpuscles and would not be found if the electrolytes 
of the blood were all in simple, homogeneous solution, reveal themselves. 
The most striking difference is found in the case of the NH,Cl mixtures, 
the combination, blood + NH,Cl + formaldehyde having a distinctly 
smaller conductivity than the combination, blood + formaldehyde + 
NH,Cl. The main reason for this, I suppose, is that the penetration of 
the NH,Cl into the unhardened corpuscles in the first combination is more 
complete, and perhaps its union with some of the constituents of the 
stroma more intimate than in the case of the hardened corpuscles in the 
second. In addition, it is probable that a great part of the formaldehyde, 
which, as a non-conductor, depresses the conductivity of the extra-cor- 
puscular liquid, is in the second case fixed in the corpuscles. 
combinations blood + NaCl-+ formaldehyde, and blood + formaldehyde 
+ NaCl, do not differ much in conductivity. The slight diminution of 
conductivity seen in both cases, when the mixtures are allowed to stand for 
a considerable time, is probably due to a slow penetration of the NaCl 
into the corpuscles, the permeability of which for NaCl, as was shown in 
my previous paper,’ is somewhat increased by formaldehyde, at any rate 
in the first stages of its action. 

combination, blood + saponin + formaldehyde, has a somewhat smaller 
conductivity than the combination, blood + formaldehyde + saponin. 
As mentioned in a previous paper,? a similar difference, but a much 
more pronounced one, was found in the case of dog’s blood, especially 
when, in the second combination, blood which had been acted upon by 
the formaldehyde so long that no laking took place on the addition of 
saponin, was employed. I have already discussed the cause of the differ- 
ence in that paper. I do not know whether the smaller amount of the 
difference between the two combinations in the case of fowl’s blood 
than in the case of dog’s blood is accidental or not. It may be due to the 
fact that in Experiment III, the saponin in the first combination had a 
greater influence than usual in bringing electrolytes out of the corpuscles, 
and thus increasing the conductivity, since it was added in somewhat 
more than the usual amount (in the proportion of 0.4 c.c. of the saponin 


' STEWART: Journal of physiology, 1901, xxvi, p. 470. 


* STEWART: Journal of experimental medicine, 1902, vi, p. 257- 


Behavior of Nucleated Colored Blood-Corpuscles. 107 


solution to 4 c.c. of blood, instead of 0.4 c.c. of the saponin solution to 
5 c.c. of blood). I have shown ' that the action of saponin on formalde- 
hyde blood is to increase the conductivity of the corpuscles, without 
necessarily causing any of their contents to pass bodily into the serum. 

The combination blood + water + formaldehyde has a smaller conductivity 
than the combination blood + formaldehyde + water, especially when the 
measurement is made a considerable time (e.g. twenty-four hours), after the 
addition of the formaldehyde. ‘The result is to be attributed to the de- 
pressing effect of the blood-pigment and the excess of formaldehyde in the 
liquid portion of the water-laked blood on its conductivity, which more 
than compensates the influence of any electrolytes brought out of the 
corpuscles by the water. 


In Experiment IV the behavior of heat-laked fowl’s blood to the 
four tests is investigated. It will be seen that, as in dog’s blood, there 
is only a slight difference in the conductivity of specimens of blood 
to which NH,Cl and NaCl have been added. Only a slight difference 
in the permeability of the ghosts for the two salts can accordingly 
exist. What difference there is is in favor of the NH,Cl. Since the 
nuclei are still present in the heat-laked blood, and are little if at all 
altered in microscopical appearance, the experiment indicates that 
these exercise no noticeable preference for NH,Cl, as compared with 
NaCl. Of course it may be that the temperature necessary for heat- 
laking has deprived the nuclei of this property. But there is other 
evidence, including the behavior of the unheated nuclei in ammonium 
chloride solutions, that there exists in this respect a sharp contrast 
between the nuclear and the extranuclear portion of the intact cor- 
puscles. It is perhaps worthy of mention that the small difference 
which appears to exist in the permeability of the ghosts to NH,Cl and 
NaCl is greater soon after the addition of the salts than later on. I 
have noticed the same thing in dog’s blood. The explanation may be 
that the ghosts take up NH,Cl in preference to NaCl, but later on, 
perhaps on account of some action of the NH,Cl itself on the envelope 
or stroma, become incapable of holding it. Even in unlaked blood, 
both fowl’s and dog’s, the minimum conductivity of the NH,Cl mix- 
ture, and the maximum conductivity of the NaCl mixture, are reached 
soon after mixture. Later on, there is a tendency, although it never 
goes very far, for the conductivity of the former to increase a little, and 
for that of the latter to fall a little. Even very shortly after mixture, 


1 STEWART: Journal of experimental medicine, 1902, vi, p. 257. 


108 G. NV. Stewart. 


in the case of heat-laked blood, the difference is always insignificant, 
in comparison with that in ordinary defibrinated blood or formalde- 
hyde blood. On the other hand, saponin and water produce a great 
relative increase in the conductivity of heat-laked fowl’s blood, just 
as in heat-laked dog’s blood. When heat-laked fowl’s blood is di- 
luted with water, the diminution of the conductivity is relatively 
smaller than when the unlaked blood is correspondingly diluted. 
A remarkable effect of heat-laking in fowl’s blood is that its con- 
ductivity may be greatly diminished by the laking, ¢. g. in Experi- 
ment IV, in one specimen of blood from 49.41 to 34.07, and in the 
other from 49.41 to 29.75. A diminution is often seen in mamma- 
lian blood also, but it is never, in my experience, so great as this. 
Another point of difference is that fowl’s blood after heat-laking is 
much thicker and more viscid than dog’s blood. 

The fowl’s defibrinated blood used in Experiment IV was the same 
as that used in Experiment III, and it will be seen that in Experiment 
IV the addition of 0.4 c.c. of the saponin solution to 5 c.c. of the heat- 
laked blood, whose conductivity was 34.07, raised the conductivity to 
66.19; while in Experiment III the addition of the same amount of 
the saponin solution to 4 c.c. of the unheated blood, whose conduc- 
tivity was 49.32, raised the conductivity only to 69.26. This shows, 
pretty clearly, I think, that the final conductivity after the action of a 
sufficient amount of saponin to cause its full effect, depends on the 
quantity of electrolytes originally present in the serum and corpuscles, 
and not on their distribution at the moment the saponin is added. 

Before concluding this section of the paper, I may mention that the 
serum and defibrinated blood of the fowl have, on the average, a con- 
siderably higher conductivity than dog’s blood and serum. For eight 
fowls, the average conductivity of the serum was 93.61, and of the de- 
fibrinated blood 47.81; while for observations on thirty dogs, taken at 
random out of a series embracing considerably over one hundred, the 
average for the serum was 82.71, and for the blood 33.00. If the rela- 
tion between the proportion of serum and corpuscles to the conduc- 
tivity of the blood and serum is the same for fowl’s as for mammalian 
blood, this would indicate that the average proportion of serum is 
higher in the fowl’s blood. 


Behavior of Nucleated Colored Blood-Corpuscles. 09 


II. THe NuCLEATED CORPUSCLES OF THE BLOOD OF THE EMBRYO 
AND OF THE BONE MARROW. 


Embryo’s blood. —I have not been able to obtain a sufficient quan- 
tity of embryonic blood for freezing-point determinations. Measure- 
ments of conductivity could, of course, be made with a very small 
amount of blood, but I preferred to devote such scanty supplies as 
came into my hands to microscopic investigation. The observations 
here recorded are mainly qualitative. I hope to have another oppor- 
tunity of making quantitative measurements of the relative resistance 
to laking of the various kinds of corpuscles, etc. The following is a 
typical protocol. 


Killed a pregnant albino rabbit by decapitation. Collected and defibrinated 
the blood. Immediately opened the uterus and exposed four embryos, 
6.5 cm. long. They showed lively movements. Wiped them dry with 
blotting paper, then cut the umbilical cord and decapitated each in turn, 
collecting and defibrinating the blood. The blood of the embryos only 
yielded a scanty amount of fibrin. From the maternal blood, fibrin was 
obtained in normal quantity. 

Action of sapotoxin. —Twenty-five minutes after the death of the mother, a 
small drop of the fcetal blood was placed on a slide, and at alittle distance 
from it a small drop of a 2 per cent solution of sapotoxin (Schuchardt) in 
0.9 per cent salt solution. A coverslip was gently laid on the two drops. 
The sharp interface between the blood and sapotoxin solution was now ob- 
served with the microscope. Rapid laking of the colored corpuscles was 
seen to goon. Numerous ghosts are left. The nuclei of the corpuscles are 
not well stained by methylene blue (in 0.9 per cent NaCl) either before 
or after laking. They seem to break up into granules which stain with 
methylene blue, and this is the case after all the methods of laking tried. 
This is in marked contrast to the bird’s corpuscles, the nuclei of which 
stain very well after laking, and retain their original shape, and usually 
also their original position in the ghosts, the oval outline of which may be 
made out around them ; although, occasionally, the nucleus is displaced and 
may lie with its long axis in the short axis of the corpuscle, or even be 
extruded from it altogether. When the 2 per cent sapotoxin solution was 
stirred up with the foetal blood, complete laking occurred immediately. 
The same was true of the maternal blood. 

Sodium taurocholate (a 2 per cent solution in 0.9 per cent salt solution) 
also laked both at once. But it caused a greater disintegration than 
the sapotoxin, and only granular colored masses could be made out, 
and not complete ghosts. In view of the fact mentioned in a previous 


G. NV. Stewart. 


paper’ that the taurocholate is a more intense leucolytic agent than 
sapotoxin, it would be of interest to make a quantitative comparison of 
the relative resistance of embryonic and mature erythrocytes to these 
substances. Hitherto, however, I have not been able to do this. 


Dog’s serum which had stood for ten days in the cold (at a temperature of 5° 


to 8° C.) had practically no effect on the maternal blood. A very small 
amount of hemoglobin was seen after twenty-four hours to have gone into 
solution, the mixture having been kept at ordinary temperature. On the 
foetal blood, there was a somewhat greater laking effect, and under the 
microscope it could be seen that a few of the corpuscles were decolorized. 
Fresh dog’s serum (obtained twenty hours after the blood was drawn, from 
blood-clot kept in the cold) caused (at room temperature) complete, or 
nearly complete laking of maternal and fcetal blood, although the laking 
proceeded slowly. Under the microscope, it was seen that the nucleated 
corpuscles discharged their hamoglobin. But the nuclei did not seem to 
stain on the addition of methylene blue (in salt solution), only some of 
the ghosts showing a diffuse staining, or a number of scattered stained 
granules. 


Water \akes the foetal blood in the same way as the maternal. A watery solu- 


tion of methylene blue also, of course, causes complete laking. In the 
foetal blood, numerous blue granules of various sizes are seen, some free, 
others apparently lying inside the ghosts. ‘The nuclei seem to break up 
into these granules. When methylene blue in watery solution is added to 
foetal blood, to which an excess of 4 per cent formaldehyde solution in 0.9 
per cent NaCl solution was added twenty-four hours previously, a diffuse 
staining of the corpuscles is caused, the nuclei not being well defined. 
Ehrlich’s triacid stain causes the nucleated formaldehyde-hardened cor- 
puscles to assume a reddish color, the nuclei being lighter than the rest of 
the corpuscles. 


A 1 per cent solution of NH,C7 completely lakes both maternal and feetal 


blood. About four volumes of NH,Cl solution were added to one of 
blood. In both cases there is ‘a period of resistance,” during which the 

corpuscles maintain their shape and color. Then they become large and 

perfectly round, while still retaining their hemoglobin. Then, at about the 

same moment, great numbers of them lose their pigment rather suddenly. 

Some corpuscles retain their pigment a considerable time longer than the 

majority. This is true both of the maternal and foetal blood, and indeed 
of all kinds of blood when submitted to the action of all the laking agents, 
I have used. Eventually these more resistant corpuscles fade out too. 

A o.9 per cent or a 1 per cent solution of NaCl causes no laking either in 

the maternal or feetal blood. The embryonic corpuscles, then, like the 

adult, are freely permeable to NH,Cl and not to NaCl 


! STEWART: Journal of medical research, 1902, iii, p. 268. 


ee eee er 


Behavior of Nucleated Colored Blood-Corpuscles. 111 


Amyl alcohol in small amounts lakes foetal and maternal blood with equal 
readiness. 

Heat-laking. — Both maternal and foetal blood mixed with several volumes of 
0.9 per cent NaCl are laked on being heated to 63°-65° C. In 
the foetal blood there is almost complete destruction of the corpuscles, 
only a few ghosts being seen, and no more being revealed by the addition 
of methylene blue. Numerous granules of blood pigment are seen in 
clumps. In the blood of the adult mammal the ghosts persist after heat- 
laking. No nuclei could be seen after aking in the foetal blood. ‘This is 
in contrast to the behavior of bird’s blood. 


Bone marrow. — Red marrow was obtained from the proximal epi- 
physis of the femur of a young dog. Portions of it teased in 0.9 per 
cent NaCl solution were examined under the microscope and sub- 
jected to the action of various laking agents. 


Sapotoxin. — With 2 per cent sapotoxin solution in 0.g per cent NaC] solution, 
all the cells containing hemoglobin, large and small, nucleated and 
non-nucleated, are quickly decolorized, and some of them completely 
broken up. 

Water also decolorizes all the heemoglobin-containing cells. 

NH,Cl in 1 per cent solution also causes laking of all the colored elements, 
while NaCl solution of the same strength has no such effect. We must, 
therefore, conclude that the ancestors of the colored blood corpuscles in 
the marrow, both erythroblasts and normoblasts, exhibit, at any rate from 
the moment they acquire hemoglobin, the preference for NH,Cl, as com- 
pared with NaCl, which is so characteristic of the adult corpuscles. 

After all three methods of laking, methylene blue stains the nuclei of the nor- 
moblasts and erythroblasts. The outline of the ghost can be easily dis- 
tinguished, enclosing the nucleus, which is often eccentric in the laked 
corpuscles. 

Blood from a case of pernicious anzemia. — Blood was obtained, after death, 
from the kidneys and lungs of a case of pernicious anzemia in which, dur- 
ing life, large numbers of nucleated colored corpuscles were present in the 
blood stream. There was pronounced diminution in the total number of 
red corpuscles, and also a diminution in the number of white corpuscles, 
although in proportion to the red corpuscles the white were increased. 
Many of the colored corpuscles were crucible-shaped, with a cup-shaped 
depression at one end which might simulate a nucleus. Many were 
highly crenated. A good many were larger than normal. All the 
colored corpuscles are pervious to NH,Cl, since they are laked in a 1 
per cent solution of that salt. They are also permeable to urea, being 
laked in a 1 per cent solution of it. They are impermeable to NaCl, 


112 G. NV. Stewart. 


lake in the normal way with sapotoxin, and lake also when suspended in a 
1 per cent NaCl solution and heated to 62°-64° C. 

Bone marrow from a case of pernicious anemia. — Some red bone 
marrow from a case of pernicious anaemia was also examined. Colored 
corpuscles decidedly larger than the normal (macrocytes) were rather 
plentifully present. Nucleated red corpuscles of ordinary size were also 
seen. All the colored elements were laked by sapotoxin, sodium tauro- 
cholate, 1 per cent NH,Cl solution, and t per cent urea solution. The 
large colorless marrow cells did not seem to be destroyed easily by any 
of the laking agents used. Sodium taurocholate, which is so efficient a 
leucolytic agent for the leucocytes of mammalian blood, exerted no 
special effect on them. 


Ill. THe ActTIon oF LAKING AGENTS ON THE COLORED 
CORPUSCLES OF NECTURUS. 


The biood was obtained by making a snip in the heart through the 
skin. Collected in this way, it clots with great readiness, but by 
vigorous stirring could be sufficiently defibrinated for microscopic use. 
The copious and viscid skin secretion, which seemed to be increased 
while the animal was struggling, was wiped off before the incision 
was made, so as to prevent it from mingling with the blood. 

One of the most striking characteristics of the blood, and one 
which, in addition to the great size both of the colored and of the 
colorless corpuscles, renders it particularly useful for the investi- 
gation of the action of hzmolytic agents, is the ease with which the 
hemoglobin crystallizes, both outside of, and in the corpuscles. 
This occurs spontaneously when the blood is allowed to stand for 
some days at room temperature, and instantly, on the addition of 
sodium taurocholate, sapotoxin, Loffler’s methylene blue (owing, 
mainly at any rate, to the alcohol in it), water and watery solutions 
of substances like NH,Cl, which readily penetrate the corpuscles. 
In extraglobular crystallization the crystals usually assume the form 
of massive elongated rhombic prisms (Fig. 2). This is the case also 
in intraglobular crystallization, if the haemoglobin is well diluted, or a 
portion of it liberated from the corpuscle before the rest crystallizes, 
as is apt to be the case when crystallization is induced by the addi- 
tion to the blood of water (Fig. 1), which, passing into the cor- 
puscles, causes them to swell. It may be here remarked, in passing, 
that the fact that water induces intraglobular crystallization, is of 
itself enough to show that the hemoglobin in the corpuscles cannot 
be in the form of an ordinary watery solution, since the dilution of 


— i. SS 


— a ee ee 


Behavior of Nucleated Colored Blood-Corpuscles. 113 


such a solution with additional water would hinder instead of deter- 
mining crystallization. 

With agents like sodium taurocholate (Fig. 4), or Léffler’s 
methylene blue (Fig. 19), which often cause crystallization of the 
whole or the greater part of the hemoglobin in a corpuscle before 
there is time for the liberation of any or much of it, and which do not 
cause the corpuscles to swell and become globular, as water does, the 
intraglobular crystals are apt to be more irregular in shape, being 
apparently moulded, by their mutual pressure and by the contour of 
the corpuscle, into polygonal masses. The genesis of these masses 
can be watched under the microscope. After the addition of sodium 
taurocholate (a 2 per cent solution in 0.9 per cent NaCl solution was 
always used), although a few corpuscles become broader in the trans- 
verse diameter, most of them preserve approximately their original 
dimensions. Longitudinal wrinkles, and less frequently transverse 
wrinkles, appear in many corpuscles, due apparently to the puckering 
of the envelope for whose existence evidence will be presently pro- 
duced. The hzmoglobin-containing substance of the corpuscle breaks 
up into round or oval globules of different size. There may be seven 
or eight or more in one corpuscle. The edges of the globules are 
faint at first, but rapidly grow more distinct, the outlines at the same 
time becoming less round and smooth, and ultimately angular. 
Either the whole mass of the hemoglobin passes bodily from the col- 
loid to the crystalline condition, or, if we assume the existence Of a 
stroma in which the hemoglobin is contained, and for the existence 
of which there seems to be a considerable amount of evidence, the 
relation of the pigment to the stroma is altered by the hemolytic 
agent and the hemoglobin crystallizes zz sztz in thestroma. Pressure 
on the coverslip does not generally displace the crystals, which are 
evidently kept in position in some way in the corpuscle, probably by 
the stroma. As the crystals grow more distinct, the outline of the 
nucleus does so too, while the border of the corpuscle becomes 
fainter, although it can often be seen, still in its original position, 
after complete formation of the crystals, following their angles and 
sinuosities so as to include them all. Later on, the envelope may 
completely disappear, as if dissolved by the bile salt. When a drop 
of blood and a drop of the taurocholate solution are allowed gradu- 
ally to mingle on a slide, complete laking, with no intraglobular crys- 
tallization may be observed near the interface, where the solution of 
the laking agent is strongest, and where accordingly the haemoglobin 


II4 G. N. Stewart. 


is liberated at once, before it can crystallize in the corpuscle. A 
little farther from the interface, typical intraglobular crystallization 
may be observed. Sometimes the intraglobular crystals are very nu- 
merous, small, and granular looking. Sometimes the opposite extreme 
is seen, nearly the whole of the pigment in a corpuscle crystallizing in 
one mass, in which the nucleus is imbedded. The nucleus may par- 
tially project from the crystalline mass (Fig. 22), or it may be com- 
pletely enclosed in it. In this case, the ends of the corpuscle are apt 
to be colorless, as if the haemoglobin here had been liberated 
before, it had time to crystallize inside (Fig. 20, 22). It may indeed 
not infrequently be seen that the poles of the corpuscle are earli- 
est attacked by hemolytic agents, and sometimes one before the 
other. A fraying out of one pole, as if it were being corroded, is 
sometimes observed. Occasionally, however, the opposite may be 
seen, namely, acrystal of hemoglobin occupying each pole of the cor- 
puscle, the rest being entirely colorless and filled up with the swollen 
nucleus (Fig. 28). The taurocholate does not at first affect the size 
of the nuclei, which continue to show clearly the granules that 
represent the nodes of the intranuclear network. Later on, the nuclei 
become much swollen, though still oval in shape, filling the whole 
corpuscle, except a small portion at the two poles. The intranuclear 
network now has the appearance of fine instead of coarse granules, as 
at first. That there is something (resistance of the nuclear mem- 
brane to distention or resistance of the stroma pushed towards the 
poles) which even in the completely laked ghost prevents the nucleus 
from filling the whole corpuscle, is well shown in Fig. 26, 27. 
When swelling of the nucleus takes place after the haemoglobin has 
crystallized in the corpuscle, the crystals may be seen to separate 
farther from each other as the nucleus swells, being doubtless mechan- 
ically forced apart. The same thing may be observed under the 
influence of other agents, and even in hardened corpuscles, when no 
crystals are present, a fragmentation or fissuring of the corpuscle 
may be produced by the swelling of the nucleus. Sometimes the 
taurocholate may cause complete disappearance of the nucleus after 
swelling. 

Sometimes the corpuscles show a curious twist at one or both ends. 
It was seen that this was caused during the spinning of the cor- 
puscles. Probably the rigidity of the corpuscle is lessened by the 
bile salt, and bending under stress is apt to take place where it is 
thinnest, namely, at the poles. The twist may be seen both in cor- 


ee SE eee 


Behavior of Nucleated Colored Blood-Corpuscles. 115 


puscles which retain their haemoglobin (Fig. 21), and in completely 
laked corpuscles (Fig. 26, 27). 

The leucocytes swell little, if at all, under the influence of the bile 
salt, although they may become dim. 

Water. — Water causes well-marked agglutination of the colored 
corpuscles before laking. The corpuscles then swell and become 
globular. Some of the nuclei may escape from the ghosts and swell 
up, the network disappearing. They then stain, as all swollen nuclei 
do, diffusely and faintly with methylene blue. Inthe corpuscles which 
have not yet lost their hemoglobin, the nuclei stain deeply and show 
the network. The best way to get intraglobular crystallization with 
water, is to put not too large a drop of water on a slide, and let it 
gradually mix with the blood under the slip. In some corpuscles, a 
single hemoglobin crystal forms, the pigment in the rest of the cor- 
puscle remaining for a time uncrystallized. When this is the case, the 
corpuscle is pale in the immediate neighborhood of the crystal, indi- 
cating either that all the pigment from this portion of the stroma has 
been gathered into the crystal, or that the liberation from the cor- 
puscle of a portion of the hemoglobin of this part of the stroma has 
gone hand in hand with the crystallization of the rest. Sometimes 
when a crystal stretches right across a corpuscle, the ends may be seen 
to be curved in correspondence with the boundary of the corpuscle, 
and not to project through it. In some corpuscles the whole of the 
haemoglobin may be seen to be crystallized. As already remarked, 
this cannot be reconciled with the theory that the hemoglobin exists 
in the corpuscles in ordinary aqueous solution. Indeed I do not think 
that anybody who studies the influence of reagents on these magnifi- 
cent objects, and particularly the phenomena of intraglobular crystal- 
lization, can believe that the corpuscles are merely little bags or 
vesicles containing hemoglobin solution, as some physiologists seem 
to suppose. 

Marked swelling of the leucocytes is caused by water. Their nuclei 
also swell, and many of the granules in the protoplasm of the leuco- 
cytes seem also to become distended into vesicles. The leucocytes 
of Necturus blood are normally of great size and numerous relatively 
to the red corpuscles. After the action of water, some of them may 
become even larger than the swollen erythrocytes, and as many as 
_ $ix or more swollen nuclei may be seen in some of them. 

Urea in I-per cent solution in water rapidly lakes the red corpuscles. 
The nuclei swell and their boundary becomes very distinct, irresistibly 


116 G. N. Stewart. 


suggesting a membrane. The laked corpuscles are also much swollen. 
In partially laked corpuscles, what look like radial strize passing out 
into the corpuscle from the nucleus may be seen. A similar appear- 
ance may be observed in water-laked corpuscles stained with methy- 
lene blue, the striae being blue:(Fig. 3). These striae might possibly 
be thought to represent fine channels by which the nucleus gets its 
nutriment from the blood-plasma. This would seem, especially in 
such large corpuscles, advantageous to the nucleus, whose metabolism 
is presumably more intense than that of the rest of the corpuscle. 

Sapotoxin (2 per cent solution in 0.9 per cent NaCl solution was 
always used) causes rapid laking with copious formation of crystals 
of various shapes outside the corpuscles. Some are “knife rest” 
shape, some almost cubical. The nuclei of the erythrocytes remain 
intact, are not swollen, and show the normal coarsely granular appear- 
ance of the intranuclear network. They stain well with methylene 
blue. Ghosts can, as a rule, be seen surrounding the nuclei, and their 
boundary suggests a membrane sometimes frayed. Sometimes an 
actual gap in the stroma can be seen, the envelope being continued 
over the gap. Sapotoxin causes most of the leucocytes to disappear, 
but the large coarsely granular ones are left. 

Amyl alcohol produces rapid laking without special feature. 

Loffler’s methylene blue. — Its action in causing intraglobular crys- 
tallization has already been described. It need only be added that 
immediately after the crystals have appeared in the corpuscle the en- 
velope of the latter may be seen enclosing them. It afterwards disap- 
pears from some of the corpuscles. Intraglobular crystallization may 
take place in corpuscles whose nuclei are not as yet stained in the 
least. This indicates that the crystallization is caused by the alcohol. 
Many of the corpuscles are completely laked. A large granular pre- 
cipitate, the granules of which are stained blue, is thrown down around 
and on the still unlaked corpuscles. Some of the ghosts seem to dis- 
appear entirely, and give rise to the granular precipitate, as heaps of 
granules may be seen around the deeply stained nuclei, and also in 
corpuscles which are not yet laked, though globular. The whole ex- 
tranuclear part of the corpuscle becomes granular, the granules re- 
sembling exactly in color and size those outside. Sometimes what 
looks like an envelope may be seen extending partly around a mass 
of granules of the same shape and size as a corpuscle. Many of the 
erythrocytes are seen to become globular, namely those which do not 
show crystals in their interior and which still retain their haemoglobin. 


eee ee Eee 


Behavior of Nucleated Colored Blood-Corpuscles. 117 


Heat-laking. -—- The temperature necessary for heat-laking seems to 
be a little less than that required for mammalian blood. In one ex- 
periment nearly all the corpuscles (suspended in 0.9 per cent NaCl 
solution) were laked at 58°, and all at 59° C. At 51° C. there was no 
laking. A specimen from another Necturus laked at 60°, but not below 
it, e.g. not at 59°. It was kept ten minutes at 60°. In neither case 
did the liquid become quite transparent, even when all the haemoglobin 
was discharged. The ghosts were particularly well preserved, and had 
a granular appearance, as if heat-coagulation of the stroma had taken 
place. They still retained the oval shape of the intact corpuscle. The 
nuclei occupied their normal position in the ghosts. Sometimes the 
ghosts exhibited a “ waist” opposite the middle of the nucleus, as if 
the envelope of the ghost had there become tacked to the nucleus. 
Methylene blue in 0.9 per cent NaCl solution stained the nuclei well, 
although the intranuclear network was not quite so distinct as in the 
intact corpuscles. The envelope and stroma did not stain. But ina 
specimen heated in Necturus serum to 36°, some of the granules in 
the ghosts stained slightly. The addition of sapotoxin (2 per cent 
in 0.9 per cent NaCl solution) to the ghosts, after staining with 
methylene blue, caused much of the stain to come out. Some- 
times a portion of the stroma seemed to escape or become dissolved, 
and the gap could then be seen to be bounded by an envelope (Fig. 
17). The addition of Léffler’s methylene blue now caused deeper 
staining of the nuclei, but no other change. 

Spontaneous laking. — In four days at room temperature, some of the 
corpuscles were seen to be laked, and some hemoglobin crystals were 
present outside of them. It is not now so easy to obtain intraglobular 
crystallization by the addition of laking agents, the hemoglobin being 
very readily brought out of the corpuscles and then crystallizing out- 
side. Some intraglobular crystallization is, however, caused, ¢. g. by I 
per cent NH,Cl solution. Intraglobular crystallization may be very 
easily got, if a little of the blood is allowed to dry partially on a slide 
without addition of anything. Many of the corpuscles also seem to 
become Jaked. The addition ofa drop of 0.9 per cent NaCl solution 
to the partially dried blood causes instant laking of practically all the 
corpuscles which still retain their hamoglobin. The nuclei are often 
surrounded by groups of hemoglobin crystals. These phenomena are 
not seen if the blood is mixed with the NaCl solution before it has 
dried at all. 

In blood which had stood nine days, and was long since sponta- 


118 G. N. Stewart. 


neously laked, a considerable number of round bodies colored with 
haemoglobin were seen. They are much smaller than the original 
erythrocytes, but seem to represent them and their nuclei, as may be 
shown by adding Loffler’s methylene blue, which deeply stains the 
nuclei as well as the envelope (Fig. 15, 16). 

I have already described! certain changes produced in mammalian 
corpuscles fixed with formaldehyde by treating them with ammonia 
and heating, and in sublimate-fixed corpuscles, by heating them in 
water or by treating them with H,S or (NH,),S. Still more interest- 
ing results have been obtained with Necturus corpuscles hardened in 
formaldehyde, Hayem’s solution and osmic acid. 

Sublimate-hardened corpuscles.— Addition of H,S to corpuscles 
hardened for forty-eight hours in excess of Hayem’s solution, and then 
washed many times with water, causes laking of the corpuscles. The 
nuclei are not swollen, and the intranuclear network remains unaitered. 
The ghosts become faint, but do not disappear. Addition of a little 
NH,, after the H,S has acted, at once causes great swelling of the 
nuclei, which remain oval in shape and become very indistinct. The 
nuclei still stain pretty well with Loffler’s methylene blue. 

Addition of (NH,),S to the washed sublimate corpuscles causes 
liberation of the hemoglobin. The nucleus at first remains distinct 
and surrounded by a black border, which is sometimes divided into 
small portions of about uniform size by narrow light lines. Possibly 
this appearance is produced by the beginning of the swelling of the 
nucleus leading to distention of the structure represented by the 
dark border. The nucleus now swells greatly, and often appears to 
burst the corpuscle at one side, so that the ghost assumes the appear- 
ance of a “bowler” hat (Fig. 23, 24). This may also be seen in cor- 
puscles which still retain their haemoglobin or some of it. The 
phenomenon was first observed after the addition of (NH,),S, but it 
may be at once remarked that NH, also produces it, and save in one 
point to be mentioned presently, it is doubtful whether (NH,),5 solu- 
tion acts in any other way than by reason of the NH, init. The ap- 
pearances depend on the amount of (NH,),S or NH, added _ If 
much is added, the nuclei swell and become exceedingly indistinct, 
as do the ghosts. All the stages may be traced at different distances 
from the edge of the slide, if a drop of the reagent is run in under it. 
The nucleus seems to begin to swell before the rest of the corpuscle 


! STEWART: Journal of medical research, /oc. cé?. 


EEO OEE OEOEOEEEeEEEEOOOEEOEOEOEOEOEeEEEeEeEEee—eeeeeeeee ee ee 


Behavior of Nucleated Colored Blood-Corpuscles. 119 


is affected. If the action of the (NH,),S is gradual, the nucleus may 
swell without rupturing the corpuscle, which swells also to some ex- 
tent and loses haemoglobin. The boundary of the nucleus is picked 
out in black, either as a continuous line or as the interrupted line 
already described. The black intranuclear network opens out its 
meshes (Fig. 12, 13) as swelling of the nucleus advances. The hat 
appearance may not be seen at all at one part of the slide, while all 
the corpuscles at another part may exhibit it. The addition of Loffler’s 
methylene blue instantly causes deep staining of the nucleus, which 
at once contracts greatly in size, as does also the whole ghost, the 
original shape of the corpuscle being resumed. In some of the cor- 
puscles, an envelope stained blue can be made out. 

When the sublimate corpuscles are first acted on by NHsg, and then 
by Loffler’s methylene blue, the envelope is not so well revealed as 
when (NH,),S is used. The best method of all is to add first H,S, 
then NH,, and then Loffler’s methylene blue. To a drop of a suspen- 
sion of the corpuscles in water, a drop of H,S solution of equal size 
is added, and then a much smaller drop of NH,. A large amount of 
NH, seems to disintegrate the corpuscles too much, and they disap- 
pear. Ldoffler’s methylene blue is now added. A small amount of it 
stains the nucleus without coloring the rest of the corpuscle or its en- 
velope. A larger quantity causes the nucleus to stain intensely blue. 
The envelope of the corpuscle and the nuclear membrane are also 
colored blue (Fig. 10). While the great majority of the nuclei 
shrink to about normal size on the addition of Loffler’s methylene 
blue, some remain much swollen, and are colored diffusely and much 
more faintly than the others. It seemed as if there was no envelope 
around the corpuscles whose nuclei did not shrink. This suggests 
that the swelling of the nucleus in these cases might have been so 
great that the envelope was ruptured, and that the osmotic process 
on which the shrinking of the corpuscles and nucleus depends, could 
not take place in its absence. As a matter of fact, the shrunken cor- 
puscles exhibit numerous wrinkles all over their surface, the wrinkles 
appearing more deeply stained than the rest of the corpuscle. In an 
endeavor to produce heat-fixing of the envelope without destroying 
the power of the nucleus to expand and to shrink, some of the washed 
sublimate corpuscles were heated to 60° C. in water for two minutes. 
The hemoglobin was not discharged. There was, however, in a few 
corpuscles, either some swelling of the corpuscle or shrinking of the 
nucleus, since the latter was separated from the surrounding stroma 


120 G. N. Stewart. 


by a concentric space. The nuclei swelled on the addition of (NH,),S 
as readily as if the corpuscles had not been heated. 

Some corpuscles were heated to 65° C. for fifteen minutes. The 
haemoglobin was not discharged. The addition of (NH,),S caused 
swelling of some of the nuclei, but not of all. Addition of H,S caused 
the corpuscle to become pale, and its boundary, which was originally 
black and sharply marked, to become indistinct. H,S caused no 
swelling of the nuclei, which, however, either became tinted with the 
blood-pigment, or showed the tint better when the corpuscles were 
rendered pale, since they were distinctly yellow. The intranuclear 
network was still distinct after addition of H,S. Some corpuscles 
were less affected by H,S than others. The addition of NH, now 
caused all the nuclei to swell enormously. The corpuscles also 
swelled somewhat, but perhaps only passively, owing to the disten- 
tion of the nucleus, which practically filled the whole corpuscle. The 
corpuscles did not become spherical, but still retained their oval shape. 
Addition of Loffler’s methylene blue now brought out the envelope 
and nuclear membrane (Fig. 5, 6, 7, 8, 9). 

The preliminary heating of the corpuscles to 65° C. did not seem 
to make much difference in the result. Presumably, therefore, there 
is no heat-fixing at this temperature of the portions of the corpuscle 
which are “ mobile” under the influence of the reagents used. 

To some of the washed sublimate corpuscles, H,S, NH,, and Loffler’s 
methylene blue were added in succession, with the results just stated, 
and then sodium taurocholate (in NaCl solution). The blue was 
partly discharged from all the corpuscles. What looked like a par- 
tially separated envelope (Fig. 11) seemed to grow less distinct 
under the influence of the bile salt, although it did not completely 
disappear. The taurocholate caused in many of the corpuscles 
renewed swelling, after the shrinking produced by the methylene blue 
solution. Addition of NH, now caused further swelling, with prob- 
able disappearance of some corpuscles. 

Formaldehyde-fixed corpuscles. — The corpuscles were fixed in 4 per 
cent solution of formaldehyde in 0.9 per cent NaCl solution, then 
washed thoroughly with water, and suspended in water. NH, causes 
the same change of color and of spectrum as in dog’s formaldehyde 
corpuscles.’ Some of the corpuscles swell somewhat in the cold. 
On heating to 60° C., they are considerably swollen and less distinct. 


} STEWART: Journal of medical research, /oc. cét. 


— 


Behavior of Nucleated Colored Blood-Corpuscles. 121 


The nucleus is not swollen, and is as distinct as before. The contrast 
between the behavior of the formaldehyde and the sublimate cor- 
puscles is very striking, and indicates a marked difference in the 
structures fixed or in the mode of fixing in the case of the two hard- 
ening agents. 

Although the Necturus corpuscles look decidedly pale under the 
microscope, after being heated to 60° in water containing NH, it is not 
difficult to prove that the laking is mainly apparent. One way of doing 
this is to filter the corpuscles off from the solution, when it is seen that 
the blood-pigment is nearly all in the corpuscles. Another method is 
to cause the corpuscles to shrink, ¢. g., by the addition of a solution of 
hydroxylamine hydrochlorate. They become at once smaller, more dis- 
tinct, and as deeply tinged with blood-pigment as before. Matthes,! 
in connection with the fact discovered by him, that dilute HCl 
brings the pigment out of sublimate-fixed dog’s and frog’s corpuscles, 
mentions an experiment in which he supposes that on standing the 
ghosts may again take up the hemoglobin from the liquid. The same 
possibility suggested itself to Dr. S. Peskind and myself, when we saw 
formaldehyde-fixed dog’s corpuscles apparently lake under the influ- 
ence of NH, and heat, and sublimate-fixed dog’s corpuscles apparently 
lake when heated in water, and yet observed the ghosts in both cases 
become again tinged with blood-pigment on the addition of hydroxyl- 
amine hydrochlorate. We proved clearly? that when this happened 
the laking was only apparent, the corpuscles being swollen and the 
blood-pigment, therefore, more dilute in the interior of the corpuscle. 
Dr. Peskind, while confirming the statement of Matthes that dilute 
HCl in a certain strength does very easily lake sublimate-fixed mam- 
malian corpuscles, which swell enormously, has since shown that if 
the swollen corpuscles are caught at a certain point, at which they ap- 
pear free from hemoglobin under the microscope, and their suspen- 
sion in water appears laked to the eye, they can be caused to shrink 
by hydroxylamine hydrochlorate, when their color returns. At this 
stage also it can be shown, by filtering off the swollen corpuscles, that 
they contain most of the blood-pigment. If this critical point be 
allowed to pass, the corpuscles discharge their hemoglobin, and their 
color cannot be restored by hydroxylamine hydrochlorate. Many of 
them break up entirely. I find that Necturus corpuscles fixed by 
Hayem’s solution are rapidly decolorized by 0.2 per cent HCl, without 


1 MatTTHES: Miinchener medicinische Wochenschrift, April 29, 1902. 
2 STEWART: Journal of medical research, /oc. cit. 


122 G. N. Stewart. 


appreciable swelling either of the corpuscle or its nucleus. Many of 
the ghosts disappear if more than the minimum amount of acid needed 
for laking be used. Some shrinking of the ghosts is caused by 
hydroxylamine hydrochlorate, but I could not demonstrate any return 
of the hemoglobin color. 

The Necturus formaldehyde corpuscles which have been heated to 
65° C. in ammoniacal water, stain deeply with Loffler’s methylene 


blue, the whole corpuscle staining well and the nucleus more deeply. - 


When only a trace of the methylene blue is added, only the nucleus 
stains, the rest of the corpuscle remaining unstained. The intra- 
nuclear network is well seen. There is a great difference between 
ordinary formaldehyde corpuscles, and those treated in the way de- 
scribed, in the facility with which the extranuclear portion stains. 

Formaldehyde corpuscles suspended in water containing N Hg, and 
heated several minutes to 60° C., and one minute to boiling, be- 
came pale and considerably increased in size. The nucleus was not 
materially increased in size. The intranuclear network was not so 
distinct as in ordinary formaldehyde corpuscles, nor did Léffler’s 
methylene blue bring it out so distinctly, although the nucleus, as well 
as the rest of the corpuscle, stained deeply. In some of the corpuscles, 
after the methylene blue, a finely serrated or scalloped outline could 
be seen. 

Osmic acid-hardened corpuscles. — The corpuscles were hardened in 
I per cent osmic acid for two days, and then washed many times with 
water. Sodium taurocholate produces no change in them. NH, 
causes considerable swelling of the corpuscles, which become corre- 
spondingly paler and less distinct, though the blood-pigment does not 
seem to come out. The shape always remains oval. The nucleus 
swells a little, and the network becomes less distinct. Loffler’s methy- 
lene blue causes some of the corpuscles to become smaller, and stains 
the extranuclear portion an intense blue, nearly as deep as that of the 
nuclei (Fig. 18). Many of the corpuscles do not shrink, but some 
shrink very greatly, becoming markedly smaller than the normal cor- 
puscle. The border of the nucleus is very prominent, in contrast to 
that of the corpuscle. 

On heating washed osmic acid corpuscles in ammoniacal water to 
boiling, partial laking seems to take place. The corpuscles are greatly 
swollen and indistinct. Some of them are fissured. The nuclei are 
quite distinct, and show the network, 


ee OE 


Behavior of Nucleated Colored Blood-Corpuscles. 123 


Jo 


IV. MISCELLANEOUS OBSERVATIONS ON MAMMALIAN CORPSUCLES. 


Is the power of NH,Cl to penetrate the corpuscles dependent upon a 
toxic influence ?— I have shown! that NH,Cl only penetrates perfectly 
fresh corpuscles, at least in notable amount, after a relatively long 
‘period of resistance.” It might be supposed that during this period 
the NH,Cl exerted an injurious influence on the envelope of the cor- 
puscle, which changed its natural properties, and that it is only 
through an injured envelope that NH,Cl is capable of passing. Ex- 
periments V, VI, and VII, were designed to throw light on this point, 
the idea in all being to see whether the NH,Cl, when it exists in less 
than a certain concentration in the serum, is unable to penetrate the 
corpuscles. If the permeability of the corpuscles for this salt 
depends upon a poisoning of them by it, this, of course, might be 
expected to be the case. 

In Experiment V the behavior of blood which had stood for forty 
hours in the cold, after being drawn, was investigated with regard to 
a mixture of the solutions of NaCl and NH,Cl employed in the pre- 
vious experiments, in the proportion of one volume of the NH,Cl 
solution to nine volumes of the NaCl solution. In spite of the fact 
that the conductivity of the NH,Cl— NaCl mixture is greater than 
that of the NaCl solution, the conductivity of the blood, after the 
addition of the mixture, is less than that of the blood to which the 
NaCl solution has been added. The reason, no doubt, is that NH,Cl 
has penetrated the corpuscles. Of course the conductivity of the 
blood to which the ordinary NH,Cl solution has been added, is smaller 
still in comparison with that of the NaCl blood, since the whole of 
the dissolved substance added to it is capable of entering the cor- 
puscles, while in the mixture only about one-tenth of the dissolved 
substance added, at most, can do so. 

In Experiment VI, perfectly fresh blood was treated with the ordi- 
nary NH,Cl and NaCl solutions, mixtures of the NH,Cl and NaCl 
solutions, and solutions of NaCl to which, instead of NH,Cl solution, 
the corresponding volume of water had been added. In all cases 
there is evidence that some of the NH,Cl penetrates the corpuscles. 
Since, however, the conductivity of the blood after the addition of 
NH,Cl — NaCl mixture is always markedly greater than after the 
addition of the corresponding NaCl — water mixture, it would appear 


1 STEWART: Journal of physiology, 1901, xxvi, p. 476. 


124 G. N. Stewart. 


that the whole of the NH,Cl added to the serum never penetrates the 
corpuscles, or at least does not become bound there in such a form 
as to prevent it from taking part in the conduction of the current. 
It seems not improbable that the presence of NaCl in the serum hin- 
ders, to some extent, the penetration of the corpuscles by NH,Cl, or, 
in other words, exerts a protective influence on the corpuscles, but 
further observations, including investigation of the sera separated 
from the mixtures, would be necessary to determine this point. 

In Experiment VII, the conductivity of perfectly fresh blood to 
which quantities of NH,Cl, ranging from about I gm. per 100 c.c. 
of blood to about 0.02 gm. per 100 C.c. 
of blood were added in watery solution, 
was compared with that of the blood to 
which approximately equal quantities of 
NaCl were added. It was supposed that 
if the NH,Cl penetrates in virtue of 
some toxic power, there should be some 
abrupt change in the curve of conduc- 
tivity at the point where, with increas- 
ing concentration of the NH,Cl, this 
power first becomes manifest. As is 
seen in Figure I, in which the conduc- 
tivities plotted from the results of this 
experiment are laid off along the verti- 
cal, and the number of c.c. of NH,Cl 
and NaCl solution added to 15 c.c. of 
blood along the horizontal axis, no such effect can be observed. 

A further argument in favor of the view that this is not the ex- 
planation of the permeability of the corpuscles for NH,Cl is found in 
the observation that this salt, when added in substance to blood, does 
not produce laking, even when the blood is saturated with it. The 
laking action of watery solutions of NH,Cl appears, therefore, to be 
essentially the same as water-laking. 


FIGURE I. 


Effect of amyl alcohol on sublimate-fixed corpuscles. — When dog’s 
corpuscles which have been fixed by Hayem’s solution, and thoroughly 
washed, are suspended in water and treated with amyl alcohol, the 
blood-pigment after a time is entirely liberated and goes into solu- 
tion in the water. Mercury is liberated at the same time, and may 
be detected in the solution of haemoglobin. Ammonium sulphide 
throws down a reddish flocculent precipitate insoluble in water, and 


Behavior of Nucleated Colored Blood-Corpuscles. 125 


evidently containing all the blood-pigment. On washing this pre- 
cipitate thoroughly with water, although all the pigment is still 
present, no mercury can be demonstrated in it. Laking by amyl 
alcohol does not succeed, if the sublimate-hardened corpuscles have 
been treated with 0.9 per cent NaCl solution for some time before 
the addition of the amy] alcohol. 


SUMMARY OF RESULTS. 


Section I. 1. Nucleated blood corpuscles, like non-nucleated, exer- 
cise a marked preference for NH,Cl as compared with NaCl. This 
is the case even when the blood has been treated with formaldehyde, 
although the difference is less, especially when the conductivity is 
measured immediately after mixture. When the mixtures of for- 
maldehyde blood with NH,Cl and NaCl are allowed to stand, the 
difference increases continually with the time. Even when the 
substances were added to blood on which formaldehyde had already 
acted for twelve days, the difference was very distinct, when the 
measurements of conductivity were made fifteen hours after mixture. 
The hardening action of the formaldehyde, accordingly, appears to 
render the penetration of the NH,Cl into the corpuscles slower, or 
their power of binding it less, although it does not abolish their 
characteristic property of taking up NH,Cl in preference to NaCl. 
The preference is equally marked in stale and in fresh blood. In 
bird’s blood, therefore, as in mammalian blood, the preference does 
not depend on the life of the corpuscles. 

2. Saponin produces a notable increase in the conductivity of fowl’s 
blood, as it does in the conductivity of dog’s blood. An equal, or 
even a somewhat greater, increase is produced in both kinds of 
blood, when saponin is added after the blood has been hardened by 
formaldehyde. This shows that the increase of conductivity is not 
dependent on, or associated with, the escape of the hemoglobin from 
the corpuscles, since in the formaldehyde blood the hemoglobin is 
fixed and the saponin does not cause laking. As in the case of mam- 
malian blood, the increase of conductivity is due to an increase inthe . 
permeability of the corpuscles to electrolytes. 

3. The conductivity of fowl’s blood is diminished by dilution with 
water to a much smaller extent than would be the case if serum 
or a solution of electrolytes were correspondingly diluted. This 
is due to the participation of the electrolytes of the corpuscles 


126 _G. N. Stewart. 


in the conduction of the current after the addition of water. The 
relative increase of conductivity is quite as well marked in formal- 
dehyde-hardened as in fresh blood. Here, too, there is evidence 
that the electrolytes of the corpuscles participate in the conduction. 

4. Heat-laking of fowl’s blood is caused at about the same tem- 
perature as that of mammalian blood. The nuclei are not destroyed. 
The conductivity of fowl’s blood may be markedly diminished by 
heat-laking, unlike that of mammalian blood. The action of sapo- 
nin on the ghosts of heat-laked fowl’s blood causes a marked 
increase of conductivity, and the action of water a marked relative 
increase of conductivity, just as in the case of mammalian blood. 

Section II, 1. The nucleated colored corpuscles of the blood of the 
mammalian embryo behave in the same way as the non-nucleated 
corpuscles of the adult, to most of the laking agents investigated 
(sapotoxin, sodium taurocholate, amyl alcohol, water, foreign serum). 
On being heated to 63° —65° C., they lake like the adult corpuscles, 
but, unlike them, leave few ghosts behind. Their nuclei also seem 
to be broken up by heat-laking, unlike those of bird’s corpuscles. 

2. All the hemoglobin-containing elements of the red bone mar- 
row of a young mammal were found to behave in the same way as 
the adult mammalian colored blood corpuscles to the agents investi- 
gated (sapotoxin, water, solution of NH,Cl). 

The same was true of the colored corpuscles of the blood of a case 
of pernicious anemia to sapotoxin, NH,Cl solution, urea solution, and 
heat. 

Section III. 1. Intraglobular crystallization of the hemoglobin of 
Necturus blood is very readily obtained by the action of various 
hemolytic agents. The observations on this point show that the 
hemoglobin cannot exist in the corpuscles in ordinary aqueous 
solution. 

2. The various laking agents do not cause similar changes in the 
shape and size of the corpuscles and their nuclei. | 

3. Necturus corpuscles hardened by Hayem’s solution are laked by 
H,S without swelling of the nucleus. Ammonia causes great swelling 
of the nucleus. 

4. By treating Necturus corpuscles hardened by Hayem’s solution 
successively with H,S, NH,, and Loffler’s methylene blue, and in 
other ways, an apparent envelope can be demonstrated. 

5. Necturus corpuscles fixed by formaldehyde swell under the 
influence of NH, and heat, while the nuclei retain their original size. 


Behavior of Nucleated Colored Blood-Corpuscles. 127 


The same is true of corpuscles fixed by osmic acid. These facts 
indicate the existence of differences in the nature and point of attack 
of the fixation by different hardening agents. 

Section IV. The permeability of the colored corpuscles for NH,Cl 
does not depend on a toxic effect of the salt on the corpuscle. 


EXPERIMENT I. 


Effect of water on defibrinated blood, and a sediment of defibrinated blood rich in 
corpuscles. The blood was obtained from fowls by inserting a cannula into the carotid 
or jugular. The blood was stirred for a long time (as much as thirty minutes) in order to 
make sure that it would not clot. 


} oo ie ee 
ost Oreo 
Azcl aa 
‘ lw os | meee |% oF 
Defibrinated blood. r} eet Sediment. Pm fecda= 
Oe = | | ¢Otrat ane 
gE° | age 
oA | Mxsa 
| 
Hen’s defibrinated blood | 48.28 Sediment from cock’s de- | 
fibrinated blood (after 
+ 1 vol. water leecdeodll L216 centrifugalization ) 17.25 | 
+ 1 vol. water (not 
+ 3 vols. water (cor- laked) LOM] |) = 169 
puscles only par- 
tially decolorized) 17.84 | 2.70 + 3 vols. water (laked) | 10.42) 1.65 
+7 vols. water 112.43} 3.88 + 7 vols. water 7.84 | 2.20 
Sediment of the defibri- | serum (from clot) 91.67 
nated blood | oO:oL 
After settling a longer | 
time | 22.63 
Serum (from clot) 102.35 


1 As in all the tables, X is an abbreviation for A (5°) & 10%, the conductivity (at 
5° C.) being expressed in reciprocal ohms X 10°. 


128 G. NV. Stewart. 


EXPERI 


April 138, 1901. At 1.50 P.M. obtained the blood of four fowls by decapitation and 
defibrinated it. At 2.52 p.m added to 50 c.c. of the defibrinated blood 15 c.c. of an 8 per 
cent solution of formaldehyde in NaCl solution. Call the mixture A. On April 14 at 
11.40 A. M. added to another quantity of the defibrinated blood one-fifth its volume of the 


Defibrinated blood. Time. Formaldehyde blood. A.. A 
Defibrinated blood 46.43|| 4.09p.m.| A (already brownish) | 55.70 
dato SS eae 56.58 
3.05 “ .| Defib. blood + NH,Cl.! T2200 | IAS INES CI: 
r 7.244 “ jane 88.02 
313 « Not laked, but a little Afterstanding 13 hrs. | 74.34 
darkened 57.35 Top 99.80 
348 “ Sie 56.33|| 7.36 “ | A+ NaCl. 
7.383 “ Hoes 91.52 
3.06 “« Defib. blood + NaCl | Afterstanding13 hrs. | 91.85 
| Top | 116.82 
Gil eae eel otc 85.93}| 6.573 “ | A + Saponin. 
7.11 “ | Not laked nor dark- 
3.07% “ Defib. blood + saponin.! ened | 68.33 
After standing 14hrs. | 68.14 
215 Be Distinctly darker. Bottom 64.52 
7.56 “ | A-+ as much of NaCl 
3.20 “ | Quite dark, but not per- sol. used in making 
fectly laked. saponin sol. as was 
3.28 ‘ | Fairly well laked 54.99 added of the sapo- 
nin solution 59.09 
4.01 “ | Completely laked 59.93 After 38 hrs. Top | 93.54 
6.554 “ | A -+ water. | 
tees Defib. blood + water.t 7.04 “ Not laked nor dark- 
ened 29.00 
3.39 « Not completely laked 21.99 After standing 143 h. | 28.97 
| : Top 33.75 
| A of defibr. blood _, u April 14 
dof diluted blood ~~" 10.02 a.m. A 57.88 
Serum fr’m A (free fr’m | 
| Percentage of serum in e blood-pigment) 94.24 
defeated blood, | 10.37“ | A + NH,Cl. 
63.3. 10:40)“ 89.26 


After standing 20} h. 

Not laked at all 81.31 

Top (free from blood- 

pigment, but some- 

what turbid) 105.19 

10:45 “ | A-+ NaCl. 


After standing 20} h. | 91.85 
Top (free from blood- 
—_—__—_ —— pigment, not turbid) | 119.02 


Fal &¢ a ; 

1 As inall the experiments, “+ NH,Cl,” aac a | 67.24 
“+ NaCl,” without further comment, may : After standing 20} h., | 
be interpreted as meaning that to the __ no laking 67.78 
blood its own volume of the solution of Top | 95.30 
: wae hi Bottom (separated by 
NH,Cl or NaCl was added. “+ saponin centrifuge. Very 
means that 0.4 c.c. of the saponin solu- || sticky and thick) | 28.58 


tion was added to 5 c.c. of the blood. ||11.20 “ | A + water. | 

a “ / 2 
+ water” means that toa given volume | 11.24 After standing 20% hrs 
of blood two volumes of distilled water ; a raking: a 29.72 


were added. Top 35.01 


“ 


- 


Behavior of Nucleated Colored Blood-Corpuscles. 129 


MENT II. 


8 per cent formaldehyde solution. Call this mixture B. For the NH,Cl and NaCl solu- 
tion, whose effect on the blood was compared, A = respectively 142.41 and 139.30; for the 
NaCl solution used in making the formaldehyde solution, AX = 13930; for the saponin 
solution, A = 83.94; for the NaCl solution used in making saponin solution, \ = 78.97. 


Time. | Formaldehyde blood. B. | 


| 
Time. | Defibrinated blood. 


| = 
April 14 | April 14 | 
10.22 A.M. | Defibrinated blood | 45.51 || 11.49 a.m. | B 53.26 


| Serum from defib. blood | 93.54 | 12.15 p.m. | B-+ NH,Cl. 


12.15 P.M. | Defib. blood + NH,Cl. ||12.24 “ Considerably dark- 
ened and seems par- | 
12.24 * Little, if at all dark- | tially laked. 
ened. After standing 223} h. | 71.01 
After standing 223h.! | 58.41 /|12.17 “ | B+ NaCl. 
12.17 “ | Defib. blood + NaCl. | After standing 224-h | 87.41 
12.19 ‘* | B + saponin. 
After standing 224 h. | 83.39 
12.19 “ | Defib. blood + saponin. 12.25 “ | Somewhat darkened 


but not laked. 
12.25 “ | Quite dark and par- | After standing 24 hrs. | 62.44 
tially laked. Very thick, so that 
After standing 2+hrs. | 69.64 some air-bubbles | 
12.21 “ | Defib. blood + water. got into U-tube, and 
this Ais too low. | 
17.25°- “6 Partially laked. 12.01 “ | B+ water. 


After standing 25 hrs. | 20.71 
12.25 “ | Darkened but not laked. | 
After standing 25 hrs. | 26.30 


12 p.M.| Serum from B 95.30 
Biel 0) es lle; 53.37 
3.344 “ | B+ NH,CI. 


cy Me AS | 88.02 
After standing 20 hrs. | 79.22 

3.42 “ | B+ NaCl. 

3.444 *° | 89.89 


After standing 20 hrs. | 91.20 
3.51 “ | B+ saponin. 


ios) 
mn 
aS 
an 
mn 
Lo) 
iS 


After standing 20 hrs. | 67.96 
4.00 “ B + same amount of 
s NaCl sol. used in 
making saponin sol. 
as was added of 
saponin solution 54.05 
After standing 20 hrs. | 56.46 
4.07 “ | B + water. 


1 Unless otherwise mentioned, the 
blood-mixtures were kept, for all periods 
exceeding a few hours, in a cold room 
whose temperature varied from 5° C. or 
less to about 8° C. 


120 G. N. Stewart. 


EXPERIMENT III. 


April 19, 1901. Fowl’s blood obtained by decapitation at 2.30 p. M., and defibrinated. 
At 7.20 p. M. added to 50 c.c. of the blood 25 c.c. of a 4 per cent formaldehyde solution in 


NaCl solution. Call the mixture A. 
For the NH,C! solution A = 142.41; for the NaCl solution A = 139.30. 
For the saponin solution A = 137.03; for the formaldehyde solution A = 112.17. 


Time. Defibrinated blood. r Formaldehyde blood. A. 


April 19 
7.43 p.M.| Defibrinated blood 


April 20 
2.23 p. M. | Serum from defib. blood | 2.12 p.M.| Serum from A (only the 
(only very faint trace faintest trace of 
of Hb in it) blood-pigment in it 
April 21 April 21 
8.55 A.M. | Defibrinated blood | §.42A.M.| A 
9.04 4 c.c. defib. blood + 4 c.c. 9.04 “ |6 cc. of A + 4 ce. 
NH,Cl. NH,€1 

Not laked but darker 9:55 one 

than defib. blood + 

NaCl No laking 


After standing 26 Snoc 
hours more After standing 21] hrs. 
To 4 cc. of the mixture longer 
of defib. blood and Top 
NH,Cl added 1 c.c. || 9.08 a.M.| 6c.c. of A+ 4c.c. NaCl 
of the formaldehyde 
solution. 
This mixture 
After standing 24 hrs. | 65.84}/ 11. S08 
4 c.c. defibrinated blood After standing 23 hrs. 
+ 4¢.¢, NaCl. longer 
: Top 
.|6 cc. of A + 04 c.c. of 


saponin solution. 


After standing 27 
hours more 
To 4 c.c. of the mixture Saae 
of defib. blood and After standing 24 hrs. 
NaCl added 1 c.c. longer 
of the formaldehyde Top 
solution. 3 ‘ |6cc. of A + 04 c.c. of 
This mixture 93.5 the NaCl solution 
After standing 24 hrs. | 87. used in making the 
4 c.c. defibrinated blood saponin solution. 
+ 0.4 ¢.c. saponin. Rc 
Well laked. | After standing 26 hrs. 
longer. 
69.26 Top 
16 cc. of A+ 8 cc. of 
69.26 | water. 


Behavior of Nucleated Colored Blood-Corpuscles. 


Time. 


EXPERIMENT 


Defibrinated blood. 


10.39 A.M. 


EB.O: 

12.10 P. M. 

12.25...“ 
Z.34 “ 


ie Ny At 
9.13. A.M. 
10.57. “ 


2.17 P.M. 


2.46 “ec 


4 c.c. defib. blood + 0.4 
c.c. of the NaCl so- 
lution used in mak- 
ing the saponin 
solution. 


4 c.c. defib. blood + 0.4 
c.c. saponin. 


To 4.4 c.c. of the mix- 
ture of defib. blood 
and saponin solu- 
tion added 2 c.c. of 
the 
solution. 

This mixture 


4 cc. defib. blood + 8 | 


c.c. water. 


After standing 28 hrs. 

Top 
To 6c.c. of the mixture 
of defib. blood and 


water added 1 c.c. | 


of the formaldehyde 
solution. 

This mixture 

After standing 24 hrs. 


Top 


formaldehyde 


56.46 


68.33 


I1L1.— (continued ). 


Time. |Formaldehyde blood. A.| 
10.48 “ Spa 
After standing 26 hrs. 
longer 
| Top 
Mayl | 
3.35 P.M.| A 
S05. | 6c.c. of A + 0.4 c.c. of 
saponin solution. 
£005 
TO 
405 “ |6c.c. of A+ 0.4 c.c. of 
the NaCl solution 
used in making the 
saponin solution. 
410 “ 96 fe 
4.22 “ |2 cc. of A +133 ce. 
NH,CE: 
After standing 15 hrs. 
Ae 1 2°C.c. 108 JAY -- 1.33),6.6)| 


NaCl. 
After standing 15 hrs. 


131 


132 G. N. Stewart. 


EXPERIMENT IV. 


April 22, 1901. Heat-laked fowl’s blood. The blood was the same as that used in 
Experiment III. The solutions of NH,Cl, NaCl, and saponin were also the same. 


Time. Heat-laked fowl’s blood. 


6.24 P. M. Defibrinated fowl’s blood . 

6.00 Heated some of the blood for fifteen minutes to 60° to 
64° C. Becomes dark and laked. Is thicker than laked 
dog’s blood. 

The heat-laked blood 

Heat-laked blood + NH,Cl. 

"After standing 14 hours . 

Heat-laked blood + NaCl. 


"After standing 14 hours 


Heat-laked blood + saponin. 


Heat-laked blood + water. 


Heated another specimen of the blood to 60° C. It darkens 
and lakes, but less readily than at 64°. It is also thick. 
It begins to darken a little above 50°. 

This specimen of heat-laked blood . 


Behavior of Nucleated Colored Blood-Corpuscles. 133 


EXPERIMENT V. 


Defibrinated blood obtained from a bitch forty hours before, and kept in cold. 


The defibrinated blood 
NH,CI solution . 
NaCl solution. 


NH,Cl-NaCl mixture (containing 1 vol. of EE solution 
to 9 vols. of NaC] solution. é 


To 5 c.c. blood added 5 c.c. of the NH,Cl—-NaCl mixture. 


Is not noticeably darkened or ake 
After standing 22 hours longer . 


To 5 c.c. blood added 5 c.c. of the NaCl solution. 


"After standing 22 hours longer ; 
To 5 c.c. blood added 5 c.c. of the aateely solution. 


" Now much darkened : . 
After standing 22 hours longer : 


Fresh defibrinated blood obtained from a dog one hour before the beginning of the 
experiment. 


Time. 


G. N. Stewart. 


EXPERIMENT VI. 


The defibrinated blood 
NH,CI solution . 
NaCl solution. 


NH,Cl-NaCl mixture containing 1 volume of the NH,Cl 
solution to 9 volumes of the NaCl solution (Mixture A) 


NH,Cl-NaC] mixture containing equal volumes of Bie 
and NaC] solutions (Mixture B) neers : 


A mixture containing 9 volumes of the NaCl solution to 1 
volume of water (Mixture A’) : 


A mixture containing equal volumes of the NaCl solution 
and of water (Mixture b’) . 


To 10 c.c. of blood added 10 c.c. of Mixture B. 
Slightly darkened 
After 24 hours a little Hb i in colution: in serum 


To 10 c.c. blood added 10 c.c. of Mixture B’. 

After 24 hours serura distinctly red, although not nearly 
so much Hb in it as in the serum of the blood to which 
its own volume of the NH,Cl solution was added 

To 10 c.c. blood added 10 c.c. of the NH,Cl solution. 

Darkened and laked in a few minutes a Le 

After 24 hours. 


To 10 c.c. blood added 10 c.c. of Mixture A. 


After 24+ hours serum not red 


To 10 cc. blood added 10 c.c. of Mixture A’. 


Not at all darkened . : 
After 24 hours no Hb in serum . 


To 10 c.c. blood added 10 c.c. of the NaCl solution. 


After 24 hours no Hb in the serum 


20.63 
136.29 
130.67 


131.35 


13135 


118.46 


67.96 


57.88 
57.61 


61.38 
63.06 


66.36 
67.78 


Behavior of Nucleated Colored Blood-Corpuscles. 135 


EXPERIMENT VII. 


Defibrinated blood obtained from a dog forty-two minutes before the beginning of the 
For the fresh defibrinated blood, AX = 30.47; thirty hours later, 29.82, and 
For the serum obtained from clot, A = 98.97. 


For the NH,Cl solution, A = 136.29 


observations. 
fifty-three hours after it was drawn, 30.32. 
Percentage of serum in defibrinated blood, 55.4. 


for the NaCl solution, \ = 130.67. 


Interval between Interval between 


2 C.c. of mixture and mixture and 
Quantity NaCl or | measurement of aS of measurement of US of 
of blood NH,Cl | dof NH,Cl1 mixt. NH,Cl A of NaCl mixt. NaCl 
in C.c. solution 


added. = 2 
Hrs. Min. Hrs. Min. 


mixture, mixture. 


136 G. N. Stewart. 


EXPLANATION OF FIGURES.! 


Fics. 1 to 28 were drawn with Leitz Oc. 1V, Obj. Pantachromatic 3 mm. Tube not drawn 
out. All except 1-4 were drawn in outline with the camera lucida. For 5-19 the 
paper was at the level of the stage; for 20-28 at the level of the foot of the micro- 
scope. In reproduction 5-28 are reduced to about }; 1-4 to 3. 

1. Intraglobular crystallization produced by a watery solution of methylene blue. - 

2. Hzmoglobin crystals formed outside the corpuscles after addition of a 2 per 
cent sapotoxin solution in 0.9 per cent NaCl solution. 

3. A water-laked corpuscle stained with methylene blue, showing fine radial striz 
extending out from the nucleus. 

4. Intraglobular crystallization produced by 2 per cent solution of taurocholate 
of sodium in 0.9 per cent NaCl solution. ‘The envelope of the corpuscle has apparently 
disappeared between the hemoglobin crystals, but dark streaks due to the wrinkling 
of the envelope are seen over the crystals, either because the wrinkled envelope still 
persists in this situation, or because the wrinkling present during the formation of the 
crystals caused them to become correspondingly grooved. 

5. A corpuscle fixed by Hayem’s solution, then thoroughly washed with water, 
then heated in suspension in water to 65° C., then treated with H,S, then with NHs, 
and then with Loffler’s methylene blue. 

6. A corpuscle treated in the same way as in 5. It shows nuclear membrane and 
also envelope of the corpuscle. 

7. A corpuscle subjected to the same treatment, showing envelope of corpuscle 
and a gap from which the nucleus has escaped. 

8. A corpuscle, treated in the same way, showing rupture of the envelope by swell- 
ing of the nucleus. 

9. A corpuscle, treated in the same way, showing the envelope partly detached. 

10. A corpuscle fixed by Hayem’s solution, washed, treated with H.S, then NHsg, 
then Loffler’s methylene blue. It shows a bulging as if the envelope was tacked 
at one point to the nucleus. 

11. A corpuscle fixed by Hayem’s solution, washed, then treated with HS, NU, 
and Loftler’s methylene blue, and then with 2 per cent solution of sodium taurocholate 
in 0.9 per cent NaCl solution, and then again with NH;. The nucleus fills the 
whole swollen corpuscle, and therefore is faintly stained. At one side the nuclear 
membrane and the envelope of the corpuscle seem to be detached, the former being 
the more deeply stained. 

12. A Hayem-fixed corpuscle washed with water and then treated with NHg. It 
shows great swelling of the nucleus and opening out of the meshes of the intranuclear 
network. The corpuscle preserves its elongated shape, and the nucleus does not 
encroach on the poles. 

13. A corpuscle treated in the same way as in 12. It is more globular than the 
one shown in 12; the intranuclear network is more strongly marked, and nearly the 
whole corpuscle, including the poles, is occupied by the nucleus. 

14. A washed Hayem-fixed corpuscle treated with NH, and then with Léffler’s 
methylene blue. The latter caused the swollen nucleus to shrink. At one side the 
envelope seems to be tacked down to the nucleus. 


-" 


' 
: 
j 
} 
} 
Hy 


‘ 1 am indebted to my pupil Mr. Samuel S. Berger for drawing Figs. 1 to 4. 


‘oo == es 


Behavior of Nucleated Colored Blood-Corpuscles. 137 


* 


138 G. NV. Stewart. 


EXPLANATION OF FIGURES — (continued). 


15. A very small body with the characters of a colored corpuscle, left after spon- 
taneous laking. Stained with Léffler’s methylene blue. The envelope seems to be 
ruptured at one side, and the nucleus has escaped. 

16. A small corpuscle after spontaneous laking, stained with Loffler’s methylene 
blue. What seems to be its nucleus lies beside it. 

17. Fresh blood-corpuscle suspended in 0.9 per cent NaCl solution and heated to 
60° C. Then treated with methylene blue in 0.9 per cent NaCl solution, and then 
with 2 per cent solution of sapotoxin in 0.9 per cent NaCl solution. A gap is seen 
near one end, which is bounded by the envelope. 

18. A portion of a corpuscle fixed by osmic acid, washed with water, then treated 
with NHg, then heated and then treated with Loffler’s methylene blue. Where the 
corpuscle is fractured the nuclear membrane seems to protrude. It is stained a 
deep blue. 

19. Fresh corpuscle suspended in 0.9 per cent NaCl solution, treated with L6ffler’s 
methylene blue. Intraglobular crystallization of the hemoglobin has taken place. 
At two points on opposite sides of the nucleus there is no hemoglobin, the pigment 
having apparently been dissolved out before it could crystallize. Here what appears 
to be the stroma is seen bounded by the envelope. 

20. Fresh corpuscle treated with the sodium taurocholate in salt solution. The 
haemoglobin is still retained in the greater part of the corpuscle. 

21. A fresh corpuscle treated as in 20. The corpuscle is twisted at both poles, 
where the outline of the envelope can be traced, forming a loop. 

22. A fresh corpuscle treated in the same way. The hemoglobin remains in the 
whole corpuscle except at the poles. 

23. A washed Hayem-fixed corpuscle treated with (NH,),S. (NHg has the same 
effect.) The nucleus has swollen enormously and apparently burst through one side 
of the corpuscle, which has assumed the form of a hat. 

24. A corpuscle treated as in 23. The nucleus has burst through both sides of the 
corpuscle, though the greater part of it protrudes through one side. 

25. A corpuscle treated as in 23, seen on the flat. The swelling of the nucleus 
seems to have caused fissures in the corpuscle. ‘The same thing is seen when NHsg is 
used instead of (NH,4).S. The envelope of the corpuscle is not usually visible across 
the fissures, perhaps because it is dissolved by the N Hs. 

26. Fresh corpuscle stained with a solution of methylene blue in 0.9 per cent NaCl 
solution, then treated with Na taurocholate. The haemoglobin is completely dis- 
charged. The ghost is twisted at both poles. The nucleus is not excessively swollen 
and is therefore pretty deeply stained. 

27. A fresh corpuscle treated as in 26. The nucleus is more swollen than in 26 
and more faintly stained. 

28. Fresh corpuscle treated in same way. ‘he nucleus fills the whole corpuscle 
except at the poles, each of which is occupied by a hemoglobin crystal. 


Soe EFFECT OF DIMINISHED EXCRETION OF 
SODIUM CHLORIDE ON THE CONSTITUENTS 
OF THE URINE. 


By R. A. HATCHER ann TORALD SOLLMANN. 


[From the Pharmacological Laboratory of Medical Department of Western Reserve Uni- 
versity, Cleveland, Ohio. | 


CONTENTS. 


Page 

MEmBETILL GG UCLONY Mag 8 arate NPs rea crete SoMa, | cfg) ao anh Be eee tr 139 
MGIC ALT HISLONICS. 2 a0 oh. i lertacl os: yd no he a) ote ces) wer ee oe wn wee LS 
MEMEECLHOUS OMANALYSIS | coc Ua! soc cst es ee es Sy ee ee ce os ee, AO 
umlealculationGe resultS “7-4 § 2 3 6 © 8 ee ol a Te ee ew ts Se FS 4 
TRESS i O. Bie Aer er ee ae aera ES er ae es fh ee] (| 
SII SESSION OLMESULES I ee) we, esc ac Ore, SN ee le Bee, te os 45 
Ci poNGix Lables andjcurves, (oe 5 2.5 8 sew ee Ss Oe wt ee 4G 


I. INTRODUCTORY. 


re is well known that the chloride content of the urine is notably 

diminished in febrile diseases. This alteration in the composition 
of the urine should be accompanied by changes in the composition of 
the other urinary constituents, if the physical theories of urine 
formation hold true. These changes, if they occur, would probably 
throw some light upon the mechanism of chloride-retention. A 
minute study of the composition of chloride-free urines seemed there- 
fore desirable. These considerations led us to undertake the research 
recorded in this paper. The courtesy of Dr. L. W. Ladd enabled us 
to work on two cases of typhoid fever at Lakeside Hospital, Cleveland. 
The urines of these patients were almost devoid of chlorides, and 
were therefore well adapted to our purpose. The main results agree 
so well in both cases, that we may consider them as typical. 


II. ABSTRACT OF THE CLINICAL HISTORIES OF THE CASES. 


Frank Nachtigall, Med. No. 1937. Admitted July 5. rgor, with diagnosis of 
typhoid fever. Widal test is positive. Temperature 105° F. A milk 
diet is begun on July 6. The patient is sponged with water of go° F. 
whenever the fever rises above 102° F. Small doses of strychnine are 

139 


140 R. A. Hatcher and Torald Sollmann. 


also prescribed. The urine contains a faint trace of albumin. On July 
12 the diarrhcea has ceased. ‘The temperature varies between 102° 
and 104°. On July 17 the temperature has fallen considerably, and on 
July 25 it has returned to normal. ‘The patient is discharged as well 
on Aug. 9. 

James Boles, Med. No. 2099. Age 29. Admitted Aug. 19, 1901, with 
diagnosis of typhoid fever, the onset of the disease being given as Aug. 
12. The Widal test is positive on Aug. 20. Milk diet was begun on 
Aug. 19, also sponging and strychnine. Tub baths are begun on 
Aug. 20. The temperature to Aug. 25 ranges between 102° and 104° 
F. On Aug. 29 the fever begins to fall, and continues to fall until 
Sept. 3, when it rises again, and reaches 103° F. It remains about 
this until Sept. ro. From this date until the 2oth it varies between 98.5° 
and ror.5°. On Sept. 23 it is practically normal. On Sept. 6 a 

« femoral phlebitis set in which subsided on Sept. 19. On Oct. 4 the 
patient suffered a relapse. He was discharged as well on Oct. 30. 


III. MrtTHops oF ANALYSIS. 


The urines were co/lected from 7 A.M: to 7 A.M. in glass-stoppered 
2-litre bottles, which were brought to the laboratory at 10 A.M. In 
the first case (Nachtigall) the gwantity was measured at the hospital. 
As we found reason to doubt the accuracy of these measurements, 
we directed in the second case (Boles) that the level of the urine be 
marked ona slip of paper gummed to the side of the bottle. The 
urine quantity could in this way be measured by ourselves. The 
Sreesing-points were determined by the Beckmann apparatus, the mean 
of three observations being taken. All the other determinations were 
made in duplicate. The ch/orides were estimated in 10 c.c. of urine 
by Salkowski’s modification of Mohr’s method. The phosphates 
were determined volumetrically with uranium, in 50 c.c. of urine, 
using cochineal as indicator. The total sw/phates were estimated 
gravimetrically, by boiling 50 c.c. of urine for five minutes with 
10 c.c. of HCl, and precipitating with barium chloride. The éota/ 
nitrogen was determined by the Kjeldahl method, using 4 c.c. of 
urine. For the éota/ solids, 10 c.c. of urine were evaporated and dried 
at 110° C. for forty-eight hours. The residue was incinerated for 
ash. The difference was calculated as organic residue. There is 
reason to suppose that there was considerable loss of ammonia during 
the drying and incineration. Indeed, it sometimes happened that the 
nitrogen was greater than the organic matter, and the sodium chloride 


Sodium Chloride and the Constituents of the Urine. 141 


greater than the ash. We therefore place but very little value on the 
data of solids, ash, and organic matter. 

The analyses and calculations were made almost exclusively by 
R. A. Hatcher. 


IV... CALRGULATION OF THE RESULTS. 
The salts were calculated as NaCl, Na,SO, and P,O,. 


Ct =the fofal concentration, was obtained by dividing the depression of the 
freezing point by 1.89. 

Cu =the concentration of urea, was computed by calculating the nitrogen 

' (grams per litre) as urea (multiplying by 2.14) and dividing by the 
molecular weight of urea (60). 

Cel = the concentration of the chlorides, was similarly computed by dividing the 
grams of NaCl per litre by the molecular weight and multiplying with 
(1+a). a, the degree of dissociation, was deduced from the weight of 
the ash. 

The concentration of the sulphates was calculated by an analogous 
formula. 

A X ec. and C X cc, the daily molecules, were obtained by multiplying A or 
Ct by the daily quantity of urine, expressed in c.c. 

8 = the metabolic molecules, (Claude and Balthazard) are computed by multi- 
plying the NaCl per cent by 0.61 and subtracting the product from A. 
The factor represents the depression of freezing-point due to the non- 
chloride molecules, on the assumption that the dissociation of the 
chlorides is that of a 1 per cent NaCl solution. 

A : ; A 
~~,» represents the ratio of total molecules to chloride molecules ; —, that of 
NaCl % ) 
total molecules to ‘ metabolic ” molecules. 


V.. RESUETS. 


We will present our results in condensed form. These may be 
controlled by the table of averages (Table 1), and more exactly by the 
detailed tables, and by the curves from the case of Boles given in the 
appendix. The principle changes in the urine are to be seen in its 
quantity, and in the sodium chloride, total molecules, and ae , or 

a /9 
corresponding factors. The variations are smaller in the case of 
nitrogen, and least with SO, and P,O,. It could also be seen that 


with the nitrogen the daily variations are less than the variations in 


142 R. A. Hatcher and Torald Sollmann. 


percentage, whereas with SO, and P,O, the percentage was the more 
constant. 

As was to be expected, the changes in composition are not 
rigorously uniform under similar conditions. They exist rather as 
tendencies which may be obscured on individual days. Whilst 
they may be recognized by a careful study of the daily figures, they 
are seen more strikingly in averages of typical days, the accidental 
variations being in this manner largely neutralized. 

These averages are presented in the table shown on pages 143-4. 

I. Salt starvation. — (Period of pure milk regime.) —This lessens 
in the first place the quantity of chlorides, and raises the factor 


A se Dad 
Nac] Yo'Y greatly —to 72.3. The quantity of the urine is less than 
during the period of convalescence; A is also somewhat lower: 
The daily molecules are therefore somewhat diminished. The daily 
output of nitrogen is lessened, but its percentage is increased. This 
increased concentration of nitrogen causes the A to be near to nor- 


mal, notwithstanding the low concentration in salts. The factor 


5 is of course abnormally low. The other urinary salts, SO, and 
P,O;, are slightly lessened in concentration, and particularly in daily 
quantity. 

The composition of the urine under the influence of salt withdrawal, 
or retention, appears to be regulated mainly by three factors: — viz. 

1. A tendency to maintain A near 1.0 to 1.5. 

2. A tendency to maintain the normal chloride content of the 
body. 

3. A tendency to excrete a definite amount of nitrogen. 

2. Salt administration. — This causes in general the reverse changes 
to salt withdrawal. The sodium chloride rises greatly, and may reach 


16.9 per litre, or 25.6 gm. per day. ie 

NaCl 
minimum of 1.25. The quantity of urine is increased, and its concen- 
tration rises considerably ; the daily molecules therefore show a great 
increase. The concentration in nitrogen is diminished; but this 
diminution is slight, and in the case of Boles the percentage is even 
increased a trifle. The daily excretion of nitrogen is considerably in- 
creased — the maximum increase being from 10.5 to 15.2 (Nachtigall), 
and from 12.6 to 17.8 (Boles); the average increase is from 10.46 to 


is greatly depressed, to a 


13.67 (N.), and from 12.6 to 14.47 (B.). is of course considerably 


A 
oy 


Sodium Chloride and the Constituents of the Urine. 


TABEE I. 


AVERAGES OF PERIODS. 


The first figures represent the averages, those in parentheses the extremes. 


PERIOD OF CONVALESCENCE 
Bee ee! July 27, 28, and 29. 


(Sorr DIET). 
Boles, Pee 4 to 21. 


Per aay. 


14 


838 (591-1124) 
1280 (900-1980) 


1179.86 (887.3-1596.1) 
1241.3 (876.6-1597.4) 


5.91 (3.65-9.50) 
3.0 (2.47-3.60)! 


2.14 (1.84-2.59)} 


4 


J 


SALT STARVATION. 
Boles, Aug. 22 to 23 


Per day. 


Per litre. 
¥ 
Quantity } a 
K (N. 1.324 (0.976-1.787) 
te. 1.043 (0.711-1.488) 
DGG C. } AE 
A (N. 1.31 (1.11-1.61) 
5 ) B. 1.41 (1.27-1.70) 
$ iN, 44 (3.0-6.1) 
NaCl ) B. 4.76 (3.2-8.3) 

(N. 2.32 (1.41-3 36) 
NagsO, 4p. 2.65 (2.56-2.74)! 
a, (N. 1.75 (1.07-3.13) 

205 ) B. 1.91 (1.75-2.05)! 
N { N. 11.41 (6.87-18.16) 

: ) B. 8.72 (4.69-11.55) 

A (N. 4.03 (2.53-5.96) 
NaCl ) B. 2.43 (1 69-2.89) 

: PERIOD OF 
Nachtigall, July 11 to 19. 
Per litre. 
Quantity at 
rm § N. 1.163 (0.859-1.446) 
) B. 0.994 (0.991-0.997) 

ASC cc 
A {.N. 1.02 (1.00-1.04) 

5 iB: 1.045 (1.03-1.06) 
= {N. 0.5. (0.2-1.0) 
NaCl 1B. 0.77 (0.55-0.9) 

i N. 2.51 (2.10-2.97) 

NaSO4 4p. 2.19 (2.18-2.21) 
Ee. jN. 1.95 (1.20-3.00) 

a5 eB: 1.28 (1.06-1.51) 

1; jN. 12.63 (9.48-16.10) 
yek . B. 12.12 (12.07-12.18) 
A )N. 32.15 (14.24-72.3) 
NaCl 1B. 14.56 (11.01-18.12) 


972 (650-1597) 
802 (680-925) 


0.49 (0.130-0.857) 
0.56 (0.51-0.61) 
W72)(L-21-2.37) 
1.76 (1. 50-2. 02) 
1.82 (1.07-2.65) 
1.06 (0.72-1.40) 
12.13 (8.52-16.92) 
9.67 (8.28-11.15) 


1 13, 14, and 21 only. 


144 R.A. Hatcher and Torald Sollmann. 


TABLE I— (continued). 


PERIOD OF SALT ADMINISTRATION. 
Nachtigall, July 22 to 23. Boles, Aug. 25, 26, 27, and 30. 


Perlitre: Per day. 


1335 (1183-1488) 


Quantity 1154 (960-1425) 


1.955 (1.609-2.301) 
1.368 (1.220-1.498) 


N. 
18}. 
N. 
B. 
N. 


2655.3 (1903.4-3407.3) 
1342.4 (1150.6-1767.6) 


va, lee) 
eo) 


Za 
— 


9 58-25.58) 
3.04-10.26) 
1.80-2.44) 

2.49-5.92)! 
3.03-4.26) 
1.08-2 69) 
12.13-15 22) 
12.63-17.81) 


NaCl 


Ze 


NagSO, 


—_ 


Ze 
DAH YHDNYDOUWD 
"4 


Za 
ied 
We NON WN WAN FE 


— 


onl 
mm 


ca 


5 
l 
S 
l 
j 
l 
) 
l 
) 
l 
j 
l 
j 
} 
) 
l 


1 25, 26, and 30 only. 


increased (toa maximum of 1.94). SO, and P,O, are increased, both 
in concentration, and more especially per day. 

The height of all these changes is reached on the second day, 
the salt being given for one or two days. On the third day there 
is a considerable fall, and the normal is reached on the fifth or 
sixth day. 

The daily quantity of urine pursues quite uniformly a rather 
peculiar course: It is increased on the first day, reaches its maximum 
on the second, falls dc/ow normal on the third, is again high on the 
fourth, and continues high for some time. 

3. Effects of variations in the quantity of urine. — The total con- 
centration, as well as that of the individual constituents, tends to be 
roughly inverse to the quantity of the urine. During the salt starva- 
tion, the main effect falls upon the urea. The daily quantity of the 
constituents tends to vary directly as the diuresis. 

4. The concentration of the individual salts have little relation to 
each other. — SO, follows the urea, but the increase of SO, during 
the salt starvation is not as great as the increase of nitrogen. P,O; 
pursues its own course, being probably largely dependent upon the 


o> 


Sodium Chloride and the Constituents of the Urine. 145 


food, whilst its quantity is not large’enough to be influenced by the A, 
as is that of urea. 

5. The height of the fever appears to have but little influence 
upon the urine. 


VI. DiscussION OF RESULTS. 


Our results show that the disappearance of the chlorides from 
the urine does not lead to any very large changes in the other urinary 
constituents — not sufficient at least to alter the physical properties 
of this fluid. Nor does the administration of sodium chloride produce 
such changes. The bearing which these facts have upon the theory 
of urine secretion becomes evident when they are considered in con- 
nection with other data, as will be done in the next paper. 

We would, however, draw attention to several practical considera- 
tions which are suggested by our results. The first relates to 
diagnosis, the second to therapeutics. 

1. According to the school of v. Koranyi, heart disease can be 
diagnosed from a high a joined with low A x c.c. Our re- 
sults show that the same change results, in at least an equal degree, 
from milk diet. Unless due value is given to this fact, a mistake in 
diagnosis may easily occur. 

2. The addition of sodium chloride to a milk diet results in a 
greater excretion of urine and of metabolites. The latter are at the 
same time diluted. These effects seem to us desirable, especially in 
fevers. The retention of chlorides shows further that the body 
requires a certain amount of sodium chloride, which is not supplied 
by the insufficient quantity of this substance contained in the milk. 
For these reasons the addition of sodium chloride to a milk diet 
appears commendable. The salt may be added directly to the milk 
without inconvenience. It should be distributed so that about fifteen 
grams are taken per day. 


146 hk. A. Hatcher and Torald Sollmann. 


VII. . AP 
TABLE 
COMPOSITION OF 


I. Nachtigall. Milk diet from July 6th to 26th. 


Grams pet litre. 
pate, | Qn | 26 aa 
~ | AEG) wach | a5S0,!| P.O, |) oN) 7 2a eee eee 
July 11 650 0.2 2 305 165 | 16.10 | 22.42 3.42 | 19.00 
oe les 1035 0.2 2.65 | 1.46 | 12.67 14.75 4.21 10.54 
eile 887 0.3 Zoe 1.40 | 1141 16.51 4.34 Lay, 
Seals 1597 0.5 2.07 1:20 | 9:48 15.85 3.80 12.10 
‘ Al6 828 03 2.10 3.00 10.29 13.10 5.85 7.25 
cee ald’ 769 0.8 ZOD) 2.77 11.93 17.10 6.00 11.10 
— als 1153 0.7 2.97 2.30 15.54. 16.35 6.15 10.20 
The patient received soft diet on the 19th, but returned to milk diet on the 20th. 
July 19 857 1.0 2.94 1.85 13.65 16.05 6.27 | 9.78 | 
Three slices of bread and one egg on 2st. 
Seal 1479 2.8 2.24 2.15 10.57 15.90 7.15 8.75 | 
120 gms. of NaCl were given during the two days, beginning with 7 A. M. of the 21st | 
and ending with 7 A.M. of the 23d. | 
July 22 1183 8.1 2.97 | 3.60 | 12.91 28.45 14.07 | 14.38 
fo BS 1488 16.9 eed 2.04 8.01 27138 20.81 | 1652 
Sodium chloride stopped. | 
Sime 2 532 146 | 248 2.76 10.81 31.80 19:35) |) ues | 
« 95 | 1035 | 73 | 249 | 345 | 1323 | 2601 | 138 | 1468 
Soft diet begun on 26th. | 
“ 2 | 1183 | 44 | 388 | 5.10 | 1781 | 2846 | 1292 | 15.54 
27 | soi | 30 | 336 | 313 | 1816 | 2660 | 1040 | 16.20 
28 798 3.1 1.93 1.27 10.29 13.18 | 650 | 668 
B29 1124 56 | 259 1.55 10.32 17.75 9.19 | 8.56 
Aug 9 | 72 | 61 | 141 | 107 | 687 | 1483 | 953 | 50 


1 Not known, but high. 


Sodium Chloride and the Constituents of the Urine. 


PENDIX. 


il 


THE URINES. 


147 


II. Boles. Milk diet from August 19th till September 4th. 
Grams per litre. 
mae | SB a 
(c-c-) | Nacl. |Na,SO,.| PO; | N. | Total | ash, | Organic 
a solids. | | residue. 
Aug. 22 680 0.90 2.206 1.06 | 12.18 16.89 1.32 15.57 
or. 23 925 0.55 2.182 1.51 | 12.07 15.50 1% | 1366 
Cua Sodium chloride, 30 gms. in two days (24th and 25th). 
Os 1080 0.95 2.460 1.37 12.53 14.02 efhe, 12.23 
2 1425 370 4.736 1.14 12.50 13.62 3.26 9 36 
Sodium chloride discontinued. | 
me 26 960 4.70 2.582 1.13 13.16 Weil 4.18 13253 
eee TL 1050 2.90 72 | 1:37 LOSS 3.98 1535 
hilo $35 3.00 | 14.66 
Sodium Ehiawde! 45 gms. on oth, 
eee | 740 5.50 2.938 2:39 13.72 18.48 6.65 11.83 
Sodium chloride discontinued. 
oS atv 1180 | 8.70 2.468 1.40 | 11.37 | 19.68 9.64 10.04 
ceil 825 | 4.40 2.378 1.85 | 11.86 | 18 63 | 6.13 | 12.50 
Sept. 1 1410 | 1.90 8.75 | 
2, 1720 | 0.92 8.29 
3 1350. ||. 095 7.28 | 
Soft diet begun on 4th. | 
«4 | 1160 | 4.00 8.91 | 
“5 | 1980 | 480 4.69 
«7G 1140 | 3.20 
13 L350) |) 4210 8.98 
ales 1145 | AS 8.00 
| 10.59 


148 R. A. Hatcher and Torald Sollmann. 
TABLE 
DAILY 
I. Nachtigall.! Milk diet from July 6th to 26th 
Grams per day. 
Date. Q ae ve 
NaCl. Na.SO, P,O;. INC 
July 11 650 0.130 eA: 107 10.46 
pall 1035 0.206 1.90 1.51 13.11 
so 887 | 0.266 1.64 124, 9.02 
JS 1597 0.798 2.29 1.91 15.14 
Samal 828 0.248 AL 2.48 8.52 
Sie ly 769 0.615 1.36 213 se} 
els 1153 0.807 NEY | 2.65 16.92 
The patient received soft diet on the 19th, but returned to milk diet on the 20th. 
July 19 857 0.857 e714: 1.58 14.70 
Took three slices of toast and an egg. 
tO IA 1479 “ell! 2.30 3.18 15.63 
From 7 A.M. of the 21st to 7 A. M. of 22d, took 120 gms. NaCl. 
eae 1183 9.58 2.44 4.26 | 15.22 
a3} 1488 | 25a15 1.80 3.03 12.13 
Sodium chloride discontinued. 
hot 532 7.91 0.93 1.50 5.86 
ser es: 1035 (fests) 1.79 Si 13.69 
Soft diet begun. 
“ 26 1183 5.20 3.17 6.03 | 21.97 
a 21 591 iid lsy/ 1.85 | 10.74 
25 798 2.47 1.06 | 1.01 8.21 
29 1124 6.30 2.01 1.74 | 11.60 


! The daily quantities in the case of Nachtigall must be considered unreliable. 


Sodium Chloride and the Constituents of the Urine. 149 


ee 
URINE. 
II. Boles, Milk diet from August 19th to September 4th. 
Grams per day. 
| Date. mies 
NaCl. NaySO,. P.O;. N. 
: a a 
| Aug. 22 680 0.61 HSON eh Oe | 8.28 
| 93 925 0.51 2.02 | 140 | 11.16 
| Sodium chloride, 30 gms. | | 
“r 94 1080 1.03 | BG. | 1.48 13.53 
“95 1425 5.28 592 | © 1.62 17.81 
Sodium chloride stopped. 
a 960 4.51 | “<249 1.08 12.63 
ies 27 4 6| | 1080 3.04 1.81 14.04 
| wes | 835 | S250 | 12.24 
| Sodium chloride, 45 gms. 
ae. | 740 4.07 BAT 1.74 10.15 
| Sodium chloride discontinued. | 
«30 | uso | 102 #«+'| 291 2.65 13.42 
«31 | a5 | 3s ate 1.53 979 
| 
Sep 1. | 1410 2.68 12.04 
si Re ep) 1.59 14.27 
72. 3 1350 1.28 9.83 
Soft diet. 
| 4 1160 | 4.64 10.34 
ares | i980, | 9:50 | 9.29 
ar 1140 3.65 | 7.42 
e313 1350 5.53 3.60 2.59 15.17 
4 1450 | 4.65 Bos nt 2.00 10.78 
pee 900 TAT 2.47 | 1.84 10.40 


R.A. Hatcher and Torald Sollmann. 


150 
TABLE 
CONCENTRATION 
I. Nachtigall. Milk diet from July 6th to 26th. 
Date, | anti:|) ta acu eae 
July 11 650 0.7651 0.0063 0.0405 0.5750 
coals 1035 | ~~ 0.5677 0.0064 | 0.0484 0.4525 
ogy 887 0.5339 0.0096 0.0457 0.4075 
walls 0.4545 0.0160 0.0372 0.3387 
UG 828 0.5143 0.0095 0.0353 0.3675 
sep di7 769 | 0.6794 0.0250 0.0427 0.4262 
1S 1153 0.6556 | 0.0224 0.0543 0.5500 
The patient received soft diet on the 19th, but returned to milk diet on the 20th. 
July 19 857 0.7534 0.0311 0.0489 0.4875 
| Three slices of toast and one egg. 
Zi 1479 0.5910 0.0881 0.0401 0.3775 
120 gms. of sodium chloride from 7 A.M. of the 21st to 7 A, M. of the 23d. 
22, 1183 0.8513 0.2472 | 0.0476 | 0.4612 
“93 1488 1.2174 0.5026 | 0.0387 0.2862 
| Sodium chloride discontinued. | 
she 24 0.9624 0.4441 0.0370 | 0.3862 
Se, 25 0.8794 0.2258 0.0389 | 0.4725 
Soft diet begun. | | | 
26 0.9688 0.1369 | 0.0627 | 0.6362 
ee 0.9455 0.0933 0.0554 0.6487 
28 0.5883 0.0973 0.0328 0.3675 
oN aie) 0.7513 0.1733 0.0407 0.3687 
Aug. 9 0.5164 0.1898 0.0233 0.2437 


Sodium Chloride and the Constituents of the Urine. 151 


EV; 
IN MOLECULES. 


II. Boles. Milk diet from August 19th to September 4th. 


| 
See) ee teen, (> Nese) ee 
| Aug. 22 680 | 0.5243 | 0.0292 | 0.0305 0.4350 
P= 23 | 995 | osazs 0.0178 | 0.0301 0.4312 
Sodium chloride, 30 gms. | 
ee 1080 0.5354 0.0310 | oo4l 0.4475 
< 95 1425 0.6455 | 0.1164 0.0799 0.4462 
Sodium chloride stopped. 
26 960 | 0.7170 0.1513 0.0435 0.4587 
Ce 1050 0.7392 0.0892 tae 0.4775 
28 835 | 0.7290 0.0944 hae 0.5237 
Sodium chloride, 45 gms. 
} © 99 740 0.8661 | 0.1692 0.0471 0.4900 
Sodium chloride tonact 
80 1180 0.7926 0.2697 0.0396 0.4062 
“31 825 0.7397 0.1380 0.0388 0.4237 
| Sept. 1 1410 0.5058 | 0.0593 sat 0.3050 
| Callas’ 1720 0.3815 0.0300 ie 0 2962 
3 1350 0.3418 0.0308 Ase 0.2600 
Soft Bice. 
rae 1160 | 0.5508 0.1244 tebe 0.3175 
eet. 25 1980 0.3762 0.1503 aed 0.1675 
6 1140 | 0.4069 0.1012 caee 0.2337 
a 1S 1350 | 0.6265 0.1276 0.0450 0.4012 
4 Vgsue9)  Dseds 0.1305 0.0432 0.3437 


« 9] 900 | 0.7873 0.2554 | 00440 0.4120 


ese R. A. Hatcher and Torald Sollmann. 


TABLE 
VARIOUS 
I. Nachtigall. Milk diet from July 6th to 26th. 
Rt, Daily Daily | Metabolic | | 
Date. any A. molecules. | molecules.| molecules. | te Sane A 
(Sc) AXce.| CX ce (8) | % NaCl | 8 
July 11 650 1 446 S40!0) |) 497-31 1.4338 72.30 1.01 


weeds L035 LOTS 1110.0 587.59 1.0608 | 53/65/02 le 
oder 887 1.009 895.0 | 473.56 WOOO 79) 35:63 1.00 


als: 1597 1) 01859 1370.8 698 89 0.8285 | 17.18 1 OF 


S2a6 828 0.972 804.5 425.84 0.9537 | 32.46 1 02 

LS lly 769 1.284 OST || o22-45 225250 eGo 1.04 
| 

KS! WSS 12239 1428'5) "| ¥d9-99. |) 1.1963 17.70 1.03 


| | 
The patient received soft diet on the 19th, but returned to milk diet on the 20th. 


July 19 857 a 1220.3 645.86 1.3630 Woress | dO: 


et All 1479 1.117 1652.0 | 874.09 | 0.9462 a99 


Three slices of toast and one egg. | | 

| 1.18 

120 gms. of sodium chloride were given during the two days, beginning with 7 A. M. 
of the 21st and ending with 7 A. M. of the 23d. 

July 22 | 1183 1.609 1903.4 | 1006.08 1.1049 1.97 | 1.45 

| 


eg) 1488 2.301 3407.3 1800.49 | 1.2701 1.40 1.81 


Sodium chloride stopped. 


/ 
Pees |) Sa2 We Aseh? 967.71 511.00 0.9390 1225 1.94 | 


ys | 1035 1.662 1720.17 910.18 | 1.2176 2.28 «|. 36 


Soft diet begun. 


26 1183 1.831 2168.07 , 1146.08 1.5526 4.16 1.11 
27 581 1.787 1056.12 | 558.80 1.6040 5.96 Leu} 
LOLS 798 LUIZ 887.37 | 469.46 0.9229 3.59 | 1.20 
29 1124 1.420 1596.08 844.46 1.0784 2.53, | “Te 


Aug. 9 eat 0.976 eee | sretere 0.6039 1.60 1.61 


Sodium Chloride and the Constituents of the Urine. 153 


Vis 


FACTORS. 


II. Boles. Milk diet from August 19th to September 4th. 


a Daily | -Daily Metabolic 
Date. heared A. molecules.) molecules.) molecules. ue A: 
| oS Oe Gal) Ge Seecic: (8) fo NaCl 8 
Aug. 22 680) |) .0:9911 673.8 356 52 0.9301 11.01 1.06 
mee. || 925° || 0.997 922.2 || 487-94 0.9635 18.12 1.03 


Sodium chloride, 30 gms. | 
met |) | LOSO: |. L012 1093.0 | 578.23 0.9538 10.60 1.06 
ee yA) 1425 1 220 1738.5 919.84 0.9933 3.30 1.23 


Sodium chloride stopped. 


“ 96 960 | 1.355 | 13008 | 687.32 1.0683 2.88 1.27 

« 97 | 1050 1.397 14768 | 77616 1.2201 4.82 1.14 
| 

« 98 | 9835 1.378 1150.6 608.80 1.1950 4.59 16 


| 
Sodium chloride, 45 gms. | 
j | 


ape | 740 1.637 | 1214 640 61 1.3015 2.98 1.28 


| 


| 
Sodium chloride discontinued. 


« 30 | uso | 1498 |. 17676 | 935.29 | 0.9673 yeh)! <a055 
« 31 | 925 | 1308 | 11533 | o101s | 112% | 318 | 1.24 
Sept. 1 | 1410 | 0.956 | 13480 | 713.18 | 0.8401 5.03 | 1.14 
«2 | 1720 | ova | 12401 | 65618 | 0.6646 795 | 1.09 
feos | isso | 06. | S721 | 4etaz | ose7s |~ 680 | 1.10 
Soft diet. | | 

Reel sia |e. | its | esses | Orem | neo. | aa 
« 5 | 198 | o711 | 1407.7 | 74488 | 0.4182 169 | 1.70 
“6 | 40 | 0769 8766 | 46387 | 05737 240 | 1.34 
« 33 | 1350 | 1.184 | 1507.4 | 845.77 | 0.9340 289 | 1.27 
| “. 14 | 145 | 1065 | 12194 | 64521 | osu9 | 257 131 
« 2] o00 | 1488 | 13302 | 70857 | o9si7 | 179 sl 


154 Lk. A. Hatcher and Torald Sollmann. 


CURVES. 


The curves are plotted from Boles’ urine. 


explanation. 


AuGUST SEPTEMBER 


2000 ce 


etal 


F 
Hl 
i 
r 
a 


NaCl 4 
Q 


Y 
\ 


Na, SO, 25 


oR 
3 
) ee ase eee 


SALT SALT SOFT DIET 
PERIOD PERIOD BEGINS 


PERCENTAGE OF THE CONSTITUENTS. 


as 


FHLIT HId STINITIOW-WVYD 
» 
u 


>. 


AuGcUST 


Pail 
ofr ATT 
TIM Wea 

PEC 


SALT 
PERIOD 


PER DAY e2 24 26 2 30 


SEPTEMBEP 


CONCENTRATIONS. 


I 
A BACivG 


ieeesec=ess 


PERIOD 


faa 


They require no further 


AvCUST Tentewaes 
ie EY’? 


aS 
He 
a 
israel ane 


rae a 
P| 


—-— es BEGINS 


SALT SALT 
PERIOD 


GRAMS PER Day. 


THE MECHANISM OF THE RETENTION OF CHLO- 
RIDES: A. CONTRIBUTION TOG -THE THEORY OF 
URINE SECRETION. 


By TORALD SOLLMANN, M.D. 


[From the Pharmacological Laboratory of the Medical Department of Western Reserve 
University, Cleveland, Ohio.| 


CONTENTS. 
Page 
meine retention OMeniondesimieven fo ' i) 5) (2 fe) eae lw ee a e156 
II. The retention of chlorides when salts are deficient inthe food . . . . . 159 
III. The retention of chlorides after sulphate injections . ...... Aen lesi| 
IV. Do other salts suppress the chlorides in the same manner as do aie 
sulphates? . . Ae ee a OM sae uy Ko A LO 
V. Forster’s theory of shel retention nat phiurides SW ES a) OE eee | Gee OS 
VI. Cushny’s theory of the reabsorption of chlorides . . . : 167 
VII. Is the reabsorption a sufficient explanation for the aieciice af chiorides'3 in 
thepurime'* 2) ©. Sel ce a aces my HLL CG)2) 
VIII. Is the glomerular fluid Esened = flération: or be SEGrehionts sas pee ee NTL 
TEPCENICIUISIONS el <5 ofr andl Sos as dad Jah Bo isd wie ap ton Sete d oti Sa. US 
PESTO ITOD TAP de, Berea aie el eer So Recess ls ee, peel sh Mad ip, ah AS 


HE urine becomes almost or quite chloride free in infectious 

diseases, in salt starvation, and on intravenous injection of 
Na,SO, solution. The research recorded in the preceding paper led 
me to inquire into the mechanism of this chloride retention. 

Three very different theories have been advanced to account for 
the phenomena: one theory assumes that nochloride is filtered. This 
was first definitely formulated by Forster. Another theory assumes 
that the chlorides are filtered in the glomeruli, very much as when 
retention does not occur; but that the chloride is reabsorbed. This 
has been supported by Cushny. The third, which grants to the glom- 
erular cells the property of vital secretion, assumes that these cells 
secrete no salts when the proportion of salts in the blood falls below 
a certain amount. 

These explanations involve respectively the theories that the urin- 
ary salts are filtered, reabsorbed, and secreted. Each of these theories 
has been called into doubt, so that the study of the question of chloride 


155 


156 Torald Sollmann. 


retention offers a peculiar interest, and may be expected to throw 
light on the whole subject of urine secretion. 

In this study I have availed myself very freely of the published 
records of other observers, but have also introduced my own experi- 
ments when these seemed more suitable. In presenting the subject 
I will first analyze the conditions under which the chloride-reten- 
tion is seen, and then the theories by which the retention has been 
explained. 


I. THE RETENTION OF CHLORIDES IN FEVER. 


The earlier literature relating to the chlorides of the urine in fever 
is given by Rohmann (1).! It appears that Redtenbacher (2) in 
1850 was the first to discover that the chlorides disappear com- 
pletely, or almost completely, from the urine in genuine pneumonia. 
Unruh (3) extended this observation to typhoid and other acute 
fevers, as also to abscess formation, trichinosis, aortic insufficiency, 
Ete: 

Such a striking phenomenon could not fail to elicit inquiry and 
bring forth attempts at an explanation. Deficiency of the chlorides 
in the food was perhaps the most apparent cause, and this was one 
of the earliest explanations advanced. Other suggested explanations 
are, that the chlorine is retained in the exudates, or excreted by other 
channels, — especially by the stools, if there existed diarrhoea, — by the 
sputum, sweat, etc., or that a storage of the chlorine occurs. 

1. Is there a storage of chlorides in the body ? — KGhmann (1) seems 
to have been the first to compare carefully the income of chlorine, 
and its elimination by the urine and faeces. He found that there is 
an actual retention of chlorine during the febrile period, often per- 
sisting for several days after the crisis. The difference between the 
income and output of chlorine is, however, quite small, usually less 
than two grams per day. He proves that this retention cannot be 
accounted for by faulty absorption of the salt from the alimentary 
canal. 

Rohmann explains the cause of the storage of chlorine by the 
febrile changes of metabolism, on an adaptation of Forster's theory. 
Forster (5) believes that the greater part of the salts of the body is 
united with proteids, forming proteid-salt combinations which cannot 
be excreted by the urine. Rohmann considers that the altered met- 


' The numbers refer to the Bibliography at the end of this paper. 


Lhe Mechanism of the Retention of Chlorides. 157 


abolism of fever causes tissue proteids to be converted into “ circulat- 
ing” proteid, and that the latter requires more chlorine. Critics of 
this theory have not been wanting. It has been pointed out that it 
would first need to be demonstrated that circulating proteid contains 
more chlorine; secondly, that the circulating proteid is increased. 
As to the former, Langlois and Richet (4) have recently shown that 
the blood contains two or three times more chlorine than any of the 
organs, so that the circulating proteid-containing liquids may be con- 
sidered to be the richer in chlorine. But the theory that the circu- 
lating proteid is increased in fever, is not supported by facts. Nor 
is it easy to see why such an increase of circulating proteid should 
occur. To the contrary, fever causes an increased destruction of 
proteids; and this, by Forster’s theory, should lead to a liberation 
and excretion of salts, toa loss of salts, instead of a salt retention. 
That an increased destruction of proteids does lead to a loss of salts 
seems amply demonstrated. Kast (9) administered poisons which 
caused a destruction of blood-corpuscles — such as CO, pyrogallol, 
and toluylendiamin — or such as increase the destruction of tissue, 
as phosphorus. These caused an increase of chlorine excretion, 
generally parallel to the N, and fairly independent of the quantity 
of urine. A similar increase of chlorine excretion accompanies the 
pre-mortal rise of N in starving animals. This was observed by 
Forster (5) and also by Kast (9g). Our figures for Nachtigall, com- 
paring the beginning and end of starvation period (July 11 and 17), 
also show an increase of chlorine, but since this is not proportional 
to the nitrogen, its cause is probably extraneous. 

These facts are certainly adverse to Réhmann’s theory. For the 
only effect of fever on proteids which we know with certainty is, that 
there is an increased destruction, which should lead to a loss of salts. 
Indeed, the results of Gramatchikov (6) disagree entirely with those 
of Rohmann. Gramatchikov found that the body actually lost a 
small amount of chlorine in fever. The difference is perhaps to be 
explained by the original content of the patients in chlorine, or by 
other factors, such as the diet and its content in chlorine, or the 
amount of tissue destruction, etc. 

Rohmann’s results show that there may undoubtedly be a retention, 
even with as small a chlorine income as exists in fever. That a tem- 
porary storage may occur when larger quantities of salt are admin- 
istered, is well known. We wished to confirm this fact on our cases 
by observing the total income and elimination of chlorides, but had 


158 Torald Sollmann. 


to abandon the attempt, as it was found impossible to control the 
food and collect the excreta of the patients with sufficient precision. 
However, the fact that the excretion of chlorine does not reach its 
maximum until the second day after the administration of sodium 
chloride, and does not return to normal until about the fifth day, 
argues in our results also for a chlorine retention under these con- 
ditions. It is supported by all other experiments on sodium chloride 
administration. 


Rohmann found that the maximum excretion fell usually on the second day, 
and remained high for three days, when the entire administered quantity 
was excreted. Lauder-Brunton (7) quotes from Ludwig’s Lectures 
(1869-1870) that, if a definite quantity of sodium chloride be taken daily 
for some time, the quantity excreted by the urine becomes practically 
equal to that ingested. If the consumption is now increased, the excre- 
tion does not increase (correspondingly ?) for about three days, so that a 
storage must occur. After further three days, the quantity excreted will 
again equal that ingested. Diminution of the sodium chloride of the 
food has a corresponding result, requiring the same time to produce 
equilibrium. Krummacher (8) finds that when sodium chloride is in- 
jected hypodermically, two days are required for practically complete 
excretion. 


Whilst the administration of sodium chloride certainly leads to a 
storage of sodium chloride, this is purely temporary if the body pos- 
sesses a normal amount of sodium chloride to begin with. Langlois 
and Richet (4) did not find the chlorine content of the tissues mod- 
ified by adding sodium chloride to the food. 

2. Retention of chlorides in effusions. — Giving all possible weight to 
Kohmann’s data and experiments, it must be said that the binding of 
salts by the “circulating” proteids can play but a very subordinate 
rdle in the retention of the chlorides. A much more important factor 
is the formation of effusions, when such occur. 


This explanation was already advocated by Redtenbacher (2). M. Schu- 
bert (10) has studied this phenomenon in four cases of ascites from 
hepatic cirrhosis. Whenever puncture was performed and the liquid 
drawn off, the sodium chloride of the urine sank greatly and recovered 
only slowly. ‘The decrease existed only in the daily excretion, the per- 
centage remaining normal. ‘There was also a similar decrease in the 
daily excretion of nitrogen, and in the quantity of urine. This reten- 
tion therefore differs radically from that encountered in fever and hunger, 


The Mechanism of the Retention of Chlorides. 159 


as may be seen, for instance, from our results on the typhoid urines: 
In the latter the percentage of the sodium chloride in the urine is re- 
duced, as well as the daily quantity, whilst the nitrogen and the quantity 
of the urine are but little altered. The two conditions can therefore 
be readily distinguished. 

I have only encountered a single statement concerning the physico-chemical 
constitution of the urine in ascites. Waldvogel (11) found the factor 


NaCl raised. ‘This indicates that the chlorine retention is greater than 
Na 


the retention of other solids ; it might, however, be explainable by the 
interference of the ascites with the renal circulation. 

As a converse, it has been found (12) that a very large quantity of chlorine is 
excreted by the urine during the absorption of effusions. The chlorine 
excretion is said to be much larger than what would correspond to the 
nitrogen excretion. 


3. Retention of chlorides through the formation of other secretions, of 
blood, and of solid tissues. — That the loss of chlorides through the 
sputum, through diarrheic stools, etc., will lead to diminished excre- 
tion of chlorides by the urine, may be granted a priorz. 

A retention of chlorides also occurs when chlorine is lost through 
hemorrhage. This has been shown on animals by Kast (Q), and has 
also been demonstrated on man (14). The effect is precisely the 
same as described in the formation of effusions. 

An increased growth of tissue, joined with exudations, also dimin- 
ishes the chloride excretion, as may be seen in many cases of 
carcinoma (15). 


II. THE RETENTION OF CHLORIDES WHEN SALTS ARE DEFICIENT IN 
THE Foop. 


After a liberal value is assigned to Réhmann’s theory, and to the 
retention of chlorides by effusions, etc., the greater part of the 
phenomenon of chloride-retention in fever is still unaccounted for. 
The real explanation is to be found in a deficient salt income. The 
diet of fever patients is always scanty, but is particularly poor in 
chlorides. The composition of the urine corresponds almost precisely 
to what is seen after a corresponding diminution of chloride-income 
in health. Iam therefore inclined to regard the chloride-retention of 
fever as depending practically purely on the diet, and consider it no 
more a reliable index to the severity of the disease, than is the 
appetite of the patient, by which it is controlled. 


160 Torvald Sollmann. 


The great influence of the diet on the salt excretion justifies a close 
study of its effects. I have already presented these in a contem- 
porary paper in “American Medicine” (Oct. 25, 1902). 


Falk (1848) is said (16) to have been the first to prove that the sodium 
chloride of the urine varies with that of the food. Bidder and Schmidt 
in 1852 (17) observed in hunger the practical disappearance of chlorine 
from the urine at a time when there was still a large quantity of chlorine 
in the body. Forster (5) in 1873 shows that the chlorine of the urine 
is reduced to mere traces when food nearly free from salt is given; but 
that some trace is excreted until the death of the animal. The great 
diminution of chlorine in fasting can also be seen in an interesting man- 
ner in the observations of Hoover and Solimann (18) on metabolism 
during fasting in hypnotic sleep. In this case no solids were taken for 
a week, but water was allowed ad Zébztum. ‘The percentage of chlorine 
sank immediately and rapidly and after a few days remained at a fairly 
constant, low level. Lindemann (13. p. 15) finds the percentage of sodium 
chloride in fasting (five cases) to fall to 0.17 to o.20 per cent, mean o.18. 
This is not quite as low as the figures obtained by Hatcher and myself. 
However, there is considerable reason to believe that the other salts of 
the milk save sodium chloride from excretion; and even the proteids 
of the milk may have a saving influence on the chlorine, by saving the 
proteids of the body from destruction. In support of this, Forster found 
(5, p- 354) that the less the quantity of salt-free food, the greater was the 
loss of P,O; in the urine; and again (p. 358) that more salts are excreted 
in hunger than when salt-free food is given. Langlois and Richet (4) 
found that fasting, with or without administration of sodium chloride, 
does not lower the content of the tissues or blood in chlorides (zZ. é., the 
excretion of chloride corresponds to the destruction of tissue) ; but feed- 
ing with chloride-poor food causes a diminution of to per cent in the 
sodium chloride of the tissues (no diminution in the blood). This is 
reconcilable with Forster’s results, on the theory that the body weight 
did not diminish as greatly during the experiment as in plain fasting. 
The addition of other salts to chloride-poor diet did not alter the loss 
in either direction. 

The effects of a chloride-poor diet on the other salts of the urine have been 
given in the preceding paper, so that it will not be necessary to repeat 
them in this place. 


The effects of a deficient chloride income on the tissues and fluids 
other than the urine, have also been investigated. Langlois and 
Richet (4) showed that chloride-poor food lowered the chlorine of 


The Mechanism of the Retention of Chlorides. 161 


the tissues, but not that of the blood; the blood indeed retains its 
chlorine with great tenacity, and is the last component of the body 
to participate demonstrably in a chlorine loss. 

The effect of salt starvation on the chlorine of the gastric juice has 
been investigated by Forster and by Cahn. Forster (5) found that 
the gastric juice still contained 0.07 per cent of chlorine when the 
urine had become free from chlorine. Cahn (10, p. 532) confirms the 
persistence of sodium chloride in the gastric juice in advanced salt 
starvation, but observes that the secretion of free hydrochloric acid 
ceases when the chlorine content of the body falls below a certain 
amount, and recommences as soon as Chlorine is administered. 

Salt starvation, therefore, diminishes the chlorine of the gastric 
juice, but not as readily nor as profoundly as it reduces the chlorine 
of the urine. 


II]. THe RETENTION OF CHLORIDES AFTER SULPHATE INJECTION. 


Magnus in 1900 (21, II, p. 417) discovered that the chlorides 
disappear almost completely from the urine when a solution of sodium 
sulphate is injected intravenously. The fact was also discovered 
independently by me in 1899 (22, p. 20). In the three experiments 
which I made, the disappearance occurred only in the two (Exps. VI 
and VII) in which a considerable diuresis was produced. In the third 
(Exp. IX) the excretion of the urine was only moderately increased, 
and in this the percentage of sodium chloride in the urine only fell to 
Gay per cent. 

Further experiments in this direction have recently been published 
by Cushny (23). His results are very contradictory. He states 
(p. 436) that injection of sodium sulphate alone usually zzcreased 
the chlorine of the urine. On the other hand, when mixed solu- 
tions of sodium chloride and sodium sulphate were injected, the 
chlorine adésappeared entirely from the urine in one and one-half hours 
(Exp. I, p. 432). Pototzky (41) also finds that sulphate injection 
increases the chlorine excretion in salt-starved rabbits. 


Cushny’s and Pototzky’s experiments were made on rabbits, whilst Magnus 
and I used dogs. ‘This difference in the animals may account for the 
different results ; for Magnus has shown that these animals do not possess 
the same retaining power for different salts. He states (21, II, p. 405) 
that when sodium chloride was injected, both rabbits and dogs retained at 
the end of the experiment two-thirds of the injected salts; but when 
sodium sulphate was injected, the rabbits retained two-thirds, and the 


162 Torald Sollmann. 


dogs only one-tenth. A difference exists also in the excretion of phos- 
phates; Bergmann (24) claims that carnivorous animals excrete the major 
part of the phosphates by the urine, whereas herbivorous animals excrete 
injected phosphates almost entirely by the feeces.' 

In all the sodium sulphate injections made by Magnus and by me, it can further 
be seen that the sodium chloride does not lJeave the blood by any other 
channel. On the contrary, the total amount of chlorine in the blood is 
increased after the injection of the sulphates. The increase is probably 
to be attributed to the diminished partial osmotic pressure of the chlorine 
ions in the serum. 


IV. Do OTHER SALTS SUPPRESS THE CHLORIDES IN THE SAME 
MANNER AS DO THE SULPHATES? 


I. Sodium nitrate.— I had planned to extend my former experi- 
ments on salt injections to other salts; but so far I have only found 
time for a single experiment with sodium nitrate. A dog was used. 
The method was the same as in my former experiments (22), employ- 
ing 75 c.c. per kilo of a solution of sodium nitrate, A 0.481. The 
urine contained at first 4 per cent of sodium chloride; in one and 
three-quarter hours it contained 2.8; in six hours the sodium chloride 
was still 2.4 per cent. The diuresis was quite profuse. The result 
differs so greatly from that seen with sodium sulphate that there 
need be little hesitation in affirming that sodium nitrate does not 
suppress the chlorine of the urine. This conclusion is confirmed by 
the effect which the oral administration of potassium nitrate produces 
in animals whose urine has been rendered chlorine free by salt 
hunger. 

A. Cahn (9) mentions that potassium nitrate causes chlorine to 
appear in this condition. He does not give the source of this state- 
ment, but confirms it by some experiments of his own. He found that 
in the dog the administration of potassium nitrate raised the chlorine 
from a trace to 0.119 gm. per day (p. 528). The urine happened to 
be reduced from 1500 to 100 c.c. He states (p. 523) that sodium 
nitrate was ineffectual, but this conclusion may be challenged, since 
only a single experiment was made; as the salt was given by the 
mouth, there could have been defective absorption or some other 
interfering factor. 


' In this he differs, however, from Cushny (23, p. 442 and 443) who recovered 
77 per cent of the injected P,O, from the urine in three hours. Cushny worked 
with rabbits, Bergmann with sheep. 


The Mechanism of the Retention of Chlorides. 163 


Langlois and Richet (4) also found that the chlorine of the tissues 
and blood was greatly reduced on the injection of large doses of 
nitrates into starving animals. 

2. Sugar.— Cahn, as also Langlois and Richet, found that this 
behaved very similar to nitrates. 

3. Urea.— That urea does not cause a retention of chlorine may 
be argued from the normal occurrence in urine of large amounts of 
both substances. Katsuyama (29, p. 237), giving urea by the mouth 
to rabbits whose urine had been rendered chlorine free by fasting, 
found that this caused a very considerable excretion of chlorine, some- 
times quite independent of diuresis (Exp. 3). 

In this respect urea bears a close resemblance to the nitrate. 
Caffein and theobromin also produces a similar effect. 

4. No data regarding the chlorine excretion in phosphate injection 
could be found in the literature. 


V. ForSTER’S THEORY OF THE RETENTION OF CHLORIDES. 


1. The serious difficulty offered to the filtration theory of urine 
secretion by the retention of chloride in salt starvation, when the 
composition of the blood is but slightly altered, was recognized by 
Forster (1873). He proposed the following explanation (5, p. 318) : 
The larger part of the soluble salts of the organism exists in firm 
combination with the organic substances, particularly with the 
proteids of the tissues, juices, and blood. Another, much smaller 
quantity, is present in simple solution, in a free state. The former — 
the salt-proteid compounds — are incapable of filtration in the kidneys. 
Whenever the proteids are saturated with salt, any additional quantity 
of salt is filtered off rapidly into the urine. The excretion of salts is 
therefore practically equal to the salt income. When salts are with- 
held entirely, their excretion should also cease completely. That a 
small quantity of salt is still excreted in salt hunger, he explains by 
the combustion of proteids, during which the combined salts are 
liberated. When the proteid combination is under-saturated, as 
must occur when salt-free proteids are fed, this liberated salt should 
combine with the unsaturated proteid, and should thus be utilized 
over and over again, without being excreted. He assumes that this 
utilization occurs, but that the combination requires time; that 
meanwhile the salt circulates free, and thus a fraction passes through 
the urinary filter. 


164 Torald Sollmann. 


2. This theory of Forster’s, that the urine is a filtrate containing 
not the entire ash, but only the free salts, seems to have fallen con- 
pletely into oblivion; yet it is a very plausible hypothesis, which 
agrees excellently with the experiments made by Forster. It would, 
moreover, explain a large number of other obscure physiological 
phenomena, such as the remarkable tendency of the blood to pre- 
serve a constant composition; the fact that sugar is not excreted 
until its proportion rises above 0.2 per cent, but that all above this 
is eliminated; the fact that no urine is excreted unless a “ harnfahige 
Substanz,”’ —7z. ¢., a substance capable of being secreted by the urine 
— is added in the blood, as seen from the experiments of Munk, of 
Nussbaum, Beddard and Halsey, of Magnus, and of Spiro. 


I. Munk (25) showed that if the blood of a fasting dog is circulated through 
an excised dog’s kidney, no secretion occurs, but that secretion is started 
when chloride, urea, sugar, etc., are added to such blood. In Nussbaum’s 
experiments (26) it was noted that no secretion of urine occurred after 
ligation of the renal artery, until urea was injected. This was confirmed 
by Adami, and by A. P. Beddard (27, pp. 26 and 27). An experiment 
made by Magnus is most interesting in this connection (21, III, pp. 213- 
216). He finds that increasing the quantity of blood in a rabbit or dog 
by 84 per cent, by transfusion of blood from another animal which has 
been kept under identical conditions, does not cause diuresis. In this 
he confirms Ponfick’s observation (quoted /ézd, p. 212). If the composi- 
tion of the injected blood is not the same — as when the animal has 
received an injection of sodium sulphate — diuresis results. 

Spiro (28) finds (p. t50) that injections of solutions of colloids — gums 
or gelatine — into animals which have fasted twenty-four to forty hours, 
gave no diuresis, although the quantity of blood was increased by 
78 to go per cent. If, on the other hand, the animals had been fed, and 
especially, watered freely, a good diuresis resulted (p. 152). This could 
be interpreted as showing that these colloids hold their own “ aggregate ”’ 
composition against the urinary filtration tendency; and that diuresis will 
only result from them if filtrable substances are present in excess. Indeed 
even caffein cannot produce diuresis after the injection of colloids unless 
such filtrable substances are present. 


Forster’s theory could also be made to explain why one animal 
reacts to a given amount of a diuretic by a large diuresis, whilst 
another animal may show but very little response: an animal poor 
in substances capable of passing into the urine would react less 
readily than one which contains an excess of such substances. 


The Mechanism of the Retention of Chlorides. 165 


The difference of diuretic effect of equimolecular solutions of 
different salts could also be explained on physical principles by this 
theory : if it be assumed that sodium chloride enters more readily 
into non-filtrable combination with the colloids than does sodium 
sulphate, the latter should produce the greater diuretic effect. 

A further support of Forster’s theory could be sought in the modern 
theories of the physical condition of colloids. In accordance with 
these, the non-filtrable proteid salt compounds assumed by Forster's 
theory could be conceived as ‘‘ molecular aggregates.” 


It must be granted that all the physiological phenomena for the 
explanation of which I have invoked Forster’s theory, could be 
equally well explained by the theory of vital secretion, by assuming 
that an abnormal composition of the blood serves as a necessary 
stimulus to the excreting cells, causing them to eliminate what 
substances are responsible for the abnormality. 

The physical theory of Forster possesses the advantage of simplic- 
ity. Attractive and adequate as this theory appears on cursory 
examination, it must be said to lose very much of its force when it is 
examined more critically. It can only be allowed to stand if it is 
proved : that the ratio of salts and colloid varies only within certain 
limits; that proteids do combine with salts ; that filters can be con- 
structed which retain the proteids and the greater part of the salts 
of the serum, but which allow added salts to pass freely. 

3. Is the ratio of salts to colloids constant ?— Forster’s theory con- 
ceives the combination of proteids and salts to contain variable 
proportions, tending, however, toward a certain constant. When the 
salts fall below this constant, they tend to be retained ; if they exist 
in greater proportion, they will be excreted. Asa matter of fact, it 
is found that when sodium chloride is administered, a certain amount 
is retained, and the quantity retained is the greater, the lower the 
original salt-content of the body. The experiments of Spiro on the 
injection of colloids also speak for the fact that colloids poor in 
salts tend to retain sodium chloride in the body. It can also be 
noticed that the ratio of organic matter and ash of the serum varies 
within considerable limits in different animals. All these facts agree 
well with Forster’s theory, but it must be acknowledged that an 
actual combination of salt and colloid is not necessary for their 
explanation. 


166 Torald Sollmann. 


4. Has any combination between proteids and salts been demon- 
strated? — The answer to this question must be entirely in the nega- 
tive. Bugarszky and Liebermann (33) found that no change occurs 
in the freezing-point of a sodium chloride solution when egg albumin 
or albumose are added to it. I have confirmed this on albumoses, 
and have further shown that solutions of these dissociate precisely as 
does their ash (34). Indeed the freezing-point of serum can only be 
explained by assuming dissociation of its salts. Whilst we do not 
yet know sufficient about “molecular aggregates’’ to affirm that 
their constituents cannot exist in dissociated form, yet it is scarcely 
conceivable that salts could exist combined into aggregates, and still 
possess the physical properties of free ions. The observation of 
Moore and Parker (35, p. 280) that the aggregates of soap possess 
a physical molecular weight twenty to sixty times as great as their 
chemical molecular weight, certainly speaks against this assumption. 

5. The filtration of colloid solutions.— Ludwig’s theory assumes 
the glomeruli to represent a filtering apparatus which allows the 
passage of water and salts, but not of proteids. Martin (30) has 
shown how to construct, artificially, filters of this kind by coating 
Pasteur-Chamberland filters with gelatin, or silicic acid, and filtering 
under pressure. These filters form a very close approximation to the 
theoretical filter assumed by Ludwig. They would seem to be a fair 
method of directly testing the question: Whether the filtrate from 
a colloid mixture really represents the mixture minus the colloids ? 
or in other words: whether the filtrate from, say, serum, is simply a 
proteid-freed serum ? 

The experiment was tried by E. H. Starling (31, p. 318). He 
obtained the erroneous result that the filtrate has within ;},° C, the 
same A as the serum, — ?.¢., that the filtrate is purely a proteid-freed 
serum. The error was shown by Waymouth Reid (32, p. 169), who 
investigated the subject more fully, and with due precautions. He 
finds that the filtrate is very markedly less concentrated than the 
serum; that it has been deprived, not only of the proteids, but also 
of the greater part of the non-proteid solids, and of a fair amount of 
the ash. 


‘The organic non-proteid solids were reduced from 5.06 to 0.78 per cent. 
The ash ws i? a —e * 9.42 105.05) Soe 
The A ff a = by 0.035 to 0.060°C. 


In dog’s bile the ash was reduced from 4.42 to 0.76 per cent. 


The Mechanism of the Retention of Chlorides. 167 


The fact that the filtrate is more dilute than the serum cannot, 
however, be considered as a confirmation of Forster’s theory. This 
theory demands that the salts should be mainly in the unfiltered por- 
tion, but the experiment shows that the salt content of the filtrate is 
but very little lowered. Nor can the retention of organic substances 
be regarded as due to a combination, for a similar retention occurs 
when a pure solution of dextrose is subjected to this filtration; the 
filtrate is in this case also less concentrated than the original 
solution. 

Waymouth Reid’s experiments show that filtration may retain 
certain substances, and in this way preduce a considerable diminu- 
tion in the concentration of the filtrate. But this retention is not 
due to any proteid-salt combination, but to the fact that the molecules 
of the dissolved substance pass less readily through the filter than 
does water. 

We must conclude that the direct evidence of chemical and physical 
experiments ts against Forster's theory. 


VI. Cusuny'’s THEORY OF THE REABSORPTION OF CHLORIDES. 


The theory that sodium chloride is reabsorbed from the glomerular 
fluid, presumably by the epithelium of the convoluted tubules, has found 
its best support in the recent experiments of Cushny (23 and 36). 
Cushny, in agreement with Starling (31, p. 317), looks upon the 
glomerular liquid as a simple filtrate of the serum, containing prac- 
tically all its non-proteid constituents in the original amount and 
ratio (p. 449). As this passes through the urinary tubules, water 
and the more diffusible constituents are reabsorbed (p. 429). Sodium 
chloride is very readily absorbed; P,O;, SO,, and urea with increas- 
ing difficulty. A retention of chlorides —a paucity of the urine in 
chlorides — can, according to him, be due only to two causes: to a 
. paucity of chloride in the blood which would lessen the filtration; or 
to slow excretion of urine which would favor reabsorption. It must, 
if I understand him correctly, be strictly proportional to these. 


These conclusions are based on the following experiments : 

Cushny (23) injected into rabbits a mixture of equimolecular solutions of 
sodium chloride and sodium sulphate. In the urine the equimolecular 
ratio of Cl : SQ, was (p. 434): 

In the early stages of active diuresis, *: 108 : 94; 
In the later stages of slow diuresis, :: 110 : 156. 


168 Torald Sollmann. 


In other words, the concentration in the sodium chloride was proportional to 
the diuresis, whereas the sodium sulphate was inverse to the diuresis. 

The difference between chlorine and sulphate is brought out still more 
clearly by another mode of calculating (p. 436) : 

‘The kidneys secrete in the first stage a urine in which the chlorine stands to the 
chlorine of the plasma as 2 : 3; in the last phase, as 1 : 5; sulphate, 
first phase, 2-0. srs last Grote. ur 

These facts are most readily explained by assuming that the sodium chloride is 
reabsorbed, in proportion to the time during which the urine remains in the 
kidney ; whereas the sodium sulphate is not capable of reabsorption in the 
same degree. Extending the method to urea (p. 444) and phosphates 
(p. 442), he finds that these are also not readily absorbed, urea being the 
most resistant of all. 

The dependence of the amounts of chlorides on the diuresis is still more 
strikingly established by Cushny’s experiment on partial occlusion of the 
ureter (36) ; operating as in his other experiments, but partially occluding 
one ureter and comparing the urine from the two kidneys, he finds that 
on the side of the occluded ureter the water is lessened by 66 per cent, 
chlorine by 82 per cent, sulphate by 30 per cent. 


These conclusions are further borne out by the results of numerous 
experiments made by other observers for other purposes, such as 
those of Munk and Senator (37), Magnus (21), as also by my 
experiments (22). They also agree very well with clinical cbser- 
vations. Von Koranyi’s theory, which is based mainly on clinical 
facts, coincides with Cushny’s. 

This theory involves a selective absorption of the urinary in- 
gredients. Viewing it, as we must, as a vital process, the difference 
in the absorbability of the chlorides and of urea finds many analogies. 
In my former paper (22, p. 19), in advancing this difference as 
evidence against the theory of reabsorption, I had in mind a physical, 
not a vital reabsorption. 


The reabsorption of chlorides may be considered as demonstrated. — 
I take exception, however, to two other views of Cushny, — viz., that 
the reabsorption is an adequate explanation for the practical dis- 
appearance of chlorides in the urine, and that the glomerular fluid is 
a filtrate, identical with proteid-free serum. 


The Mechanism of the Retention of Chlorides. 169 


VII. Is THE REABSORPTION A SUFFICIENT EXPLANATION FOR THE 
ABSENCE OF- CHLORIDES IN THE URINE? 


According to Cushny’s theory the amount of chlorides in the urine 
is determined by two factors: The diuresis, and the quantity of 
chlorides in the serum. It can, however, be easily shown that the 
excretion of the chlorides can be altered in either direction without 
altering these two factors in a corresponding degree; and that it is 
sometimes even the opposite of what is demanded by Cushny’s 
theory: 


In starvation, the quantity of the urine is practically unaltered, or may be even 
greater than normal. With sodium sulphate injection, the diuresis is very 
large. In both these cases the chlorides disappear from the urine. In 
other words, the results are precisely the opposite of those demanded by 
Cushny’s theory. Nor can this discrepancy be explained by a propor- 
tionately altered composition of the blood. For according to Cushny’s 
theory the amount of chlorides for a given excretion of urine should be 
strictly proportional to their quantity in the blood. Yet the chloride of 
the blood and the tissues is scarcely altered by starvation. 

In the case of sodium sulphate we have analytical data which show the insuff- 
ciency of Cushny’s explanation in a still more striking light, if possible. 

On injecting sodium sulphate, Magnus (21, II, p. 417) found that the sodium 
chloride of the serum sank only from 0.632 per cent to 0.6 per cent; 
whereas it sank in the urine to only 0.05 per cent ; sodium sulphate rose 
in the serum from 0.034 to 0.272 per cent; in the urine to 3.124 per 
cent. This could only be explained by an enormous reabsorption of 
chlorine ; yet there was a good diuresis. 

In my own experiments I found: 


EXPERIMENT VI. 


Injection of sixty-three per cent of the blood quantity of Na,SO, solution A 0 544. 


Serum. Urine. Urine 


: secretion. 
NaCl. Other salts. eee} Me 
er cent. Per cent. si 


NaCl. | Other salts. | 


Before injection Sy 2.67 4.27 


10 minutes after 5 5.62 


30 minutes after | 58 | 4.40 


170 Torald Sollmann. 


EXPERIMENT VII. 


Injection of ninety-nine per cent of the blood quantity of Na,SOy, solution A 0.606. 


Serum. Urine. Urine 
secretion. 


NaCl. Other salts. ee kg. Pe 
Per cent. Per cent. bess 


NaCl. | Other salts. 


Before injection 5. 8s 1.05 6.56 


4 minutes after 
15 minutes after 


45 minutes after 


It is seen that the excretion of chlorides is suppressed when the blood contains 
quite a high percentage of this ion, and when the diuresis is very large. 

In Katsuyama’s experiments on the effect of diuretics (29) on rabbits de- 
prived of food and water, he finds, indeed, that diuresis increases the per 
cent of sodium chloride; yet this fact, instead of favoring Cushny’s 
explanation, really controverts it; for the increase is out of all propor- 
tion. Indeed, these diuretics increase the chloride equally, whether they 
produce diuresis or not. 

Thus in Experiment III, p. 237, the quantity of normal urine was 67; Cl, 
trace. On the day after giving urea, the quantity was 64; Cl, 0.17:0. 

In Experiment VI the normal quantity of urine was 39; Cl, trace. After 
diuresis, the normal quantity of urine was 32; Cl, 0.48 per cent. 

In the case of fasting in hypnotic sleep, observed by Hoover and Sollmann (18), 
an increase of diuresis from 410 to 560 c.c, instead of increasing the 
percentage of chlorine, lessened it from 0.6 to 0.43 per cent. 


Similar instances could be multiplied. It is evident that neither 
deficient filtration, nor reabsorption from prolonged sojourn in the 
tubules, can explain these conspicuous exceptions to Cushny’s theory. 

When the diuresis is not greatly altered, there can be no doubt 
that the composition of the urine depends mainly upon the com- 
position of the serum. The very fact that the intravenous injection 
of any substance capable of passing into the urine, increases the 
excretion of this substance, is sufficient to prove this. Yet even 
much smaller variations in the blood find expression in the urine. 
Thus V. Koranyi (38, p. 10) showed that in the rabbit the factor 


A ; Je a\a a, ee : 
NaCl of the urine varied in the same direction as the corresponding 


The Mechanism of the Retention of Chlorides. 171 


factor of the serum. Only —andgthis is important — the range is 
much greater in the urine (urine, 1.13 to 13.92; serum, 0.86 to 1.25). 
The dependence of large changes in the chloride excretion upon small 
changes in the blood can only be explained by a vital theory; in the 
case of chloride retention, either by a stimulation of the absorbing 
cells, or by a cessation of the secretion of chloride. I shall discuss 
this question in the next paragraph. 


VIII. Is THE GLOMERULAR FLUID FORMED BY FILTRATION OR BY 
SECRETION ? 


By the filtration theory of urine formation the blood-pressure 
in the glomerular capillaries must exceed the ureter pressure by 
a greater amount, than the osmotic pressure of the serum exceeds 
the osmotic pressure of the glomerular fluid. Otherwise no filtration 
is possible, for the blood-pressure would not suffice to overcome the 
osmotic pressure. The composition of the glomerular fluid differs 
from that of the serum at least by the proteids. Starling (31, p. 321) 
has calculated the osmotic pressure of the serum proteid as 25 to 
30 mm. of mercury. 

In order to establish the filtration theory as the only factor in the 
formation of urine, it is necessary to prove: 

(1) That the secretion of the urine stops when the blood-pressure 
falls to less than 40 mm. mercury above the ureter-pressure. 

(2) That the osmotic pressure of the glomerular fluid differs 
from that of the serum by not more than 30 mm. of mercury. 

(1) Does the secretion of urine cease when the ureter and blood 
pressure differ by less than 40 mm. of mercury ? — Starling (31) found 
in his experiments that the secretion of the urine stops when the 
carotid pressure falls to 40 mm. mercury. He concludes from this 
that the blood-pressure is sufficient to filter a proteid-free urine. 
Moore and Parker (35) justly criticise this conclusion, on the ground 
that it is not the carotid, but the glomerular pressure which must be 
effective, and the pressure in the glomeruli must be considerable less 
than that in the carotid. 

The experiments of Gottlieb and Magnus (21, V) confute Star- 
ling’s conclusions still more effectually. These authors found that 
the urine was secreted with an arterial pressure as low as 16 mm. 
(p. 253); avery small quantity was secreted even when the arterial 
pressure surpassed the ureter pressure by only 2 mm. Hg (p. 254). 


I7o Torald Sollmann. 


(2) The osmotic pressure of the glomerular filtrate may differ from 
_ that of the serum by much more than 40 mm, mercury. — The experi- 
ments of Waymouth Reid (32) have shown that the A of the gelatin 
filtrate from serum is reduced by at most 0.035° C. The concentra- 
tion of the urine may, however, differ from that of the serum by even 
much more than this quantity. The smallest A of urine found by 
Dreser (39) was 0.16° C; this occurred after drinking one-half litre 
of Bavarian beer. Von Koranyi (40) states that the A of the urine 
may fall below 0.10°C. No observations of the A of the blood were 
made in these cases, but it is known from other sources that this 
could certainly not have been reduced to less than 0.45. This givesa 
difference of 0.35°C, between the A of the blood and that of the urine, 
corresponding to an osmotic pressure of 3140 mm. Hg, which would 
need to be overcome by the blood-pressure in order that filtration 
could take place. 

I consider that these facts establish that //tration 7s not suffictent 
to account for the phenomena of urine formation. A vital secretion 
must be assumed, and I believe this to hold true of all the urinary 
constituents. In view of this I consider that the retention of salts 


sorption. 

Whilst advocating the existence of a vital secretory process, I do 
not mean to deny that physical processes co-operate to a very im- 
portant extent in the formation of urine. The kidney cells, especially 
the cells of Bowman’s capsule, must be affected by osmotic pro- 
cesses in much the same manner as other cells. The conditions are 
particularly favorable to filtration. The dependence of the urine 
secretion upon the blood flow is well demonstrated; although this 
dependence could be explained on the secretion theory, an increased 
circulation must also be a potent aid to filtration. 

The importance of filtration becomes particularly great when salts 
are injected, or when hydraemia is produced; for the former furnish 
a larger amount of filterable constituents, and the latter causes an in- 
crease of the glomerular capillary pressure. A similar coexistence, 
and variable importance, of a secretion and filtration process may 
be noted in the intestine. No one doubts that the intestinal epithe- 
lium ordinarily seereées the intestinal fluid. But when large quantities 
of saline solution are injected, a fluid is poured into the intestine with 
a rapidity which can only be explained by filtration. The following 
may serve as an example: * } 


— 


The Mechanism of the Retention of Chlorides. 173 


Experiment VIII: 1500 c.c. of isotoffic sodium chloride was injected into the 
femoral vein of a dog of 3.9 Kg. 1050 of this was recovered from the 
intestine, and 420 c.c. of this within the first hour. 


IX. CONCLUSIONS. 


1. The disappearance of chlorides from fever urines is due practi- 
cally entirely to the deficiency of chloride income. It cannot possess 
any diagnostic value. 

2. The mechanism of the retention of chlorides is not explained by 
any physical theory, but must be a vital process. Lessened secretion 
and increased reabsorption are probably both concerned in the 
retention. 

3. The filtration theory of urine formation is inadequate. 


X. BIBLIOGRAPHY. 


1. F. ROHMANN: Zeitschrift fiir klinische Medicin, 1879, i. p. 513. 

REDTENBACHER: Zeitschrift der kaiserlich-k6niglichen Gesellschaft der 
Aerzte zu Wien, 1850. (Quoted by ROHMANN.) 

3. E.Uwnrun: Archiv fir pathologische Anatomie, 1869, xlviii, p. 227. 

4. J. P. LANGLoIs and Cu. RicHET: Journal de physiologie et de pathologie 
générale, 1900, ii, p. 742. 

J. ForSTER: Zeitschrift fiir Biologie, 1873, ix, p. 297. 

Gramatchikov: Inaugural Dissertation, St. Petersburg, 1890 (Russian). 
Quoted in “ Digest of Metabolism Experiments,” ATWATER and LANG- 
WORTHY, Bulletin 45, Office of Experiment Station, United States 
Department of Agriculture, 1898. 

7. LAUDERK-BRUNTON: “Text-book of Pharmacology, etc.,” Philadelphia, 

1885, p. 503. The statement appears to be based on the experiments 
of Voir (ROHMANN, p. 520). 

8. O. KRUMMACHER: Zeitschrift fiir Biologie. 1900, xl, p. 173: 

g. A. Kast: Zeitschrift fiir physiologische Chemie, 1888, xii, p. 267. 

to. M. ScHUBERT: “ Ueber den Stickstoff und NaCl Umsatz wahrend der 

Bildung und nach der Punktion des Ascites bei Lebercirrhose,” 
Inaugural Dissertation, Breslau, 1895. 

11. WALDVOGEL: Archiv fiir experimentale Pathologie, 1901, xlvi, p. 41. 

12. KOBLER, VOGEL, and HuSCHE: Quoted by LINDEMANN (13, p- 25). 

13. L. LINDEMANN: Deutsches Archiv fiir klinische Medicin, 1899, Ixv, p. 1. 

14. STICHER and MARKWALDT: Quoted by Kasrt. 

15. SCHOPP, 1892: Quoted by ATWATER and LANGWORTHY. 

16. FAtck: ‘*‘ Handbuch der dietatischen Heilmittellehre,” p. 128, ete. 

Quoted by Fcu. MULLER, Inaugural Dissertation, Marburg, 1872. 
17. BIDDER and SCHMIDT: Quoted by FORSTER. 
18. C. F. Hoover and T. SoLLMANN: Journal of experimental medicine, 


1897, li, p- 405. 


oo 


oe 


O2 WD Wo Ld 
td 


roe) 


bo KW WwW 
Nn 


o>) 
oe) 


ee ule. 


Oo 


a | 


Torald Sollmani. 


A. CAHN: Zeitschrift fiir physiologische Chemie, 1586, x, p. 522. 

R. MaGnus: I. Archiv fiir experimentale Pathologie, 1900, xliv, p. 68; 
II. Jéd, p. 396. III. /bzd, xlv, p. 210, IV. GoTTLiEB and MAG- 
NuS, /bzd, xlv, p. 223. V. Jbid, p."248. 

T. SOLLMANN: Archiv fiir experimentale Pathologie, 1901, xlvi, p. I. 

A. R. Cusuny: Journal of physiology, 1902, xxvii, p. 429. 

3ERGMANN: Archiv fiir experimentale Pathologie, 1901, xlvii, p. 77. 

I. MuNK: Archiv fiir pathologische Anatomie, cvii, p. 291; cxi, p. 434. 

NusssBaum: Archiv fir die gesammte Physiologie, 1878, xvii, p. 580. 

A. P. BEDDARD: Journal of physiology, 1902, xxvili, p. 20. 

K. Spiro: Archiv fiir experimentale Pathologie, 1898, xlvi, p. 148. 

K. KarsuyAMA: Zeitschrift fiir physiologische Chemie, 1901, xxxii, 
p- 235. 

C. J. MARTIN, Journal of physiology, 1896, xx, p. 364. 

Ik. H. STARLING, Journal of physiology, 1899, xxiv, p. 317- 

WayMouTH REID: Journal of physiology, 1go1, xxvii, p. 161. 

St. BuGArszky and L. LIEBERMANN: Archiv fiir die gesammte Physi- 
ologie, 1898, Ixxii, p. 51. 

T. SOLLMANN: This journal, 1902, vii, p. 203. 

B. Moore and W. H. PARKER: This journal, 1go2, vii, p. 261. 

A. R. Cusuny: This journal, 1901, vi, p. xvii. 

I. Munk and SENATOR: Archiv fiir pathologische Anatomie, 1858, cxviii, 
pide 

Fr. v. KORANYI: Zeitschrift fir klinische Medicin, 1897, xxxiii, p. 1, 
1898, XXxiv, p. I. 

DrESER: Archiv fir experimentale Pathologie, 1892, xxix, p. 303. 

Fr. v. KORANYI: Berliner klinische Wochenschrift, 1899, xxxvi, p. 782. 

C. Pororzky: Archiv ftir die gesammte Physiologie, 1902, xci, p. 584. 
(The series of articles by FILEHNE and his pupils, of which this paper 
is a part, appeared some time after the present paper was ready for the 
press. It was therefore impossible for me to discuss them as fully as 
they deserve.) 


DOES POTASSIUM CYANIDE PROLONG THE LIFE OF 
THE UNFERTILIZED EGG OF THE SEA-URCHIN? 


By F. P. GORHAM anpb R. W. TOWER. 


[From the Anatomical Laboratory, Brown University, and the Scientific Laboratory of the 
OU. S.. Fish Commission, Woods Hole, Mass.] 


INTRODUCTION. 


N a recent paper by Loeb and Lewis! on the “ Prolongation of the 
Life of the Sea-Urchin Egg by Potassium Cyanide,” the state- 
ment was made that “ there are two kinds of processes going on in 
the egg: one which leads to the death and disintegration of the egg, 
—a mortal process; and a second, which leads to cell divisions and 
further development. The latter process inhibits or modifies the 
mortal process. . . . According to this idea, death and disintegration 
are due to specific processes which take place in the egg and possibly 
in other or all living matter. These processes must be checked in 
order to render life possible.” They consider these processes to be 
chemically catalytic phenomena, and that consequently any agency 
which could stop or retard the catalytic action, without injuring the 
living matter, would of necessity check the mortal process. And they 
state that “‘ among all the agencies which act in this way, potassium 
cyanide seemed to meet this condition most perfectly. It weakens 
or inhibits a number of enzymatic processes in living matter without 
necessarily altering the constitution of the latter. When the potas- 
sium cyanide is permitted to evaporate, the original condition of the 
system may be restored.” 

In the main, Loeb and Lewis found that eggs kept in normal sea- 
water could not be fertilized after forty-eight hours. After this time 
the eggs “‘ became a sticky mass and assumed a dirty brownish color, 
and this was the beginning of complete disintegration and putrid 
decay.” But if the eggs were put into a solution of ;4,5 KCN for 
seventy-two hours, then removed to normal sea-water and fertilized, 
they would develop in every case into plutei. In some instances 


1 Logs and Lewis: This journal, 1gor, vi, p. 305. 
175 


176 fF. P. Gorham and R. W. Tower. 


plutei were obtained from eggs that had been in the poisoned sea- 
~water from ninety to one hundred hours, or about four days, while 
those that had remained for five days developed only to the four-cell 
stage. 

In Loeb’s experiments the dishes were loosely covered, and con- 
sequently the potassium cyanide solutions containing the eggs were 
constantly growing weaker. If, on the other hand, the evaporation 
was prevented by complete closure of the dishes, the eggs were so 
affected that after an exposure of sixty-six hours only a few of the 
eggs were able to segment when fertilized. For the detailed experi- 
ments the reader is referred to the original article. 

Upon reading the above paper, the authors of the present article 
concluded that many of the deductions were not warranted by the 
results of the experiments described, and that other interpretations 
were both possible and logical. There seemed to be nothing to show 
whether the potassium cyanide was acting directly on the living 
matter of the egg, or was producing some change in the external en- 
vironment which in its turn affected the egg. Could it not be 
possible that the potassium cyanide solutions prolong life only be- 
cause they kill the bacteria that ordinarily cause the egg to die? 

The fact that the potassium cyanide solution which Loeb found 
most favorable for prolonging the life of the egg is strongly antiseptic 
is important in this connection. Solutions of greater strength kill the 
egg, while weaker solutions are not so effective in prolonging its life. 

The fact, too, that the optimal solution of potassium cyanide acts 
beneficially only when allowed to become gradually weaker, does not 
prevent us from thinking that after all the potassium cyanide is 
slowly killing the egg, and is beneficial only because it acts on the 
bacteria before it does on the egg itself. 

It seemed also that there might be other factors at work which had 
been overlooked by Loeb and Lewis. In order to convince ourselves 
of the facts, we repeated, as closely as possible, Loeb’s experiments, 
first upon the eggs of the winter-flounder, and then upon the eggs of 
the sea-urchin. Experiments to determine the antiseptic value of the 
potassium cyanide solutions used by Loeb were also made. 


ANTISEPTIC VALUE OF THE POTASSIUM CYANIDE SOLUTIONS. 


In the preliminary experiments, to test the antiseptic value of the 
potassium cyanide solutions, it was found that the addition of potas- 
sium cyanide to salt-water, so that the resulting solution was 7% 


Av 


Potassium Cyanide and the Unfertilized Egg. 177 


KCN, reduced the number of bacteria from some 3,000,000 per c.c. 
to about 2,000 in twenty-four hours, while a =”, solution reduced the 
number to about 1,000 in the same time. 


EXPERIMENTS ON EGGS OF THE WINTER-FLOUNDER ( Pseudopleuro- 
nectes americana). 


The ova of this animal are not adapted to the study of fertilization 
and segmentation, but were all that were available in March, when 
these experiments were undertaken. The experiments of Loeb were 
repeated on the eggs of the flounder, and the number of bacteria in 
the different solutions was determined at various periods. 

The protocol of these experiments follows : — 


TABEEY I 


Number of | 


CRS Number of Number of 
bacteria in cao * 
ml bacteria bacteria 
: - each’e:c. ||) «= Ne 
Solutions. hart aos in each c.c. in each c.c. 
> 
meee after after 
27 hours. 48 hours. 
added. : 8 
BYU CGL) 6 sae nals vse ue os etl os 700 st Countless. 
Salt-water containing eggs . . . . 500 150,000 Countless. 
Salt-water containing sperm =. . ae 460,000 Countless. 
c=) 
Salt-water containing fertilized eggs ae 460,000 Countless. 
> oo 
qoop KCN containing eggs. . . . +00 | 180 39.500 
nt EA Nares oe : 5 c 
so00 KCN containing eggs . | 45 90 
n r T ane 2 2 
tooo KCN containing eggs . | 36 32 
| 
5p KCN containingeggs ... . 65 25 
=44 KCN containing fertilized eggs . 3 25 5 
nt ey oy eS som . > 
zo0 KCN containing eggs . .. . 21 8 
” 7 ONT aa 
zap KCN containingeggs . .. . x. 0 0 
539 KCN containingeggs . . .. |]. .: 0 | 0 
s3p KCN containingeggs . .. . 13 0 0 


From the above table we notice: 
1. The enormous number of bacteria in the salt-water containing 
the eggs and sperm. 


178 F. P. Gorham and R. W. Tower. 


2. The number of bacteria is reduced by the addition of potassium 
cyanide in proportion to the strength of the solutions. 

3. The decrease in numbers does not continue after the first few 
days, particularly in the weaker solutions. This is doubtless due to 
the gradual loss of potassium cyanide. 


EXPERIMENTS ON THE EGGS OF THE SEA-URCHIN (Arbacta 
punctulata). 


The experiments of Loeb and Lewis were carefully repeated. 
Analyses were made of the various solutions to determine the num- 
ber of bacteria present. Control experiments were made in every 
case and all the conditions prescribed by Loeb were carefully fol- 
lowed. After one complete set of experiments was carried through, 
all the solutions were destroyed, a new 7 KCN solution was pre- 
pared, and the entire experiment was repeated. The results of the 
two experiments were found to be identical in every particular, and 
are given in Table II. 

On the whole the results agree with those a Loeb with the excep- 
tion that our g7oo and yoo KCN solutions preserved the eggs the 
longest, while Loeb found the -”, and ;(79 the most potent. This 
discrepancy may be due to a difference in the purity of the potassium 
cyanide salt used. In our case a quantitative analysis showed it to 
contain ninety per cent potassium cyanide. 

In our experiments the most favorable results were obtained at the 
end of ninety-six hours, while Loeb reports seventy-five hours as the 
optimal time. This seems to indicate that some other factor than 
potassium cyanide is at work. The number of bacteria present in 
the various solutions at the end of ninety-six hours, as shown by the 
table, is in proportion to the concentration of the potassium cyanide, 
The stronger solutions are sterile. 


PROTOZOA. 

During the above experiments it was noticed that under the same 
conditions more bacteria appeared at one time than at another. 
Further investigation showed that there were present also varying 
numbers of protozoa. The relation of these organisms to the life of 
the sea-urchin larvee can be seen in Table III. 

It is evident that when the protozoa are numerous the life of both 
eggs and larve is preserved, The protozoa destroy enormous numbers 
of bacteria. Wherever there are large numbers of protozoa there are 


oii \ ie 


Potassium Cyanide and the 


Solutions. 

240 c.c. in 

each dish 24 hours. 

+- 5 c.c. of 
eggs. 

(Control) 600,000 bacteria 

Sea-water. perc.c. Eggs fer- 

1000 bacteria! tilized, and de- 

per c.c. velop normally 


into plutei. Of 
the unfertilized 
eggs none devel- 
ops. 

sz509 KCN | Same as control. 

Develop. into 
; plutei. 
| Same as control. 


too KCN 


T6000 


z000 KCN Same as control. 


zany KCN ! Nearlyallsegment 
| and develop to 


Nearly all segment 
and develop to 
swimming larve. 


tog0 KCN | Nearly all seg- 
ment and few 
develop to swim- 


ming larve. 


4, KCN Nearly all seg- 

ment. Some ir- 
| regularly. Few 

larve. 

zion KCN Few segment ir- 
regularly. Very 
few larve. 

360 KCN 

x39 KCN No egg segments. 


few bacteria. 


is due to the elimination of bacteria. 


swimming larve. | 


TA®LE II. 


4S hours. 


390,000 bacteria 
perc.c. Feweggs 
segment irregu- 
larly in first ex- 
periment. Insec- 
ond experiment 
no egg develops. 


No egg segments. 


No egg segments. 


50 per cent of eggs 
segment and de- 
velop to 
ming iarve. 

Most segment and 
develop to swim- 
ming larve. 
Some irregularly. 


No egg segments. 


Many segment. 
Some irregu- 
larly. 


No egg segments. 


No egg segments, | 


No egg segments. 


swim- | 


Unfertilized Egg. 


72 hours. 


17,000 bacte- 
May Pen icc. 
No egg seg- 
ments. A 
sticky, putrid 
mass. 


No egg seg- 


ments. 


No egg 
ments. 


Some eggs seg- 
ment and de- 
velop to plu- 
tei. 

Very few seg- 
ment. 


Very few seg- 
ment. 


No egg seg- 
ments. 


| No egg seg- 
ments. 

No egg seg- 
ments. 


No egg seg- 
ments. 
No egg 
ments. 


96 hours. 


850,000 bacteria 
perc.c. No egg 
segments. 


No egg segments. 


950,000 bacteria 


perc.c. No egg 
segments. 
$00,000 bacteria 
perc.c. No egg 
segments. 
'550,000 bacteria 
perc.c. All seg- 
ment and de- 


velop to swim- 
ming larve. 
380,000 bacteria 
per c.c. Most 
segment and de- 
velop to swim- 
ming larve. 
| 35,000 bacteria per 
c.c. No egg seg- 
ments. 


11,000 bacteria per 
c.c. No egg seg- 
ments. 


No bacteria. No 
egg segments. 


No bacteria. No 
egg segments. 
No bacteria. No 


| egg segments. 


Here again the beneficial action on the eggs or larvze 
To test the accuracy of this 


deduction a special experiment was made to determine more exactly 
the ability of the protozoa to free a solution from bacteria and thus 
preserve the life of eggs or larve. 


180 F.. P. Gorham and R. W. Tower. 


A solution was taken containing developing eggs which were being 
destroyed rapidly by large numbers of bacteria. This was inoculated 
with a considerable number of protozoa. Within twenty-four hours 
the bacteria had eaten up the dead embryos, the protozoa had greatly 
reduced the number of bacteria, and the larvze which had survived 
during the struggle were alive and swimming actively. In the con- 
trol experiments all the larva were dead. 


TABLE, Tur 


; eer en) eNUmiberm of 
Solution in which | d fie: 
aS cE REN | ays after 
€ges were kept. | fertilization. 


Condition of Number of 
larve. protozoa. 


Sea-water Dead None. 


3BGD Few swimming Many. 


16000 Alive Many. 


wt 


2 . 7 - a 
T0000 Swimming Very many. 


” 


iso0 : All alive Very many. 


All dead None. 


7000 
2000 All dead None. 


Swimming | Many. 


2000 


[000 Dead None. 


The relation of bacteria to protozoa in standing sea-water is well 
shown in the two experiments of Table IV. 


TABLE TV. 


| 


= | 
Number : Number : 
of | Bacteria Protozoa per hae | Bacteria Protozoa per 
| c ) | | mS 
pegcics BRCIGs Peper cic: bya. 
hours. 


hours. | 


0 1,000 l or 


tS 


1,800 l or 2. 


600,000 S to 10. 


25,500 


1I20'000" 15 48 390,000 Numerous. 


72 120,000 10 to 20. We, 17,000 Very numerous. 


2,300 Numerous. 


1,600 


Very numerous. 


Potassium Cyanide and the Unfertilized lege. 181 


foto) 


EXPERIMENTS WITH STERILE SEA-WATER. 


In order to determine whether the preservation of the life of the 
egg is due to the direct action of the potassium cyanide on the egg, 
or to the antiseptic action of the potassium eyanide whereby the 
injurious action of bacteria is eliminated, experiments of another kind 
were made in which potassium cyanide did not figure, and in which 
no bacterial contamination was possible. 

The ideal experiment of this kind would be to place the sterile eggs 
of the sea-urchin in sterilized sea-water and to remove a few on suc- 
cessive days for fertilization. It was found possible to collect the 
eggs from the sea-urchin in a perfectly sterile condition and to 
transfer them to sterilized sea-water, where they were kept for eleven 
days at laboratory temperature uncontaminated by bacteria. The ac- 
companying data, Table V, gives the results of an experiment of this 
kind. 

TABE ESV. 


Number of 


| } 
days before 2 days. 3 days. |4 days.|5 days.| 7 days. ll days. 
fertilization. | 


Sterile eggsin | Develop | Develop | Develop Develop into plutei. 
sterile into into into 
sea-water. plutei. | plutei. plutei. 


: : : | r; . . 
Sterile eggs in | Develop Few No egg Eggs are now a putrid mass. 
ordinary ito |) “eggs segments. 
sea-water. | plutei. | segment. 
| 


Ordinary eggs in | Develop | No egg No egg Eggs are now a putrid mass. 
ordinary | into | segments. | segments. 
sea-water. plutei. 


As will be seen, the sterile eggs that were kept in the sterile sea- 
water for eleven days could at any time be fertilized and would develop 
into normal plutei; while in all cases, eggs that were not kept sterile 
died in about forty-eight hours. At the end of eleven days, all the 
sterile eggs were alive. How much longer they would live, was not 
determined. Thus, by preserving the sterile condition of the eggs, 
their life has been “ prolonged” nearly four times as long as when 
treated with potassium cyanide solutions. 

It seems probable, therefore, from a consideration of these experi- 
ments, that potassium cyanide acts always as a poison; that any sup- 
posedly beneficial action is due entirely to its antiseptic properties. 


182 F.. P. Gorham and R. W. Tower. 


CONCLUSIONS. 


1. The action of potassium cyanide is only an indirect one, z. e. 
killing or inhibiting the bacteria, and thus giving the eggs a more 
favorable environment. 

2. Inall experiments with unsterilized sea-water, the protozoa enter 
as an important bacteria-destroying factor which must be considered 
in interpreting the results. 

3. Sterile sea-water ‘“‘ prolongs” the life of the egg of the sea- 
urchin much longer than Loeb’s most favorable potassium cyanide 
solutions. 

4. Both our own experiments and those of Loeb show that too 
strong solutions of potassium cyanide, and too long exposure to weak 
solutions, soon kill the egg. From this the reasonable interpretation 
is, that the potassium cyanide is a poison for all living matter, but it 
acts more quickly on bacteria than on sea-urchin eggs; it is in no 
sense a prolonger of life. 

5. From the fact that unfertilized eggs can be kept in sterile sea- 
water for eleven days or longer, it would seem that the specific mortal 
processes of Loeb are as yet hypothetical phenomena without any defi- 
nite experimental basis. 


We are indebted to the United States Fish Commission for ex- 
tending the courtesies of their Scientific Laboratories at Woods Hole. 


NOTES ON THE “PROTAGON” OF THE BRAIN: 
By W. W. LESEM anp WILLIAM J. GIES. 


aden years ago Chittenden and Frissell* made a study of 
the distribution of phosphorus-containing substances in the 
brain. The results obtained by them seemed to ‘indicate that 
protagon contains but a small proportion of the total phosphorus of 
the brain and that other phosphorized organic bodies, such as 
lecithins, are present, preformed in the tissue, in relatively large 
proportion.” They concluded that ‘ the dry solid matter of the brain 
contains as much or even more lecithin than protagon.” Chittenden 
and Frissell also observed that, “contrary to previous statements, 
protagon tends to undergo cleavage by long-continued heating at 
45° C. in 85 per cent alcohol, a certain amount of an alcohol-soluble 
(at o° C.) body richer in phosphorus than protagon, being split off 
while the residual protagon obtained by recrystallization at o°C. con- 
tained a somewhat diminished percentage of phosphorus.” 

Shortly after the publication of the brief note containing the above 
deductions, Dr. Gies repeated and extended the experiments begun 
by Dr. Frissell. The general conclusions of this second series of 
experiments were practically the same as those previously re- 
ported, but as the work was unavoidably interrupted, no further 
reference was made to them. Recently, however, new experiments 
on protagon have been performed by Mr. Lesem and Dr. Gies. The 
results of these experiments, to which we shall refer farther on, make 
it seem desirable to give here some of the related data of the earlier 
experiments in which the work of Chittenden and Frissell was 
repeated. 


1 This work was begun by Dr. Gres under Professor CHITTENDEN’S super- 
vision, in the Sheffield Laboratory of Physiological Chemistry at Yale University. 
It was completed by Mr. LeEsem and Dr. Gies in the Laboratory of Physiological 
Chemistry at Columbia University. 

2 CHITTENDEN : Proceedings of the American Physiological Society, Science, 
107; v- (IN. S.), p- gor. 


183 


184 W. W. Lesem and William J. Gres. 


I On THE GENERAL DISTRIBUTION OF PHOSPHORUS-CONTAINING 
SUBSTANCES IN THE BRAIN. 


The brains employed in the experiments by Chittenden and 
Frissell were taken from sheep. Although the brains were used 
within twenty-four hours after the death of the animals, it seemed 
possible that, even within that short period, bacterial changes might 
have had some influence on the results! In repeating the first 
series of experiments, this difficulty was obviated by the adoption of 
the following procedure, which is the same as that used by Chittenden 
and Frissell,? except in the steps taken at the beginning to prevent 
possible alterations through the influence of bacteria. 


First experiment. — In this experiment glass-stoppered bottles of convenient 
size, containing about 750 c.c. of 85 per cent alcohol, were accurately 
weighed and removed to the slaughter house without loss of fluid. The 
sheep were killed in the usual way. The greater portion of blood dis- 
appeared from the brain in a minute or two, when the head was opened 
with a cleaver and the entire brain quickly removed. Superficial blood 
and lymph were taken off promptly with a clean dry cloth. While the 
brains were still at practically the normal body temperature, they were 
rapidly slashed with a scalpel and at once transferred to the bottled 
alcohol. ‘Two whole brains were deposited in each of three bottles. 
Special care was taken to prevent any loss of alcohol by evaporation or by 
spilling. 

It would seem that this prompt treatment with alcohol prevented such 
post-mortem changes as exposure for several hours to the air, a lowered 
temperature, etc., might induce. We do not mean to suggest, however, 
that the alcohol itself has no transforming power on the phosphorized 
constituents. Such influence, if exerted, would doubtless have been no 
greater, nor any different, at this point than later on. 

The quantities of tissue in each bottle were 152.99, 172.19, and 
148.89 gms. 

Preliminary cold extracts. —The tissue remained in the original alcohol 
about four hours, when the filtrate was collected and the tissue very 
thoroughly macerated in a mortar. The finely divided material was next 
transferred to 750 c.c. of 85 per cent alcohol, and kept under it over 


' The results of the following experiments show, however, that no appreciable 
changes of such character could have been effected. 

* The methods employed by CHITTENDEN and FRIsSSsELL could not be described 
in the very brief abstract of the preliminary report of their work. For that reason 
we give the methods here in some detail. 


Notes on the “ Protagon” of the Brain. 185 


night, after which the filtrate was again separated. These two cold 
extracts were combined. 

Extracts at 45° C.—Extraction was next made in 85 per cent alcohol (14 litres 
for each pair of brains) for ten hours at 45° C., and the filtrate again 
collected. After standing in 2 litres of 85 per cent alcohol, at room 
temperature over night, the alcohol-tissue mixture was warmed to 45° C. 
and held at that temperature for twelve hours, after which the filtrate was 
again obtained. ‘The residual tissue was once more kept in 2 litres of 
‘85 per cent alcohol over night and further extracted in the same fluid at 
45° C. for fourteen hours, when the filtrate was preserved as before. 
After each of these filtrations, the solid substance was washed with a little 
warm alcohol (85 per cent), and the washings added to the appropriate 
filtrate. 

Extraction in boiling alcohol. — At this point the tissue remained in 1 litre of 
85 per cent alcohol over night, when the mixture was boiled on a water 
bath fora half hour. After filtering, the tissue was also extracted in boiling 
95 per cent alcohol for the same length of time. These two hot alcoholic 
extracts were combined. 

Tissue residue. — The residual tissue was finally washed with cold 95 per cent 
alcohol, then with absolute alcohol, and dried to constant weight at 80° C. 

Treatment of the extracts. —The extracts obtained at room temperature and 
in boiling alcohol: were separately evaporated in silver crucibles almost to 
dryness, and the total phosphorus content determined directly. The 
cold extract of our first preparation, however, was separated into protagon 
and filtrate therefrom by the method referred to below. 

The three extracts obtained at 45° C., in the second and third pre- 
parations, were separately reduced to o° C. with the aid of common freez- 
ing mixture, and held at that point for six hours. A heavy floccuient 
precipitate containing much crystalline cholesterin, protagon, etc., quickly 
separated from the first of each series of three extracts. The precipitate 
was considerably less in the second extract, and only a very faint turbidity 
was formed in the third. Each precipitate was quickly filtered, at a 
temperature slightly below o° C., on funnels surrounded by freezing mix- 
ture. The precipitates were washed once with cold 85 per cent alcohol, 
and then with cold ether until free from cholesterin. ‘The alcohol wash- 
ings were added to the same filtrates. The filtrates were combined and 
evaporated for the determination of phosphorus. The ether washings 
were given the same treatment. The protagon products were dried at a 
low temperature on the filter papers. Phosphorus was determined in the 
mixture of protagon and filter papers, the latter having been free from 
that element. 

Phosphorus was always determined by the usual fusion method. 


186 W. W. Lesem and William /. Gtes. 


Analytic results. — The following table gives our analytic results for 
phosphorus in the various solids and fluids separated by the above 


’ method: 
LABICE I; 


Phosphorus content. 


Ile lM | III. DEST LE IT. | Se 


[°xtracts, etc. | | 


Percentage | Percentage 
Grams. of total | of total 


solid matter.| phosphorus. 
| 
A Goldiextractsuta)iite sur 0.1423 0.1348 0.1841 | 0.35 | 0.43 | 2632 | 32.33 
@RrOtar OMe me 0.0432 
b. Viltrate ae prota, exeyh) 0.0991 
By WtractsvatetseaGs (Oi menan eee 0.2599! 0.2887 | 0.68 | 0.67 | 51.14 | 50.38 
@ Protagonsaeene- Sievers 0.0874; 0.1008| 0.23 | 0.23 | 17.30 17.380 
b. Filtrates from prota: | 
SOUSa a ra Pete 0.1370 0.1401 | 0.36 | 0.33 | 27.07 | 24.81 
. Etherw ashings SE pr O- 
tagons . Stee 0.0355 0.0478 | 0.09 | 0.11 6.77 §.27 
G Extracts’ m boiling oes 
roll a eae fine ee eee shete 0.0047 0.0054 | 0.01 | 0.01 0.75 0.75 


Dy Vissue residue 0.) ee ior 0.1098} 0.0939} 0.29 | 0.22 | 21.80 | 16.54 


Total phosphorus 0.5092} 0.5721 | 1.33 elESS 


Weight of fresh tissue . .| 148.89 152.99 172.19 
Weight of tissue residue. . eStats 15325 17.08 


Estimated solids in fresh tis- 


sue (25 per cent). . ees £55 43.05 
Estimated weight of extr acted 
MAthGr 4:54 ic) eee ees eon eeee 23.00 25.97 


That the preliminary cold extracts contained a comparatively small 
amount of protagon seems to be indicated by the results for our first 
preparation. Protagon is only slightly soluble in 85 per cent alcohol 
ato C., and is practically insoluble in ether at the same temperature. 
Thus of 2 grams of protagon, 0.03 to 0.04 gram dissolved in 500 
c.c. of 85 per cent alcohol at 0° C. The same quantity of ethereal 
filtrate from 3.6 grams of protagon, at the same temperature, con- 
tained nothing yielding a phosphorus reaction after fusion with alkali, 
It is possible that the presence of the other constituents of the 
alcoholic extract may increase or decrease this solubility. It is 
hardly probable, however, that more than an insignificant portion 


Notes on the “ Protagon” of the Brain. 187 


of the protagon remains unprecipitated on lowering to zero the 
temperature of alcoholic extracts such as the above. 

Second experiment. — We decided to repeat the experiment again, 
but with less tissue. The results of our previous experiment had 
been obtained for the whole brain. We now endeavored to ascertain 
whether the above data apply equally to all portions of the brain or 
whether there are wide phosphorus variations for the parts. This was 
accomplished indirectly without materially altering the conditions of 
the previous experiment. For the purpose indicated we took amounts 
of tissue equivalent in weight to a whole brain, but made up of 
different parts of two brains. 


The method of treatment at the slaughter house, transportation in weighed 
alcohol, extraction in 85 per cent alcohol at room temperature, at 45° C., 
etc., separation of protagon, etc., were the same in this as in the first 
experiment. Samples of the fresh tissue were used for determinations of 
solids and phosphorus. 

At the slaughter house the brains were carefully sectioned transversely into 
halves just before their deposition in the alcohol. ‘The halves were 
combined as indicated in the next table. The preliminary extracts in 
cold alcohol were united with those obtained at 45° C., and the prota- 
gon was removed from the mixture. Four extractions of each sample 
of tissue were made at 45° C. One litre of 85 per cent alcohol per 
brain was used each time. The washing of the protagons with ether was 
omitted. ; 


Table II, on page 188, gives the essential results of this experiment. 

Only insignificant differences are to be observed between the 
results of the first two experiments. The analytic data are, therefore, 
essentially the same for the anterior and posterior halves of the brain. 
The similarity of the results of this series to those of the preceding 
is especially evident from the directly comparable data given in 
Table III on page 188. 

The results of the first and second experiments show that the 
greater portion of the phosphorus of the brain is contained in sub- 
stances not precipitable as protagon. The bulk of the phosphorus 
in the preliminary cold extract (Exp. 1), and in the filtrates from the 
protagons (Exps. 1 and 2), is doubtless contained in substances as 
readily soluble in alcohol as lecithin. Some phosphate was also 
present. Probably most of the phosphorus of the ether washings 
(Exp. 1) was contained in substance which was soluble in the 


188 W. W. Lesem and William /. Gres. 


alcohol (and in the ether), but which adhered to the precipitate 
until it was treated with ether. 


TABLE Ir 


Phosphorus content. 


Extracts, ete. A. B. G. Py: 
| Ant. half of 1. | Ant. half of 2. | Ant. half of 3. Post. half of 4. 
| Post. half of 2.| Post. half of 3.| Ant. half of 4. | Post. half of 5. 
Grams. Grams. Grams. Grams. 
I. Ext. at room temp. | 
and at 45o°1Gx 3 0.2512 0.2152 0.2446 0.2353 
a. Protagons (4) . | 0.0953 0.0728 0.0900 0.0869 
6. Filtrates from | 
protagons. . 0.1559 0.1424 0.1546 0.1484 
Il, Extracts in boiling | 
alcoholi.) = 2 0.0013 0.0013 0.0015 0.0016 
III. Tissue residue . .| 0.0635 0.0432 0.0517 0.0467 
Total phosphorus: 
a. Totalinall parts | 0.3160 0.2597 0.2978 0.2836 
6. As determined | 
directly. . . aeias 0.2694 schsve | 0.3038 
Weight of fresh tissue . $4.86 72.07 86.44 82.23 
| | 


Whether these soluble substances exist ‘‘ preformed” in the brain, 
as Chittenden and Frissell and others believe, or are decomposition 
products resulting from the use of the reagents, as some infer, is not 
made clear by these experiments. The former view seems more 
probable. 


DABEE Tit: 


' Phosphorus content. 

Weight | 
of fresh 
tissue. Prota- Filtrate. I lot Tissue 

gon. extract. | residue. 

Gms. | Gm. Gm. | Gm. Gm. Gm. 


Brain. 
Total. 


First One half of III. | 86.10 0.07431 0.16217 0.0027 0.0469 0 2860 


Second GC. 86.44 0.0900 0.1546 | 0.0015 0.0517 0.2978 


! Including ether washings. 2 Including cold extract. 


Notes on the ** Protagon” of the Brain. 189 


The results of the next experiment lead to essentially the same 
conclusions as those drawn from the preceding. 


Third experiment. -— The methods of this experiment were, in general, the 
same as those of the first and second. 
treatment are to be noted. 


The following differences of 
The divisions of the brains were made longi- 
tudinally instead of transversely. The alcoholic filtrates (2), obtained at 
o° C. after separation of the protagon, were evaporated almost to dryness 
on a water bath at 35—40° C. ‘The residues thus resulting were thoroughly 
extracted several times with a moderate excess of cold ether. The 
The residue left after 
So 
little seemed to dissolve that the alcoholic extracts were evaporated with 
the ethereal. 


mostly inorganic matter, was next treated with water. 


extracts were filtered and evaporated to dryness. 
treatment with ether was extracted with boiling 95 per cent alcohol. 


The substance remaining after the extraction with alcohol, 
All of it dissolved 
very readily. This solution was then evaporated to dryness. Phosphorus 
was determined in the substance from each of these extracts and in the 


protagon, with the results tabulated below : 


TABLE IV. 


Phosphorus content. 


A B 


Extracts, etc. 


| Same lateral 


halves of 


| brains 1 and 2. 


Opposite lat- 


eral halves of | 


brains | and 3. 


& 
Opposite lat- 
eral halves of 
| brains 4and 5. 


Grams. Grams. Grams. 


I. Protagons (2) 0.0576 0.0701 0.0667 


II. Filtrates 0.1841 0.2037 


a. Substance soluble in alcohol 
and ether . Ee teen ee) Bee 

6. Residual substance soluble in 
water 


0.1603 0.1750 


0.0287 


Total phosphorus 


Weight of fresh tissue . 


II. ON THE QUESTION OF THE CHEMICAL INDIVIDUALITY OF 
PROTAGON. 


Twenty years ago Gamgee expressed himself on this subject as 
follows : “ There is no subject in physiological chemistry concerning 
which it is more difficult to give a statement, which would be accepted 


190 W. W. Lesem and Witham J. Gres. 


as correct by those who have devoted their attention to it, than the 
chemistry of the complex phosphorized fats which exist in the 
nervous tissue.”! The same may be said perhaps with equal force 
to-day, in spite of the careful work done in the mean time to solve 
the problems connected with the chemical constituents of the brain. 

Soon after Liebreich? separated from the brain the substance he 
called protagon, Thudichum® and others denied the existence of 
such a substance. Thus, Diaconow,? working as did Liebreich, in 
Hoppe-Seyler’s laboratory, obtained results which led him to conclude 
that protagon is a mixture of lecithin and cerebrin. The later re- 
searches of Gamgee and Blankenhorn,’ however, furnished data 
which were generally accepted as amply confirming the original con- 
clusions of Liebreich. The subsequent work of Baumstark,°® Kossel 
and Freytag,’ and Ruppel,® particularly, further emphasized the 
growing confidence in the existence and importance of protagon 
as a brain constituent. Until recently the matter seemed to be 
settled in the general conviction that protagon is a chemical individ- 
ual, in spite of Thudichum’s claims to the contrary. As late as 1899 
Hammarsten ® indicated, as follows, the prevalent feeling toward the 
non-concurrent conclusions in which Thudichum has persisted : 
“Thudichum claims to have isolated from the brain a number of 
phosphorus-containing substances which he divides into three main 
groups: kephalins, myelins, and lecithins. Thus far, however, his 
results have not been confirmed by any other investigators.” 

The work of Kossel and Freytag may be regarded as an approach 
to Thudichum’s position with reference to the composite nature of 
protagon. Kossel and Freytag discovered that protagon contains 
sulphur. Variations among their several products, in spite of great 
care in preparation, also led them to believe in the existence of 
several protagons. Further than this, they found that protagons 


' GAMGEE: A text-book of the physiological chemistry of the animal body, 1880, 
Ty p> Abe 

* LIEBREICH: Annalen der Chemie und Pharmacie, 1865, cxxxiv, p. 29. 

* THUDICHUM: Chemisches Centralblatt, 1875, p. 408. 

* DIAcONOW: Centralblatt fiir die medicinischen Wissenschaften, 1868, p. 97- 


® GAMGEE UND BLANKENHORN: Zeitschrift fiir physiologische Chemie, 1879, 
ili, p. 260. 

° BAUMSTARK: J/did, 1885, ix, p. 145. 

7 KOSSEL UND FREYTAG: J/6/d, 1893, xvii, p. 431. 

* Rupprec: Zeitschrift fiir Biologie, 1895, xxxi, p. 86. 


* HAMMARSTEN: Lehrbuch der physiologischen Chemie, 1899, p. 366. 


Notes on the “ Protagon” of the Brain. IgI 


readily yield several substances similar to or identical with some 
described by Thudichum,! and which he still contends are among 
the fourteen (!) different bodies contained in the protagon mixture. 
The subsequent work of Chittenden and Frissell also gave indications 
of facts in harmony with the earliest results of Diaconow and his 
view that protagon is a mixture. Lately, Wérner and Thierfelder * 
attacked the problem by improved methods, and obtained results 
which seem to show that protagon is not an individual substance, or 
else that it is a remarkably labile body, physically and chemically. 

Below we give the results of our repetitions of the experiments of 
Chittenden and Frissell bearing on the matter in question. 

Fourth experiment.—A sample of protagon which had been prepared 
by Dr. Frissell from sheep brains by the usual method — precipita- 
tion from warm alcoholic extract at o° C. and thorough washing in 
ether at 0° C.— was placed at our disposal for this experiment. 


We further purified the protagon by recrystallizing it once from alcohol. 25 gms. 
of the product was kept in 1500 c.c. of 85 per cent alcohol at 40°C. for 
twelve hours and the mixture repeatedly stirred. At the end of that 
time only about half of the substance had dissolved. 

First product and filtrate. —The mixture was filtered and the protagon sep- 
arated from the extract by the usual cooling process, etc. ‘The filtrate 
from the protagon was evaporated to dryness. 

Second product and filtrate. — That portion of the original protagon which 
remained undissolved was again subjected to treatment in the same 
amount of alcohol. Most of the substance dissolved at the end of twelve 
hours. The second portions of protagon and evaporated filtrates were 
obtained as before from the filtered extract. 

Third product and filtrate. —The protagon still remaining undissolved after 
the second extraction with alcohol was again placed in the same amount 
of warm alcohol for a similar period. Protagon was separated from the 
extract and the filtrate from it evaporated to dryness as before.’ 

Insoluble portion. — A fairly large proportion of the original protagon re- 
mained insoluble under these conditions. 

Alcohol-ether washings. — Each successive residual portion of protagon re- 
ferred to above was washed with warm alcohol and the washings added to 


1 THuDICHUM: Die chemische Konstitution des Gehirns des Menschen und 
der Tiere, 1901, pp. 54-57; 328. 

2 WORNER UND THIERFELDER: Zeitschrift fiir physiologische Chemie, Igoo, 
XXX, Pp. 54”. 

3 The crystalline appearance of these various protagon products was practically 
the same. 


192 W. W. Lesem and William /. Gres. 


the filtrates. All of the samples of freshly precipitated protagon were 
washed first with a small quantity of cold 85 per cent alcohol and later 
with moderate excess of cold ether. The alcoholic and ethereal wash- 
ings of the freshly precipitated protagon were combined and evaporated. 

Treatment of the products. — Yhe portions of protagon, and the substance in 
the filtrates and washings, were carefully determined quantitatively. Phos- 
phorus. was also estimated in each by the usual fusion method. 


The following summary gives our data in this connection : — 
TABLE V. 


Weight in Percentage of 


Protagon, etc. 
ey grams,! phosphorus. 


4. Freshly precipitated protagon : 
@.-Hrompurst extracties 0 ae 10.834 
6. From second extract .. . 7.599 
é. Exonr thind extract) =... 1.729 (20.162) 


. Insoluble protagon (residue) . . 2.009 


. Substance in filtrates from the 
freshly precipitated protagon: 
a. Of first extract . F 
6. Of second extract . 
c. Of third extract 


. Alcohol-ether washings of the 
freshly precipitated products 


Total substance recovered 


Total substance taken . 


1 The weights are for substance dried in vacuo over H,SOy, to constant weight. 


Fifth experiment. — We repeated the preceding experiment with two freshly 
prepared samples of protagon made by us from two different quantities of 
sheep brains. ‘These samples of protagon were prepared by the usual 
method and were twice recrystallized. ‘lwelve gms. of each was used. 
‘Two treatments were made with 1} litres of 85 per cent alcohol at 45° C., 
etc., as in the fourth experiment, with the results tabulated on page 193: 


Among the points to be noted in Tables V and VI is the decreas- 
ing percentage content of phosphorus in each successive protagon 
and in the final insoluble residue. Also, the unusually high though 
diminishing proportion of phosphorus in the substance of the filtrates 
obtained each time protagon was separated at o° C. 


a 


Notes on the “ Protagon” of the Brain. 193 


Our method of fractional separation was that customarily employed 
in the purification of protagon. Here it was merely repeated more 
frequently than usual. Instead of obtaining purer protagons in the 
process, however, it appears that, with each successive precipitation, 
the substance itself changed in composition and, also, that variously 
composed products were liberated into the filtrates from the prota- 
gons at the same time. The final residue was wax-like and quite 
different from the snow-white protagon of the first extracts. We 
are certain that our products were “ pure” at the start. 


TABLE VI. 


Protagon, etc. oe — 
ae Weight Percentage Weight -| Percentage 


in of phos- | in of phos- 
grams. phorus. | grams. phorus. 


A. Freshly precipitated protagon:! 
a. From first extract . . .| 5.945 L21 |i'3:659 
2.009 (5.668) 


6. From second extract . .| 2.680 (8.625) 1.01 


B. Insoluble protagon (residue) . 0.655 0.91 3.892 


C. Substance in filtrates from the 
freshly precipitated pro- 
tagon: 

a Of firstiextract. «. . % <1 1-613 ‘ 1} 1.321 
6. Of second extract . . .| 0.983 (2.596) : | 0.981 (2.302) 


Total substance recovered? . 11.876 eee 11.862 


Total substance taken. . . 12.150 1.26 12.150 


I 


1 The precipitates were washed only with cold alcohol. 


> 


2 See note | in the preceding table. 


The data of the last two experiments are in close agreement with 
the similar facts found by Chittenden and Frissell. They are in 
harmony with corresponding data recently published by Thudichum.! 

These results were obtained by applying the usual purification 
method. They show, we think, that protagon is either a mixture of 
bodies, or else a substance decomposing quite readily under the 
conditions of such experiments. If the latter conclusion appears to 


1 THUDICHUM: Die chemische Konstitution des Gehirns des Menschen und 
der Tiere, 1901, pp. 84-85. 


194 W. W. Lesem and Witham J. Gres. 


be more probable than the former, it must then be admitted that 
thus far no standard of purity for protagon has been raised which 
is not open to the objection that it is based on methods involving 
unavoidable decomposition. 

Elementary composition of protagon.—It seemed desirable at this 
point to ascertain the general elementary composition of several of 
the protagon products prepared in the preceding experiments. The 
summary below gives our results for four representative preparations : 


(AGE eB Valse 


Percentage composition of protagons.1 


Fourth experiment. Fifth experiment. 


Elements. 


ad. 


65.98 | 66.24 66.11 | 66.63 | 66.46 | 66.55 || 65.87 | 65.77 | 65.82 


H | 10.83 10.97 10.90 |10.72 10.60 10.66) 10.73 | 10.47 | 10.60 
2.09| 1.95| 2.02] 2.22, 216| 219]| 1.97| 1.99] 1.98]| 
128) eee lee .25| 1.26| 1.26 


OT || seize ON sen \eOlenay 


| 


19.97|| .. | .. |18: .. (se tee7l SP a 


1 The methods of analysis employed were those already described by us: HAWK 
and Girs: This journal, 1901, v, p. 403. 

2 The amount of ash varied between 2 and 3 per cent. It consisted very largely 
of phosphate derived during the incineration process. 


The results for elementary composition are in fairly close accord 
with those of previous observers.! Since all of our samples were 
made by practically the same method as that employed in most of 
the earlier investigations, however, this harmony proves nothing 
more than that the materials analyzed by all of us were of essentially 
the same character. The minor variations suggest that the products 
may be fairly uniform mixtures, but Kossel and Freytag’s conclusion 
that several protagons exist might also be drawn from them. In fact, 
much to our surprise, these results accord as well as many analytic 


1 See the summary lately given by Nov: Zeitschrift fiir physiologische 


Chemie, 1899, xxvii, p. 376. 


Notes on the “ Protagon” of the Brain. 195 


series given for what are undoubtedly individual substances. Our 
data in this connection, considered by themselves, would seem to 
harmonize with the older view of the integrity of protagon. In the 
light of our other results, however, they illustrate the fact that uni- 
formity in composition frequently hides chemical differences. In 
this case general uniformity seems to give no assurance of chemical 
individuality. 

Application of the methods of Wo6rner and Thierfelder. —-\Ve have 
repeated some of the recent preliminary experiments of Worner and 
Thierfelder without, however, anticipating any of the steps which it 
may be the intention of these investigators to take in furtherance of 
their work. 

Worner and Thierfelder used material from human brains. We 
used purified protagon from sheep brains. The agreement between 
their results and ours is, therefore, all the more significant. Our data 
in this connection will be given only briefly. 

We made use of freshly prepared protagon, as well as some of the 
preparations already referred to. Our protagon products dissolved 
almost entirely in moderate quantities of solutions of equal parts of 
alcohol and chloroform, or alcohol and benzol, at 45° C. The latter 
solution appeared to exert solvent action less rapidly than the other. 
The crystals obtained from such fluids, after gradual evaporation at 
40-45° C., varied somewhat with changes in the composition of the 
solvent and in the concentration of the solution. 

The residue left behind at this point, on treatment of the protagon 
with a moderate quantity of the solution, resembled that remaining 
in Experiments 4 and 5 preceding. It consisted of globular forms 
and amorphous substance. On cooling the filtrate from the melted 
matter, a bulky precipitate of snow-white ‘“‘cerebron” spheres was 
deposited. The filtrate from the cerebron, on evaporation, yielded 
microscopic needles. The filtrate from these crystals contained other 
organic matter which, however, furnished only a slight amount of 
crystalline substance on further evaporation or on longer standing. 
These experiments were repeated several times with similar outcome. 

Of these various products the cerebron was the only one we 
attempted to separate in any quantity for further examination. In 
all the ordinary tests tried on the several preparations of purified 
cerebron, we found that our products gave the reactions already 
attributed to the substance by Worner and Thierfelder. All the 
crystals figured for it by these investigators were observed in the 


196 W. W. Lesem and William J. Gres. 


various fluids. The typical transformation of the cerebron balls in 
85 per cent alcohol at 50° C. into needles, minute plates, etc., was 
also brought about several times. We were unable to make any 
elementary analyses of the cerebron, but verified the statement that 
on decomposition with acid a reducing substance may be detected 
among its cleavage products. 

In view of these results, also, it appears necessary to conclude 
that protagon is not merely an unstable substance, but a mixture of 
bodies.’ It is not at all likely that these various products arise by 
decomposition from such mild treatment. Further study of cerebron 
and its related products, also of the new substance very recently 
isolated by Ulpiani and Lelli,? and called by them, ‘“ paranukleo- 
protagon,” may throw more light on the protagon question. 


III. SUMMARY OF GENERAL CONCLUSIONS. 


(1) The protagon of the brain is a mixture of substances, not a 
chemical individual. 

(2) The mixture called protagon does not contain the bulk of the 
phosphorized organic substance of the brain. ; 


' See very recent paper by Kocu: Zeitschrift fiir physiologische Chemie, 1902, 
XXXVI, Pp. 140. 
2 ULPIANI UND LELLI: Chemisches Centralblatt, 1902, ii, p. 292. 


ena THE GROWTH. OF SUCKLING PIGS FED ON A DIET 
OF SKIMMED COW’S MILK. 


By MARGARET B. WILSON. 


[From the Physiological Laboratory of the University and Bellevue Hospital Medical 
College. | 


THE OLDER EXPERIMENTS. 


BOUT ten years ago at the Yale Medical School Dr. L. C. 
Sanford and Professor Graham Lusk reared three new-born 
pigs of the same litter on centrifugalized skimmed cow’s milk. To 
the milk of one pig 2 per cent of milk-sugar was added, to that of 
a second 2 per cent of dextrose, while a third pig received skim-milk 
alone. All the pigs had diarrhcea at one time or another, but this 
did not last long. The experiments proved that the pigs could thrive 
on a skim-milk diet and that the addition of sugars was advantageous. 
The milk was analyzed. The growth of the pigs in proteid substance 
was determined by analyzing a new-born pig of the same litter which 
was killed at birth. The percentage proteid composition was taken to 
represent that of the living pigs at the start of the feeding. Any 
gain of proteid noticed in the analysis of the pigs after the fourteen 
days of feeding could be attributed to growth. 
Another pig of a different litter was fed for fourteen days on whole 
milk of unknown composition. 
For the calorific value of this milk Rubner’s! average value has 
been assumed, —1 litre =650 calories. 
The composition of the diet of the other pigs was determined by 
analysis, as follows (see first table, page 198) : 
The calculation of the energy in the milks described in the table 
was made according to Rubner’s standard values? for milk, as follows: 


Proteid ... . . . 4.4 calories 
Raton, 2 ete 2 soe 2a 9:2 calories 
Milk-sugar . . . . . 3.9 calories 


1 LEYDEN’sS Handbuch, 1897, Bd. 1, p. 113. 
2 RUBNER: Zeitschrift fiir Biologie, 1901, xliv, p. 280: also RUBNER und 
HEUBNER: Zeitschrift fiir Biologie, 1898, xxxvi, p. 55. 
197 


198 Margaret B. Wilson. 


from which is deducted 5.4 per cent lost through the feces. Rubner 
has shown that these theoretical calculations very nearly cover the 
fuel value actually available physiologically. 


Skim. Lactose. Dextrose. 


Gms. |Calories., Gms. |Calories.. Gms. | Calories 


Proteiduy) ay be ak eee ZNE | pe 94S-Orae 2OleD 1238.6" | 302.1) |is2or 


4+ 
ee ees a al) Sona ||) Seis: 48.0 441.6 ops) 473.8 


Milk-sugat.) i) te ee aeealeo OSes 1192.6 | 581.7 + 2268.6 | 435:0 | 16965 


Dextrose 


Total 1255. Sabi. vere See Pee weave 3948.8 a ee 4199.1 

Less 5.4 per cent lost | 

through feces . . . state Sire Seay Pe corer PANES y? sorte 226.7 
| | 


Calories available physi- 
ologically . ane 


Dextrose was calculated: 1 gm. = 3692 cal. 

There can be no significance in the above figures unless the 
skimmed cow’s milk is normally absorbed. To show this the faeces 
were collected, and the intestinal contents of the pig, killed twelve 
hours after his last meal, were mixed with the faeces for analysis. 

The results may thus be indicated : 


Skim. Lactose. Dextrose. 
Grams. Grams. Grams. 


Nin milk:fed/=%. 5-7 se a eee 33.68 44.19 47.42 
iin feeces*;. a en ea ee ee SS 1.47 1.96 


N excreted in feces in percent . . BY 3:3 4.1 


Fat in milk fed 
Fat in faces 


Fat excreted in faeces in per cent . 


0 Ol Ee 


Suckling Pigs on a Diet of Skimmed Cow's Milk. 199 


Camerer! in seven experiments on children of different ages finds 
that there is a faecal excretion of 5.8 per cent of the nitrogen in a milk 
diet, and 4.7 per cent of the fat. Rubner? finds an average of 7.1 per 
cent of the nitrogen in the ingesta and 5.3 per cent of the fat, in the 
faeces of the adult on a milk diet. These figures show that the milk 
absorption in the pigs was similar to that in children. No milk-sugar 
was ever found in the urine. That the physiological utilization of the 
calories of milk is almost the same under different conditions has been 
shown by Rubner? in the following table: 


Physiologically available calories of mother’s milk. . . . . . . . . 91.6 percent 
Physiologically available calories of cows milk . . . . . . . . . . 90.7 per cent 
Physiologically available calories of cow’s milk with milk-sugar . . . . 92.2 per cent 


Physiologically available calories of cow’s milk when ingested by an adult $9.8 per cent 


The figures for the faecal contents indicate a normal absorption, 
and consequently a normal utilization of the milk by the pigs. 

All the pigs developed rapidly except the pig fed on skimmed milk, 
which seemed feeble throughout the experiment, and did not suckle 
vigorously as the others did. The following table briefly includes the 
fourteen days’ life-history of the pigs. 


Whole 


3 Skim. Lactose. | Dextrose. 
milk. 


Weight of pig when born. . . . . 550 


Weight of pig when killed . .. . | 1008 390 2000 


(Crowtbeoorams;. . 46, . . » « 458 5 | 38 | S48 
KCcomynhnnipen Gent <0) ea ay fn 82.1 
Minestechiniee, v2 hie es ew 3059 

Available caloriesfed . .... . [1989?] 


Growth in grams per litre of milk. . 149 


Growth in grams per 1000 calories fed [230?] 


i CAMERER: Zeitschrift fiir Biologie, 1881, xvii, p. 493. 
RuBNER: Zeitschrift fiir Biologie, 1879, xv, p. 115. 
3 RUBNER: Zeitschrift fiir Biologie, 1899, xxxviii, p. 380. 


200 Margaret B. Wetlson. 


The discussion of these results will be deferred until later, but it 
may be observed in passing that, except in the case of the badly nour- 
ished pig fed on skimmed milk, the growth seems proportional to the 
calorific value of the food. 

Concerning the proteid growth, it was first determined that a new- 
born pig of the same litter weighing 1137 grams, contained of his live 
weight 1.78 per cent of N or 11.17 percent of proteid. If this be the 
true value at the birth of the other pigs, we can estimate the nitrogen 
contained in them at birth and deduct this from the actual nitrogen 
found at the end of the experimental feeding. 

From this the growth of proteid matter may be calculated. 


Skim. | Lactose: Dextrose. 
. | . | rr? 
N in grams. | N in grams. | N in grams. 


Wihten Igileils Nrs,0 5) at peah Metie aek Gre 25.60 | 35.46 43.90 
\a over bYovany (egonenenwel)) 3 5 6 5 5 6 6 L780) er7Z 20.51 


Growthynsir) 2) Pee eee eee 7.80 | 16.74 23.39 


Growth in per cent’. 2-4 7 we eee ee 44.00 | §9.00°- 114.00 
Nim fond! et 3368 | 44.19 47.42 
IN; metabolize de leat ce amigere ee a Me ee ern i. 4: 24.03 


Per cent of N of food used to form new tissue : : 48.00 


From the above figures it is seen that the retention of proteid for 
tissue building amounted to 23, 38, and 48 per cent in the three 
different pigs. 

The method employed by Sanford and Lusk in these older experi- 
ments seemed to be capable of further extension, and at Professor 
Lusk’s suggestion, and with his assistance, a new research was 
undertaken. 


THE New EXPERIMENTS. 


Work has been done of late by Camerer, Rubner, Oppenheimer, 
and others, which has added much of high importance to the knowl- 
edge of the laws of nutrition in the growing child. In such exper- 
iments on children the diet must naturally be unexceptionable. It 
seemed that these experiments might properly be supplemented by 
investigating the nutritive value of skim-milk to the growing organ- 


Suckling Pigs on a Diet of Skimmed Cow's Milk. 201 


ism. So great is the popular prejudice against skim-milk, that it has 
been made a crime to sell it in Néw York.! 

For these feeding experiments, pigs were selected on account of 
their rapid growth, their vigorous metabolism, and the fact that they, 
like man, are omnivorous. Their great muscular activity is the one 
decided cause of difference between their metabolism and that of a 
helpless infant. 

Six new-born pigs of the same litter were obtained. Three of 
them were at once submitted to analysis, and three were reared on 
skim-milk. Of these three, one, the skim-milk pig, was fed on skim- 
milk alone; another, the lactose pig, received the same skim-milk to 
which three per cent of lactose had been added; and a third pig, the 
dextrose pig, was nourished with the same skim-milk containing three 
per cent of added dextrose. After sixteen days the pigs were killed 
and submitted to analysis. 

At first glance it may be said that the experiments are far too few 
in number to prove anything, and yet the general uniformity of the 
results obtained, and their accordance with the observations made on 
children, render this objection less valid. 

About forty litres of skimmed cow’s milk were obtained in winter, 
fresh from the Briarcliff Farm, which establishment also furnished 
the pigs. I desire to express my thanks to the company and its em- 
ployees for their trouble in the matter. About one third of the milk 
was treated with 30 grams of lactose hydrate per litre, and another 
third with 30 grams of dextrose per litre. The milk was measured, 
200 c.c. at one time (using Cremer’s two-way stop-cock pipette), into 
ordinary nursing bottles, which were stoppered with cotton and frozen 
in the ice machine of the anatomical department. From time to time 
the bottles were thawed, as they were needed, and warmed for use. 
The milk remained perfectly sweet throughout the experiment. It 
agreed perfectly with the pigs, and there was no diarrhcea in any of 
them at any time. 

The nitrogen of the milk was ascertained, by making Kjeldahl 
determinations: the fat, by weighing the ether extract of the milk 
dried on fat-free paper; and the milk-sugar, by the Allihn method, 
after precipitating the proteid. Calcium was determined after incinera- 
tion of the total solids. The potential energy was calculated as before, 


1 Valuable information on the nutritive value of the skim-milk may be found in: 
Milk as Food, U. S. Dept. of Agriculture, 1898, Farmers’ Bulletin, No. 74. 


202 Margaret B. Wilson. 


although a loss through the faeces of 5.4 per cent of the energy of the 
added sugars is hardly possible. Rubner has shown us, as before 
described, that there is a higher average utilization of the calorific 
power in milk after the addition of lactose. 


1000 c.c. skim-milk. 


Grams. | Calories. 


Proteid 
Hates 
Lactose 


CaO 


Total ee es 
Less 5.4 per cent 


Calories physiolog- | 


= 


ically available 


The addition of 30 grams of lactose hydrate strengthened the 
milk with 27.g grams of lactose per litre, containing 108.8 calories. 
Deducting 5.4 per cent from this (= 102.9), it is evident that the milk 
fed to the lactose pig had an energy equivalent of 474 calories per 
litre. 

Thirty grams of dextrose added to the skim-milk raised its value by 
110.8 calories (— 5.4 per cent = 104.8). The dextrose pig was, there- 
fore, fed with a calorific equivalent of 475.9 calories per litre. 

The relative calorific value of the three milks may be computed: 


Dextrose mille. 35+ . 2 1000 
TACOS) Vipdie) lee) Ie tee Se ee 
Skim=milk: 2s 2h eee cn dae 


Of 100 calories in the food there were: 
Skim. Lactose. Dextrose. 


Proteid’ <7 mo 47.0 


Rats. 


Carbohydrates 


Suckling Pigs on a Diet of Skimmed Cow's Milk. 203 


The value for proteid is here very high. Rubner,' calculating the 
gross calories of mother’s milk on the fourteenth day of lactation, states 
that the proteid contains 16.7 per cent, the fat 47.2 per cent, and 
lactose 36.1 per cent of the calories. 


Skim. Lactose. Dextrose. 


Day. nee 
Milk. | Weight.| Milk. |. Weight. | Milk. | Weight. | 1902 


Cc. Gms. BC Gms. c.c. Gms. 


1322 ee 1295 Boe 1485.7 Jan. 18 


1541.5 aes 
1525 
1380 1502 
1495 : 1610 
1601 5 1643 
1699 
1762 
1791 
1907 
iL? 2 7) 
207+ 


2090 


The amount of proteid in sow’s milk is normally high. Kdénig ? 
gives 7.24 per cent as the mean in eight sows, while six days after 
parturition it is given as high as 12.89 per cent. The pig’s organism 


1 RUBNER: Leyden’s Handbuch, Bd. 1, p. 134. 
? KONIG :.Zusammensetzung der menschlichen Nahrungsmittel, 1889, p. 350. 


204 Margaret B. Wilson. 


may, therefore, be fitted for a high proteid coefficient. The mean for 
_milk-sugar in the sow’s milk is given as 3.13 per cent, and for fat 
ASeAde bi5t 

The three pigs which were fed on the three milks above described 
grew rapidly and normally. The dextrose pig, which was heaviest at 
the start, took his food with greater slowness than the others, some- 
times suckling forty-five minutes in taking the volume of milk which 
the other two drank in fifteen minutes. There were nine feedings on 
each of the first two days, seven feedings from the third to the seventh 
day, and six feedings daily thereafter. The pigs were fed every three 
or four hours, including the night hours, as they would awake between 
three and four o’clock and cry for food. The pigs were all strong and 
active, and exercised freely about the room. 

The amount of milk fed, and the growth of the pigs is shown in the 
table on page 203. 

From this it is seen that in sixteen days the skim-milk pig drank 
10.925 litres (=4053.2 calories), the lactose pig drank 11.005 litres 
(= 5216.4 calories), and the dextrose pig drank 9.707 litres (= 4619.6 
calories). 

It is now possible to tabulate some of the facts regarding the growth 
of the pigs, and to compare these results with those of Sanford and 
Lusk. 


WILSON. SANFORD and Lusk. 


Lac- Dex- 
tose. trose. 


Lac- Dex- 


i f Skim. 
tose. trose. || 


Weight when bom . . . . . . .| 1322 | 1295 | 1485 |/1000 | 1050 1152 


Weirht whenkilled) 95 ces) sen e205 2435 | 2471 ||1246 | 1890 | 2000 


Growth peoramse. | alen nee eennne 883 1140 986 || 264 $38 S48 
Growth ny percentqn- 0a mene 66.8 88.0} 66.4}| 264 yh dy 736 
Milk fedinec. . . . . . . « .|10925 |11005 |9707 |/6826 |8836 | 9481 

| | 
Available caloriesfed . . . . . .| 4053 | 5216 | 4620 |/2339 | 3736 | 3972 
Growth in grams per litre of milk. . $l | 114 10] 38 95 SY 


Growth in grams per 1000 calories fed 


It is seen from the above that the growth of the pigs in grams is 


directly proportional to the calorific value of the food to the organism. 


Suckling Pigs on a Diet of Skimmed Cow's Milk. 205 


The one exception was that of the ill-nourished skim-milk pig of Sanford 
and Lusk. Here is an individual improperly nourished, taking too 
little food, and remaining behind his fellows in normal development. 
But that five out of six pigs of different litters, different sizes, and 
differently fed should gain in weight, respectively, 213, 213, 215, 
218, and 222 grams per thousand calories fed, seems more than a 
coincidence. Oppenheimer! has shown that the growth in grams 
of normal breast-fed children of the same age may be nearly propor- 
tional to the litres of milk fed. Here the milk presumably has the 
same calorific value, although this could not be determined. Oppen- 
heimer’s table is here reproduced : 


Growth in grams for 1 kg. Milk. 


Month. FEER. OprENHEIMER. 
I 33.8 95.0 
II isa") 201.1 
III 120.3 138.5 


In the second, third, and fourth months of these children’s lives the 
growth is uniformly proportional to the milk, or energy, fed. 

It seems, therefore, probable that the growth of sucklings in good 
health, and fed with proper and abundant food, stands in a definite 
relationship to the amount of energy in the food. 

Having determined this ratio of growth, it seemed desirable to 
estimate the constituents of the new tissue, and if possible to calculate 
the energy equivalent added to the body in the course of this growth. 

For this purpose the pigs were analyzed. Three of the new-born 
pigs were first analyzed. The results were so similar in all three, that 
it is justifiable to assume that the three bottle-fed pigs of the same 
litter started with the same percentage composition. Any increase 
above this estimated original composition may be ascribed to growth 
under the influence of diet. The proteid content was estimated by 
multiplying the amount of N found by 6.25, without! regard to the 
extractives. 


1 OPPENHEIMER: Zeitschrift fiir Biologie, Igor, xliv. p. 147- 


206 Margaret B. Wilson. 


In preparing the pigs for analysis, each animal was dissected, and 
the intestinal canal cleaned of its contents. The different portions of 
the pig were put in dishes, covered with 95 per cent alcohol, left 
for twenty-four hours, and evaporated over a steam bath. This pro- 
cess was repeated two or three times. All the organs could then be 
put through a sausage machine, except the bones. The parts were 
dried in a drying oven at 97°, ground in a mortar or put through a 
coffee-mill, and finally passed through a sieve. The skin with the hair, 
the bones, and all the other parts were thus reduced to a homogeneous 
mixture yielding equal values in duplicate analysis. Nitrogen was 
determined by the Kjeldahl method, moisture by drying to constant 
weight, fat by prolonged ether extraction, which, according to E. 
Voit,! will yield 95-97 per cent of the fat present. CaO was deter- 
mined after incineration. 
The results were as follows : 


| 
| Mk Sglisise (li. | ee oo | eee | ea par oeee 
| | 
ae | | A, ee eel | . 
Pig I 104477) 20520 13270) 96:65) alae 62.6 19.28 9.40 
Jul W422 2265 | 17.65 | oit2 136.1 | 60.1 
Seti | 1016 | 207.9 | 16 61.1 
“ skim | 2205 | 436.2 |.28:3 
** Jactose 2435 | ADS: 1 | S480 
* dextrose rile | AOS esac 


The aggregate live weight of the new-born pigs I, II, and III was 
3,247 grams, and their percentage composition may be thus computed : 


Grams. Per cent. 


Live weight . ere act hoe ory ct 

Drysolidsa; 2 sc) ane Sie oe ee eo tee 

Bata) ic gdoaclaa at. Score cee TIES 1.15 
Proteid. ¢—=2 6, 50s Be le 12.0 

CaO (average of two'pigs)/; 2 2) ee == oo 


. 


With these figures it is possible to obtain the data regarding the 
growth of the three pigs which were fed. 


1 Voit, E.; Zeitschrift fiir Biologie, 1897, lv, p. 55. 


Suckling Pigs on a Diet of Skimmed Cow's Milk. 207 


Estimated composition Composition as deter- 


at start. mined at end. Growth. 
Pig. ar a p24 pee zy a 
Solids.| Fat. | P™® | cao.||Solids.| Fat. | P™° |ca0.||Solids.| Fat. | E'S | cao. 
teid. teid. | teid. 
Skim 260.4 | 15.20| 158.90 | 24.85 || 436.2 28.34 321.3 | 36 17|| 175.8 | 13.18 | 162.4 |.11.32 


4 
Lactose | 255.1 | 14.89} 155.66 | 24.35 || 495.7 | 34.88 360.0 | 39.77|| 240.5 | 20.01 | 204.3 | 15.42 


Dextrose | 292.5 | 17.08 | 178.50| 27.92|| 495.2 | 33.24 | 349.1 | 40.27|| 202-7 | 16.16 | 170.6 | 12.35 


It is clear that the milk-sugar pig made the most progress, the dex- 


trose pig next, and the skim-milk pig the least. 

It will now be interesting to note to what extent the diet affected 
the growth. Recalling the composition and amount of the milk in- 
gested it is possible to estimate the percentage of the food constitu- 
ents absorbed and used for growth : 


Per cent used for 


Food fed. Growth. an 
growth. 
Pig. p : Lal | a is 
Proteid. | Fat. | CaO.|| Proteid. | Fat. | CaO.}| Proteid. Fat. | CaO. 
Skim 455.6 15.29 | 21.74 162.4 leypiiery MM Nes 35.6 86.00 | 52.00 
Lactose 458.9 15.40 21.90 204.3 20.01 | 15.42 44.5 130.00 | 70.00 


Dextrose 404.8 | 13.59] 19.321) 170.6 |.16:16 1.12.35 ||) 42.1 111.00 | 64.00 


| 
| 


Proteid, it is apparent, was well utilized for growth. The skim-milk 
pig used 35.6 per cent of the proteid fed for growth, the lactose pig 44.5 
per cent, while the dextrose pig used 42.1 per cent. The sugars 
naturally promoted. the saving of proteid for the organism. This 
proportion of proteid retained for growth accords with the 40 per cent 
given by W. Camerer, Jr.,! for a nine-weeks old nursing baby, and 
with the values already cited in Sanford and Lusk’s experiments, 7. ¢., 
23 per cent for the skim-milk, 38 per cent for the lactose, and 48 per 
cent for the dextrose pig. 

The increase in fat in the pigs was surprising in view of the poverty 
of the milk in this material. The newly-born pigs were all very poor 


1 CAMERER: Zeitschrift fiir Biologie, 1go2, xliii, p. I. 


208 Margaret B. Wilson. 


in fat at the start, and they seem to have absorbed and deposited most 
_of the fat ingested. The skim-milk pig deposited the least fat, while 
those fed with the added sugars apparently increased in fat more than 
would have been possible if this increase had been derived from the 
amount of fat in the food. Two explanations are here possible, either 
the fat in the milk was slightly underestimated (a small factor would 
make considerable difference) or milk-sugar was converted into fat. 
The latter would have been possible through the intestinal inversion 
into dextrose and galactose, and subsequent conversion of a part of 
the galactose into glycogen. In this way perhaps 75 per cent of dex- 
trose arises from milk-sugar, which dextrose might be partly converted 
into fat. The principal point which is indicated by this almost com- 
plete assimilation of the fat fed to the pigs, is that the carbohydrates 
fed were amply sufficient to furnish the needed energy, while a normal 
growth of tissue was progressing. No fat combustion was necessary 
for the life-processes. 

Of the calcium in the food, the skim-milk pig used 52 per cent for 
growth, the lactose pig 70 per cent, and the dextrose pig 64 per cent. 
It would seem from this that diet had an influence on the absorption, 
but a more correct idea is given when it is remembered that the dry 
solids of the three pigs contained 8.29, 8.02, and 8.13 per cent of cal- 
cium respectively. It seems, therefore, that the calcium addition 
depended rather upon the development of the organism than upon 
any specific influence of the milk constituents. It has been said that 
absorption of the calcium of skim-milk is imperfect, and that children 
are inclined to become rickety when so fed. Herter! has found strik- 
ing retardation in the development of the skeleton in‘older pigs after 
feeding them on skimmed milk for many months. He, however, notes 
no signs of rickets. It might be due to the general malnutrition in- 
duced by a monotonous or iron-poor diet like skim-milk. Certainly my 
experiments show no untoward influences. W. Camerer, Jr.,? finds 
that the amount of calcium in mother’s milk is barely sufficient to 
cover the needs of the nursing infant, if the percentage composition of 
the five-months old baby is the same as that of the new-born baby. 
The explanation of this may be derived from my experiment, which 
shows that the percentage of calcium content of the solids in the new- 
born pig averages 9.40 per cent, in contrast with 8.15 per cent at an 
age of two weeks anda half. If at the end of sixteen-days nursing 


' HERTER: Journal of experimental medicine, 1898, iii, p. 293. 
* CAMERER: Zeitschrift fiir Biologie, 1902, xliii, p. 1. 


Sucking Pigs on a Diet of Skimmed Cow's Milk. 209 


the lactose and dextrose pigs had contained 9.40 per cent CaO, an 
almost complete absorption of CaO would have been necessary. 

The study of the growth of the suckling animal would not be com- 
plete without an attempt to calculate the energy liberated in the animal. 
We may calculate the number of calories in the food, and we may cal- 
culate the number of calories contained in the proteid and fat added to 
the body in the process of growth. Rubner! has shown that the heat 
value of decomposing proteid tissue in starvation is not far different 
from the heat value of meat. Of course proteid metabolism in starva- 
tion is at the expense of all the organs. Rubner shows a heat value 
of 24.98 calories for 1 gram of nitrogen from proteid metabolism in 
starvation. This value has been adopted in the calculations. It has 
not been possible to enter into the individual heat equivalents of the 
hair, skin, hoofs, etc. The heat value of the fat has been taken as 9.3. 

The growth of the pigs in potential energy is as follows: 


N. Fat. Calories Calories =o) 
wt Sek F pase p Total. 
Grams. Grams. in proteid. in fat. 
SH Sth Re 25.98 13.18 649 ies: 772 
WRARLOSE =. << 22 32.68 20.01 816 186 1002 


682 


Dextrose 16.16 


Of the energy added to the skim-milk pig, 16 per cent was in 
fat; to the lactose pig, 18.6 per cent, and to the dextrose pig, 18 per 
cent was in fat. These data now permit the calculation of the follow- 
ing table: 


| 
Skim. Lactose. | Dextrose. 
Calories. | Calories. | Calories. 


PPE OGG LN TR BAe ey ee a! Be Bes 5 5216 


Stored in the body 

Liberated as heat . 

Fed daily per square metre of surface 

Liberated as heat daily per square metre of surface 


Retained per 1000 calories in the food 


1 RUBNER: Zeitschrift fiir Biologie, 1894, xxx, p. 94. 


210 Margaret B. Wilson. 


The calculations of the surface were made from the formula of 
Meh in which the value of & was taken to be 9.02 as suggested by 
-E, Voit.!. For g in the formula the average weight during the whole 
experiment was taken. 

The energy requirements of these pigs were much higher per 
square metre of surface than those of the human infant or of the 
adult pig.2 The human infant requires daily in the food about 1200 
calories per square metre of surface, while a more active older child 
requires 1500. The 2400 found for the lactose pig corresponds to 
that needed by a man at hard labor. The results as obtained with 
the pigs are to be explained only with the premises that the very 
active pigs nursing every three hours represented to a large degree 
a life of hard work. The bottle was usually taken with great vigor, 
and the pigs did much scrambling among themselves. 

The most striking result, however, is in the proportion of calories 
retained in the pig for growth. It has been said that the pigs gained 
in weight in proportion to the calories in the food: not only is this 
true, but the calories retained in the pigs were proportional to the 
calories in the food. The following table indicates the comparison : 


Skim, Lactose. | Dextrose. 


Gain of pig in grams per 1000 calories in the 
FOO) 0, aay ee he so ea 

Gain of pig in calories retained per 1000 calories 
in the food 


The storage of 18 to 19 per cent of the energy in the food in the 
formation of new tissue finds a counterpart in a respiration experi- 
ment cited by W. Camerer, Jr.,?> where 15 per cent of the energy in 
the food of a nursing baby nine weeks old was retained in the body, 
and in a similar experiment of Rubner and Heubner* ona baby seven 
and a half months old, where 12.2 per cent were so retained. 

The results accord with Oppenheimer’s calculations before men- 
tioned, which showed that the gain in weight of growing babies of 
the same age was proportional to the milk ingested. 


| E. Voir: Zeitschrift fiir Biologie, 1901, xli, p. 113. 
al Nee "oa tiger br Tpen e776 

$ CAMERER : Loc. cit., p. (0. 

4 RuBNER und HEUBNER: Zeitschrift fiir Biologie, 1899, xxxviil, p. 345. 


Suckling Pigs on a Diet of Skimmed Cow's Milk. 211 


But the results show further that a large and apparently normal 
growth takes place when skimmed milk, containing a plentiful supply 
of proteid, is fed, and that this growth is favorably influenced by the 
addition of sugars, which improve the nutritive value in that they 
add further calorific power to the milk. 

Bunge ! states that a suckling pig doubles in weight in eighteen 
days. The skim-milk pigs gained as much as 80 per cent in weight 
in from fourteen to sixteen days, an apparently normal growth. 

The results of this research Jead me to believe with Camerer that, 
provided the milk contains sufficient proteid, the suckling can increase 
in size and strength, whether the necessary additional fuel-value is 
furnished by fat, milk-sugar, or dextrose. 

The presence of fat in the milk greatly reduces the bulk necessary 
to supply the needed calories, and too much water is to be avoided ; ” 
but when milk-fat disagrees with an infant, there seems to be reason 
to believe that skim-milk with milk-sugar can be fed without preju- 
dicial effect. 

It is to be regretted that a comparison with pigs reared upon 
whole milk is impossible. The great labor involved does not at 
present permit the attempt of the series of experiments necessary for 
such a comparison. 


SUMMARY. 


I. Skimmed cow’s milk, with or without 2 to 3 per cent of added 
lactose or dextrose, is normally absorbed by suckling pigs. 

2. Two pigs fed on skim-milk from fourteen to sixteen days gained 
26.4 and 66.8 per cent in weight. Two pigs fed on the same skim- 
milk, with 2 and 3 per cent of lactose added, gained 79.7 and 88.0 
per cent in weight. Two pigs fed on the same skim-milk, with 2 and 
3 per cent dextrose added, gained 73.6 and 64.4 per cent in weight. 

Those fed with plain skim-milk gained 114 and 218 grams in 
weight for every 1000 physiologically available calories in the food. 
The lactose pigs gained 222 and 215 grams per 1000 calories in the 
food. The dextrose pigs both gained 213 grams per 1000 calories 
in the food. Except in the case of one ill-nourished skim-milk pig, 
the growths of the sucklings stand in a constant ratio to the calories 
in the food. 


1 Cited by ADERHALDEN: Zeitschrift fiir physiologie Chemie, 1899, xxvi, p. 497. 
2 RUBNER und HEUBNER: Zeitschrift fiir Biologie, 1899, xxxviii, p. 397. 


212 Margaret B. Wilson. 


4. The pigs fed on plain skim-milk used 23 and 35 per cent of the 
proteid in the food for tissue growth, the lactose pigs used 38 and 44 
per cent, and the dextrose pigs 48 and 42 per cent. 

5. All the pigs of the second litter gained in fat when fed on plain 
skim-milk or on skim-milk with sugars. 

6. The percentage of calcium in the bodies of the pigs diminished 
with their growth. There was considerable and normal deposition 
of calcium in the pig, and this was proportional, not to the calcium 
in the food, but to the growth of the animal. 

7. Not only is the growth in grams of the pigs proportional to the 
calories in the food, but the number of calories retained in the tissue 
substance during growth is proportional to the calories in the food. 
18 to 19 per cent of the calories in the food were found stored in the 
tissue growth of the pigs fed on the three varieties of skimmed milk. 

8. There seems to be striking evidence that the suckling pig 
reared on skimmed cow’s milk conforms to the same laws of nutri- 
tion as the breast-fed infant. 


— a 


_eSS—t—“=‘; 


MAXIMAL CONTRACTION, “STAIRCASE” CONTRAC- 
TION, REFRACTORY PERIOD, AND COMPENSA- 
nORY PAUSE, OF THE HEART. 


By. Kk. Ss. WOODWORTH. 


[From the Laboratories of Physiology in the Harvard Medical School and in the University 
and Bellevue Hospital Medical College.|} 


I. EXPERIMENTS ON THE APEX OF THE Doa’s HEART. 


HE mammalian apex preparation, introduced by Porter,” has not 

as yet been used for testing the response of cardiac muscle to 

electrical excitation. It is, however, very suitable for this purpose, 

first because, if supplied through a nutrient artery with warmed, oxy- 

genated, defibrinated blood, it beats spontaneously, and secondly 

because it is devoid of ganglion cells, and thus is a true muscle prepa- 
ration. 

It was found desirable in many of the experiments to record the 
moment of stimulation on the muscle curve itself by means of the 
spark chronograph.*? A strong battery current was passed through 
an interrupter (adjusted commonly to ;'5 sec.), an electromagnet 
designed to serve as a time-marker, and the primary coil of an 
inductorium. One pole of the secondary coil was connected to the 
recording drum, and the other pole to the apex preparation, thence 
through a fine copper wire to the writing lever (the wire also served 
to suspend the lever from the apex preparation), and by a continua- 


1 These experiments were begun in 1897-8 under the direction of Dr. W. T. 
PorTER. Left incomplete then, and interrupted by other work until recently, they 
were resumed in 1go!-2. The results of this second series, performed after so 
long an interval and in a different laboratory, as well as with slightly changed 
apparatus, agree perfectly with the earlier results and may almost be regarded as 
an independent confirmation of them. 

2 PoRTER: Journal of experimental medicine, 1897, ii, p. 391; also: This 
journal, 1898, i, p. 514. The essential parts of this apparatus are shown in Fig. I 
of Dr. CLEGHORN’s paper, This journal, 1899, ii, p. 275. 

3 For a convenient reference to the spark chronograph, see SCHAFER’s Text- 
book of Physiology, 1900, ii, p. 750. The arrangement of the electric circuits in 
my apparatus differed considerably from that described by SCHAFER. 

213 


214 . RR. S. Woodworth. 


tion of this wire inside the lever to the writing point of copper foil. 
A strong spark, by jumping from the writing point through the 
paper to the drum, knocked off a little lampblack and left a round 
spot on the muscle tracing at the instant of stimulation. A key 
was placed in the secondary circuit. To moderate the current sent 
through the heart, a shunt was sometimes used. 


MAXIMAL CONTRACTION. 


The peculiar properties of cardiac muscle, discovered by Bowditch 
on stimulation of the apex of the frog’s ventricle,! hold in full force 
for the perfused apex of the dog’s heart. Chief among these prop- 
erties is the law of maximal contraction, or of “all or none.”? Any 
contraction, however weak be the stimulus that arouses it, has the 
maximal force that the muscle can exert at that moment. Increasing 
the stimulus above its minimal effective strength does not increase 
the force of the response. The minimal stimulus is also maximal. 

I have demonstrated this property as follows. A series of con- 
tractions was first produced by minimal stimuli. When the height of 
contractions was well established, the intensity of the stimuli was 
suddenly greatly increased. After some time, the intensity was re- 
duced to its former point. But no changes in the height of contrac- 
tion were produced by these changes in the stimulus. 


STAIRCASE CONTRACTION. 


A second property of cardiac muscle, discovered by Bowditch ® in 
the frog, and holding also in the dog, is “ staircase” contraction. Ifa 
quiescent apex is excited by a series of stimuli, the first response is 
usually feeble, and the later ones increase gradually to a maximum, 
at which level they remain. If, however, the rate of stimulation is 
then hastened, a new staircase results. Or, if the rate is slackened, 
the result is a descending staircase (Fig. 1). 

The chief point of difference between staircase contraction in 
amphibian and in mammalian cardiac muscle is the length of the 
optimum interval, Bowditch* found that the strongest series of con- 
tractions was elicited from the frog’s apex by stimuli having an 


' Bowpircu: Arbeiten aus der physiologischen Anstalt zu Leipzig, 1871, p. 139. 
* Confirmed by MCWILLIAM on the mammalian heart. Journal of physiology, 


1888, ix, p. 168. 
30WDITCH: Of. cit., p. 156, ’ 4 BowbDITcH: Of. cét., p. 160. 


A i 


WS 


—-T” °°» 


Contractions of the Heart. 215 


interval of four or five seconds... In the dog’s apex, the optimum 
interval is about one second, fiayfllly somewhat less. 

The length of the optimum interval is governed by the interplay of 
two opposing factors, as will be clear froma study of Fig. 2. No long 
series of rapid beats is neces- 
sary to produce an increase 


' | HW] | Hh 

in height. A single con- 
traction following another i \ | 
at a short interval is suffi- 
cient to increase the height 
of the following contrac- 
: FIGURE 1.— Dog’s apex. Electrically excited con- 
tions. The shorter also aie ee i er meee 

: tractions. ‘‘ Staircase” resulting from increase 
the interval between two in rate of stimulation, and descending stair- 
contractions, the greater is case resulting from decrease in rate. The 
their strengthening effect interval between the slow stimuli is about 33”, 

5 and that between the rapid stimuli about 1”. 

on the following contrac- 
tions. The shortest interval in Fig. 2 is about half a second. It 
will be seen, however, that a contraction following very quickly 
after another is itself very small, and that at the beginning of the 
long series of rapid stimuli, the first two contractions are weaker and 


Hl i ili 
i] 


tel | LUELLA EE 


FiGuRE 2.—See legend of Fig. 1. Note in addition three extra contractions, probably 
spontaneous, and the increased height of the contraction which follows each. 


not stronger than those that precede, whereas at the end of the rapid 
series, the first two slow contractions are stronger than the contrac- 
tions that precede. 

The two opposing factors are then ¢he stimulating effect of a rapid 
succession of contractions, and the recuperative effect of a long pause. 
On the one hand, the following of one contraction close upon another 
acts to accelerate the production of available energy immediately after- 
ward; but on the other hand, the production of available energy is 


Between, the result of trying to tetanize the 


rhe shocks are indicated by dots, which were produced by sparks 


One spontaneous contraction at the start and two at the close. 
passing through the paper between the drum and the’writing point. 


Dog’s apex. 


> 
Pe 


FIGURE 


i 


muscle by a series of strong induction shocks, 18 per second. 


R. S. Woodworth. 


a gradual process, and a long pause enables 
more to accumulate than does a short pause. 
A short interval preceding a contraction tends 
to make that contraction weak and the follow- 
ing contractions strong. 


The strongest possible contraction is se- 
cured, therefore, by producing, first, two or 
more contractions in rapid succession, so as to 
accelerate the production of energy, and allow- 
ing after that a considerable interval for its 
accumulation. And the smallest contraction 
may be secured by first depressing the con- 
tractility by a slowseries of contractions, and 
then following the last of these by another 
contraction at the briefest possible interval. 
Bearing in mind the two opposing factors, it 
has been possible for me to produce strong or 
weak contractions at will, and to predict with 
considerable certainty the height of each spon- 
taneous contraction. The ability to control 
and predict is good evidence of the reality of 
the factors in question. 

Further evidence is seen in the tracings ob- 
tained on attempting to tetanize the heart 
muscle. Fig. 3 shows the stimulating effect 
of a rapid series of contractions, and it also 
shows that the full benefit of such stimulation 
is only got by means of a considerable pause. 

Incidentally, one may note in this tracing 
the fact that the mammalian apex, like the 
frog’s heart,! does not respond to faradiza- 
tion by a true tetanus. It responds either by 
fibrillary contraction or by a rapid series of 
beats. Walther? has found in the frog’s 
ventricle something closely resembling the 


1 KRONECKER and STIRLING: Beitrage zur Anat- 
omie und Physiologie als Festgabe zu Ludwig, 1874, 
Bi S73 

2 WALTHER: Archiv fiir die gesammte Physiologie, 
1900, Ixxvili, p. 605, p. 624. 


Contractions of the Fleart. 217 


incomplete tetanus of striped muscle, with superposition, in case the 
heart, having rested for some time, was capable of staircase con- 
traction. Walther holds that wherever staircase contraction would 
be produced by a series of single shocks, an incomplete tetanus would 
result from faradization. Now since the dog’s apex is capable of 
staircase contraction at almost any time, even when beating spon- 
taneously, an incomplete tetanus should accordingly be always 
aroused by faradization. I have never secured this result. I have 
obtained very rapid series of beats showing staircase contraction, 
but nothing comparable to the superposition of one contraction on 
another, such as appears in true tetanus. 

In Fig. 3 is seen also an alternation of stronger and weaker con- 
tractions. Essentially the same phenomenon has been noted by 
Langendorff! and by Hofmann? in the frog’s heart, and by Cushny 
and Matthews ® in the intact dog’s heart. It is more pronounced in 
the dog’s apex than in the frog’s heart, and is usually harder to avoid 
than to secure. 

The question arises, whether this alternation can be explained by 
the two factors already mentioned, namely, the stimulating effect of 
a rapid succession of contractions, and the recuperative effect of a 
long pause. 

The explanation already offered by Cushny and Matthews and by 
Hofmann takes account of one of these factors, namely, the recupera- 
tive effect of the pause. Since a very strong contraction is also 
much prolonged, the pause following it is shortened; consequently, 
‘the succeeding contraction is weak. The weak contraction, being 
also of short duration, is followed by a long pause which strengthens 
the next contraction. 

Reference to Fig. 3 will show this explanation to be possible. But 
explanation is equally possible in terms of the other factor. The con- 
traction following a prolonged contraction is practically hastened, and 
so doubtless exerts a stimulating after-effect, etc. It is probable that 
the two factors co-operate to produce the alternation, but, in the dog’s 
apex, the effect of hastening a contraction is stronger than the effect 
of prolonging a pause.* 


LANGENDORFF: Archiv fiir Physiologie, 1885, p. 287. 
HorMAnn: Archiv fiir die gesammte Physiologie, 1901, Ixxxiv, p. 145 ff. 
3 CusHny and MATTHEWS, Journal of physiology, 1897, xxi, p. 222. 
4 Fig. 3 is also of some interest as showing that the curves of returning irrita- 
bility and contractility after a contraction may differ considerably. The irritability 


1 
2 
3 


218 . Rk. S. Woodworth. 


No notice has thus far been taken of two possible objections to the 
explanation offered for staircase contraction. 

First, the question may be raised whether the stimulating effect of 
a rapid series of contractions is not rather to be traced directly to 
the electricity used to elicit the contractions. The answer to this 
is that the same stimulation appears, without electricity, when, from 
any internal cause, the spontaneous rhythm of the apex is quickened ; 
also, that no electrical stimulation, applied during the refractory 
period of successive spontaneous contractions, avails to produce stair- 
case contraction. 

Secondly, it may be questioned whether the stimulating influence 
of small interpolated contractions, such as appear in Figs. 2 and 3, may 
not be due merely to their smallness, rather than to the quickness with 
which they follow the preceding contractions. Does not, in short, 
a weak contraction tend to make the next one strong, and a strong 

contraction to make the 
next one weak, and is 
not this the cause of 
| the alternation? The 
difficulty with this 
explanation is that it 
breaks down in the 


case of staircase con- 

wiki oby se ok eas la ge traction. There, each 
FIGURE 4+.— Dog’s apex. Spontaneous contractions, at contraction 1s strong 
intervals of about 31”. Several hastened spontaneous by comparison with the 


contractions, also two contractions excited by electric one before yet shows 
: ‘ 
shocks. 


C 


no tendency to weaken 
the next contraction, 
Further, the tracing in Fig. 4 gives direct evidence for the view 
here adopted, and against the two objections just raised. This trac- 
ing shows that a spontaneous hastened contraction strengthens the 
following contractions, in the same way as a contraction hastened by 
electrical stimulus. It also shows that a contraction need not be 
weak in order to strengthen the next contraction; it need only be 
hastened. It is indeed impossible much to hasten a contraction 
without making it weak, but if the hastening is slight the weakening 
is not perceptible, and yet the following contraction is perceptibly 
strengthened. 
must be practically equal at the beginning of each contraction, since the stimulus 
is nearly constant, but the contractility is far from being always equal, 


* 


Contractions of the Heart. 219 


THE REFRACTORY PERIOD. 


During the systole or contraction-phase of the apex, 7. ¢. during the 
period corresponding to the ascending limb of the cardiomyogram, 
the muscle is refractory to stimuli. This condition persists through- 


IS | 14 13. IS 14 |}7| 8] 12 IS | 13 | 13 12 S |-3 (717) 4 | 2) 12) 


Ficurt 5.—Dog’s apex. Spontaneous contractions. Also stimulation during contrac- 
tion and during relaxation. Points of stimulation indicated by round spots along the 


1/7 


course of the muscle tracing. Time in ;;”, and numbers are in terms of this as unit. 


out the whole of the systole. Stimulation during this phase neither 
alters the force and duration of the contraction then in progress, nor 
arouses an extra contraction after it. See Fig. 5. As this is an 


Any SU 


DLE bbb rh both hhh hhh bt HOLT TT Oth nr err 


4 


FicuRE 6.— Dog's apex. Spontaneous contractions. Exact limit of the refractory 


period. Point of stimulation indicated by dot on the muscle tracing. Time in 75”. 
admitted property of cardiac muscle, its appearance in the apex calls 
for no discussion. A few details should be recorded: 

1. The refractoriness is absolute. I have used induction shocks 


220 R. S. Woodworth. 


strong enough to throw a two-millimetre spark, and also direct bat- 
tery currents of six volts, without eliciting any response. 

2. Weak contractions as well as strong, and electrically excited as 
well as spontaneous contractions, show the same refractory period. 

3. The end of the refractory period, as is illustrated by Fig. 6, 
is at the very summit of the myogram. The beginning of descent 
marks the beginning of the irritable period. 

4. The irritability is not restored all at once at the beginning of 
diastole, but by degrees. Very strong shocks are necessary to elicit 
an “extra” contraction early in diastole, but comparatively weak 
shocks suffice when relaxation is almost complete. 


THE COMPENSATORY PAUSE. 


When an intact heart, including ventricle, auricle, great veins and 
sinus, is beating normally, and an extra contraction is aroused by 
electrical or mechanical stimulus applied to the auricle or ventricle, 
a long pause follows the extra contraction.!. On measuring the 
length of this pause, it is found to compensate very exactly for the 
shortness of the interval preceding the extra contraction. The earlier 
the extra contraction has come, the longer is the pause. The time 
measured from the beginning of the last regular systole before the 
extra contraction to the beginning of the first subsequent regular 
systole is equal to two regular beat periods. The ventricle misses 
one regular beat, and waits till the regular time of the next beat. 
Thus the rhythm of the heart is only momentarily disturbed by the 
artificial stimulus, and the following beats come at the same instants 
as if no extra contraction had occurred. 

The question comes, whether the spontaneously beating mammalian 
apex shows this familiar action of the intact heart. It may be stated 
definitely that it shows nothing of the kind. An examination of Fig. 
5 sets the matter at rest. Fig. 4 is equally conclusive. Fig. 6 is less 
suitable for purposes of demonstration, since the rate of the sponta- 
neous beat is uneven, yet it shows clearly that the extra contraction 
is not followed by a prolonged pause. Not only is there no exact 
compensation, but there is not the slightest tendency to prolong the 
pause. On the whole, the pause following the extra contraction is 
shorter than the regular pause. 


1 MAREY: Travaux du laboratoire, 1876, p. 73. 


ae 


LS) (eg per ae ie 


Contractions of the Heart. 221 


These statements will be fortified by a closer examination of Fig. 5. 
The regular periods before the first extra contraction measure 13 and 
14 units of time. A double period is then 27 units. If there were a 
compensatory pause, 27 units would measure the time from the last 
regular systole before the extra contraction to the first after it. The 
actual measure of this time is 15 units. In order to compensate, the 
time from the beginning of the extra contraction to the next regular 
beat should measure 27—7 = 20 units, whereas it actually measures 8 
units, considerably less than the regular single period. 

The tracings presented are by no means exceptional. In an exami- 
nation of several hundred cases, I have found only sporadic instances 
in which the pause following the extra contraction was even approxi- 
mately compensatory. In a few cases, it is too long to compensate ; 
but in the great majority it is too short. 

In order to allow due weight to the exceptional cases in which the 
pause following the extra contraction was prolonged, I have measured 
somewhat over a hundred cases, including among them the longest 
pauses I could find, and taken the average of them. I find the average 
duration of the double period, measured from the beginning of the last 
systole before the extra contraction to the beginning of the first systole 
after it, to be 78.2 per cent of the regular double period. 

Lest any reader should still be inclined to ascribe some special 
significance to the few cases in which the pause was approximately 
compensatory, I add, in Fig. 7, the 
“ distribution curve” of the lengths 
of the double periods that included 
the extra contraction. Each is ex- 
pressed as a fraction per cent of 
the regular double period preceding. 
Each dot in the figure represents 
one case, and it is seen that the 


cases are most frequent in the 449 60 80 100 120 140 160 180 200 
neighborhood of 60 and 70 per 


cent (56-75 per cent) of the es lengths of the double period includ- 
lar double period. The cases are ing the extra contraction, expressed 
E g P 
distributed in a shape similar to in per cent of the regular double 
ies probability neue atone period. Each dot represents one case, 
’ D 

the present curve is somewhat “skew.’’ The few cases near 100 per 
cent (2. e. nearly compensatory) fall easily into the general distribu- 
tion curve, and there is no sign that they form a class by themselves 
or possess any special significance. 


FIGURE 7.— Distribution curve of the 


222 R. S. Woodworth. 


The double period including an extra contraction of the sponta- 
neously beating apex is thus a variable quantity, which occasionally 
assumes values near to the regular double period but usually values 
much less. It differs in two respects from the corresponding period 
in cases where there is a true compensatory pause: it is, on the 
average, shorter, and it is much more variable. 

The single period, measured from the commencement of the extra 
contraction to the next regular systole, averages 94.1 per cent of the 
regular single period (or 88.5 per cent, if we leave out the two very 
long pauses seen in Fig. 7, which are undoubtedly abnormal, since 
they come from pieces of apex that had not much spontaneity). In 
other words, the usual effect of the extra contraction is to hasten the 
rhythm of the apex, at least for one following beat. 

In the intact heart, it will be remembered, the earlier the extra con- 
traction comes, the longer is the compensatory pause. The reverse 
is true of the spontaneously beating apex: the earlier the extra con- 
traction comes, the shorter is the following pause. This is at least 
the average tendency. It was detected as follows: The time from 
the beginning of the last systole before the extra contraction to the 
beginning of the extra contraction (which time I call ‘interval before 
extra’) was measured, likewise the time from the beginning of the 
extra contraction to the beginning of the following spontaneous sys- 
tole (‘‘interval after extra’). Each of these intervals being then 
expressed as a per cent of the regular single period just preceding, 
the following relations appear : 


Interval before extra. Interval after extra. 
40-70%, 80.9 %, average. 
70-90 “ 102.9%, average. 


As it seemed possible that the apex might show a delayed compen- 
sation for the hastening of the extra contraction, I measured also the 
double period beginning with the first systole after the extra contrac- 
tion. I found, however, no compensation. When the double period 
including the extra contraction was short, the following double period 
was also usually short; but in the few cases in which the double 
period including the extra contraction was prolonged, the following 
double period also was prolonged. In other words, the effect of the 
extra contraction on the rhythm of the apex was somewhat persis- 
tent, usually hastening but occasionally slackening it. 


Contractions of the Heart. 223 


The ventricular muscle has then no power to mark time, or to regain 
its regular times of contraction after they have been disturbed. It 
shows no tendency to compensate for the haste of one beat by delay- 
ing the next one; it shows the contrary tendency. 

Since compensation is not a function of the spontaneously beating 
muscle, the suggestion readily occurs that it may be a function of the 
ventricular ganglia. But the presence of ganglion cells in the piece 
of muscle does not change the result, as is shown by the following 
experiment: 

In place of the apex preparation, a piece of the dase of the ventricle 
was used. When a fresh base preparation was perfused, the results 
obtained upon electrical stimulation were identical with those ob- 
tained from the apex. No compensatory pause appeared; the actual 
pause after the extra contraction was variable, usually somewhat less 
than the regular pause between beats. 

The compensatory pause is, therefore, not a property of a sponta- 
neously beating portion of the dog’s ventricle, whether provided with 
ganglia or not. 

The relation of this to other facts about the compensatory pause, 
and to the explanation of that phenomenon, will be deferred to a 
later page. 


> 


STIMULATING EFFECT OF THE EXTRA CONTRACTION. 


The spontaneous beat following an extra contraction is, almost 
without exception, stronger than the preceding spontaneous beats. 


AI AAA 


FiGuRE 8.— Dog’s apex. Spontaneous and extra contractions. 


This fact has already been presented in some of the figures. It ap- 
pears also in Figs. 8 and g. Fig. 8 shows further that the earlier an 
extra contraction comes, the stronger is the following contraction; 
and Fig.9 shows that two extra contractions are apt to be followed by 
exceptionally strengthened beats. 


22 RR. S. Woodworth. 


These stateménts, again, are based not simply on a few selected 
tracings, but on the average of all cases, so far as measured.! The 
statistical results are as follows: 

1. The height of the contraction following the extra contraction 
averages 124.4 per cent of the height of the regular contraction pre- 
ceding. In some preparations, especially where the regular beat is 


ANAM AAAA 


Ficurr 9.—Dog’s apex. Spontaneous, and single and double extra contractions. 
oD J c= 


somewhat slow and weak, this average rises to 150 per cent; in some 
where the regular beat is very strong, it sinks to 110 per cent. 

2. To compare the effect of two extra contractions with that of 
one, a special test was made on one apex. The beat following a 
single extra contraction averaged 113.7 per cent of the regular beat; 
while the beat following a double extra contraction averaged 125.7 per 
cent.2. Three or more extra contractions were found to exert a 
stronger stimulant effect than two. 

3. The statistical evidence that the beat following the extra con- 
traction is the more strengthened, the earlier the extra contraction 
comes, is presented in the following table. The interval before the 
extra contraction is expressed as a fraction per cent of the regular 
period preceding it, and the height of the following contraction as a 


per cent of the height of the preceding regular beat.® 


Interval before extra. Av. height of following beat. 
=39 156.4% 
40-59%, 129.3 % 
60-79% 1136% 
80-114 % W123'% 


! These cases were selected solely with reference to the accuracy with which 
they could be measured, and with reference to the uniformity of the spontaneous 
beats. 

2 The “ probable error” of each of these determinations is about 1.4. 

% For greater certainty, the averages of the two preceding periods and heights 


were taken as the standards. 


Contractions of the Ffeart. 225 


The shorter the interval beforesthe extra contraction, the stronger 
is the following beat. The converse of this proposition is also true, 
as is seen on grouping the cases according to the height of the follow- 
ing beat, and finding for each group the average of the interval 
before the extra contraction. Thus: 


Height of 
following beat. 
Per cent. 


Interval Interval 
before extra. after extra. 
Per cent. Per cent. 


Number of 
cases. 


—99 sys 69.8 


100-119 5 oy Pe 92.0 
120-139 ¢ S25 $8.8 
140-159 ‘ : 120.5 
160-179 

180-199 

200- 


On comparing the first and third columns, it is seen that where the 
beat following the extra contraction is strong, there the extra con- 
traction has come early. This statement, together with its converse 
obtained from the previous table, establishes a dependence between 
the earliness of the extra contraction and the force of the following 
beat.! 

On comparing the first and last columns of the above table, a 
certain amount of connection appears between the strength of the fol- 
lowing beat and the length of the pause or interval after the extra 
contraction. But this connection seems to be comparatively loose. 
On the whole the stronger beats follow the longer pauses. This de-| 
pendence comes out clearly within the limits of short series of cases. 
In some series I made the extra contractions all equally early ; the 
pauses after them varied, and the height of the following beat was 
plainly seen to vary directly with the length of the pause. 

In order to determine with more certainty the degree to which the 
strength of the beat following the extra contraction was dependent on 


1 The re-grouping of the cases is not superfluous, since it diminishes the 
danger of founding a conclusion on a chance combination of cases. ‘lo rearrange 
the cases in this way may be as valuable as to add a considerable number of fresh 
cases. 


226 R. S. Woodworth. 


each of the two factors mentioned —the earliness of the extra con- 
traction, and the duration of the pause —I have had recourse to the 
mathematical treatment of correlation devised by Karl Pearson.’ I 
have calculated Pearson’s coefficients of correlation. On correlating 
the interval defore the extra contraction with the height of the follow- 
ing beat, the coefficient was found to be 0.64. On correlating the 
interval after the extra contraction with the height of the following 
beat, the coefficient was 0.45. To get some meaning from these 
numbers, it is sufficient to know that a coefficient of unity would mean 
absolutely perfect correspondence, while a coefficient of zero would 


mean a complete lack of correspondence. The coefficient 0.64 ac- . 


cordingly denotes a fairly high degree of correspondence, and the 
coefficient of 0.45 only a moderate correspondence. To return to 
physiological terms: the strength of the following beat is closely 
dependent on the earliness of the extra contraction, and somewhat 
loosely dependent on the length of the pause.” 

Statistical treatment thus confirms the observations made on 
typical cases, and removes all doubt as to whether those cases were 
really typical. This later treatment has brought again into view the 
same two factors in the determination of the force of contractions 
that were noted in a simpler form in the staircase contraction. The 
fact that the beat following an extra or hastened contraction is stronger 
than the preceding regular beat is a fresh example of the staircase 
effect. A long pause following an extra contraction also increases 
the force of the next beat. Usually, however, in the apex prepara- 
tion, the pause is not prolonged, and therefore the strengthening 
of the next beat must be for the most part a result of the hastened 
extra contraction. 

Before finally accepting this view, we ought to consider one other 
possibility. The strength of the beat following an extra contraction 
might be an after-effect of the weakness of that contraction. It has 
indeed been suggested that the following beat simply compensates 
for the weak extra contraction. Against this view must be placed 
the fact brought out earlier in this study, that a hastened contraction 
need not be subnormal in strength in order that the following con- 


1 A convenient reference is PEARSON’s Grammar of Science, 2d ed., 1900, p. 400. 
* It is important to recall here that the earlier the extra contraction comes, the 


shorter in general is the following pause, Yet an early extra contraction and 
a dong pause both strengthen the following beat. These two influences, though 
similar in effect, must be quite independent of each other. 


Contractions of the Fleart. 227 


traction be exaggerated. A consideration of such cases leaves no 
doubt that the hastening of the extra contraction is the principal 
factor in the result. 

The view that the great strength of the beat following the extra 
contraction is simply a compensation for the weakness of the extra 
contraction, must at all events be definitely abandoned. The follow- 
ing beat often more than compensates, and its strength is hence not 
accounted for in terms of compensation. Good examples of ove- 
compensation? are seen in Fig. 4. 

Over-compensation usually comes still more plainly into view when 
several beats following the extra contraction are observed. The 
excessive height often continues for a number of beats, as in Figs. 4, 


1 The correlation method does not help us much here, because the force of the 
extra contraction is itself so closely dependent on the interval that precedes it, 
that if one be closely related to the force of the following beat, so must the other 
be as well. In fact, the coefficients are so nearly the same, 0.64 and 0.66, that 
no inference can be drawn, except this, that if we, on other grounds, admit the 
dependence of the following beat on the earliness of the extra contraction, we 
are sure that the weakness of the extra contraction has no considerable further 
effect of its own. For if it had, it would show a considerably higher coefficient of 
correlation, due to the combination of two causes of correspondence. 

2 In order to obtain a measure of the over-compensation, it would be necessary 
to add the force of the extra contraction to the force of the following beat, and 
compare the sum with the combined force of two regular beats. Unfortunately, 
it is difficult if not impossible to get an accurate measure of the force of the extra 
contraction from such tracings as are seen in Fig. 5. The actual recorded rise 
of the writing point evidently does not represent the full force of the extra con- 
traction, since much of this force was consumed in overcoming the inertia of the 
descending lever and muscle. On the other hand, the total height of the extra 
contraction, measured from the base line, is often too great to represent the force, 
as in Fig. 6. Not being able to devise an accurate measure, I have nevertheless 
used the actual recorded rise of the extra contraction as a representative of its 
force. I chose this measure, because it was the most unfavorable to the result 
that seemed likely to appear; if then the result still appeared, it was certainly not 
manufactured. Using this measure, I found the sum of the extra contraction and 
the following beat to average 98 per cent of the sum of two regular beats: and 
when it is considered that the first extra contraction in Fig. 5 was counted as 
having but one-fourth of the force of a regular contraction, and that the second 
was counted as having no force at all, and further that the same deficiency is 
present in the majority of cases, it is certain that this 98 per cent should be in- 
creased to over 100 per cent, and indicate over-compensation. Ina large share 
of the separate cases, even my defective measure showed over-compensation 
amounting to 20, 50, 70, and even as high as 175 per cent of the force of two 
regular beats. The separate cases, without regard to the average, are enough to 
show that nothing like an accurate compensatory function is in play, and that 
something more powerful than compensation is acting. 


228 R. S. Woodworth. 


8, and 9. A comparison of these with Fig. 1 leaves little room for 
doubt that both represent the same fundamental fact, viz., the stimu- 
| lating influence of hastened contractions. 

In order to determine with certainty whether the extra contraction 
does act asa momentary stimulant of the heart muscle, it is necessary - 
to consider both the force and the rate of the following beats. If the 
force were increased but the rate slackened, the resultant effect might 
be that less force was exerted in a given time. In order to bring 
both force and time into a single expression, the total force exerted 
in a given time may be divided by the time, thus giving a measure 
of the force exerted per unit of time. This quotient gives, in other 
words, a measure of the rate at which available potential energy is 
developed by the muscle. No other way of measuring stimulation or 
depression of the cardiac muscle is so complete as this. 

In calculating this quotient, the interval preceding each beat should 
be counted as corresponding to that beat, since this is the time dur- 
ing which the energy discharged in that beat is accumulated. The 
intervals would most suitably be measured from the beginning of 
diastole to the end of the next systole, but as this measurement cannot 
be accurately made in my tracings, I have measured all intervals 
from the beginning of systole to the beginning of the next systole. 

When the force per unit of time is thus measured, the stimulating 
effect of the extra contraction appears very plainly. My meas- 
urements show the following average results: 

1. The force exerted by the first beat after the extra contraction 
amounts, per unit of time, to 142 per cent of the force exerted in the 
preceding regular contractions. In other words, during the diastole 
and pause succeeding the extra 
contraction, the rate of develop- 
ment of energy is nearly 1} times 
| the regular rate. 

ad-E ale 2. This stimulating effect of the 

Noof Contraction. | I extra contraction is not exhausted 
Ficure 10.— Force exerted per unit of by the first following beat, but 

time by each of the first 20 beats fol- persists for several beats, about 8 

lowing the extra contraction. Average or 9 on the average, This can be 

seen by reference to Fig. 10. 

3. As the same figure also shows, the stimulation is not followed 
by an equal depression. The force per unit of time returns to normal 
after about ten beats, and may then sink a little below normal, but not 


measurements. 


SO eet 


ml TY are 


Contractions of the Heart. 229 


much below. In some instances the stimulating effect of several 
successive extra contractions persists for many beats, indeed almost 
indefinitely. The stimulating action of an extra contraction is, there- 
fore, real and uncompensated, or at least not promptly compensated. 

4. The stimulating action of an extra contraction is the stronger, 
the earlier the preceding diastole is interrupted by the extra contrac- 
tion. This could be inferred from previous results, since it was found 
that the earlier an extra contraction came, the shorter was the follow- 
ing pause, and yet the stronger was the next beat. In order to get a 
measure of this relation, I have correlated the length of the interval 
before the extra contraction with the rate of development of potential 
energy during the following pause, measured by dividing the height 
of the following systole by the duration of the interval after the 
extra contraction. Both the interval before the extra contraction 
and the force per unit of time of the next beat are expressed as 
fractions per cent of the regular values obtained from the preceding 
beats. 


Force per unit 
Number of of time, exerted 
cases. by the next beat. 
Per cent. 


Interval 
before extra. 
Per cent. 


—39 : 168.9 


40-49 — 3 158.2 


50-59 
60-69 
70-79 
80- 


The shorter the interval before the extra contraction, then, the 
more is the development of energy accelerated immediately afterward. 
The more the extra contraction is hastened, the greater is its effect as 
a stimulant. 

The converse statement also holds good, as is seen by re-grouping 
the cases ; that is to say, where we find an especially strong stimula- 
tion resulting from an extra contraction, there we find the extra con- 
traction to have come especially early. 

That the connection between the earliness of an extra contraction 


230 R. S. Woodworth. 


and its stimulating effect is close appears also from Pearson’s coeffi- 
cient of correlation, which in this case is 0.73. 


Force per unit of 
time exerted by beat 
after extra contrac- 

tion. 
Ber cent: Per cent. 


Number of Interval before 
cases. extra. 


=99 69.5 
100-119 66.7 


120-139 | 58.1 


140-159 58.0 
160-179 
180-199 
200- 


Finally, in order to get back from numbers to the cardiac muscle, 
attention may be directed again to Fig. 3, in which the main results 
of my measurements find a concrete illustration. On comparing the 
spontaneous beats before and after the series of extra contractions, 
the stimulating effect of hastened contractions is clearly seen, and 
has obviously no resemblance to compensation. 

By running the eye along the series of extra contractions, the con- 
nection of this stimulating effect with the staircase phenomenon be- 
comes apparent, as well as the strengthening effect of a pause or 
prolonged diastole on the following contraction. No doubt can re- 
main of the reality of these factors, and of their adequacy to explain 
the variation in force of contraction that appears on electrical stimu- 
lation of the ventricular apex. 


II. EXPERIMENTS ON THE FROG’S HEART. 


As the experiments to be reported here were devised in the hope 
of clearing points still under dispute, I shall introduce them in 
the midst of a discussion of the literature. The results obtained 
in the apex and base preparations will also be found to have con- 


Contractions of the Fleart. 231 


siderable theoretical importange. I shall confine the discussion 
to a few questions, which the facts in hand would seem to settle 
definitely.! 


THE REFRACTORY PERIOD. 


Marey, who in 1876 discovered the refractory period of the nor- 
mally beating frog’s ventricle, gave the following account of it.2. The 
refractory period was in the main coincident with the early part of 
systole. During the relaxation or diastole, the ventricle responded to 
a stimulus by an extra contraction; during systole it gave no response 
whatever, either by an extra contraction or by increased strength of 
the systole then in progress. This refractory period was not, however, 
according to Marey, of constant length, but shrank as the stimulus 
was strengthened, being gradually reduced to the first instants of 
systole, and finally, on the application of fairly strong stimuli, disap- 
pearing altogether. The response to strong shocks applied early in 
systole consisted, not in any increased height of the contraction then 
in progress, but in an extra contraction, which began late in the fol- 
lowing diastole. The latent period of the extra contraction, which was 


1 The apparatus used for recording the movements of the frog’s heart was 
patterned more or less after_the ‘“ Fiihlhebel” used by KAIseR (Zeitschrift ftir 
Biologie, 1892, xxix, p. 208). The movements of the auricle or ventricle were 
first received by a vertical piece of grass straw. The cut end of the straw rested 
directly on the auricle or ventricle. A small distance from the end of the straw 
a little opening was made into its central cavity and filled with paraffin, thus 
making the straw air-tight and helping out adhesion by suction. This vertical 
straw was attached to the horizontal lever—also a piece of grass straw — by 
means of a pin which served as a pivot on which the vertical straw could turn 
freely. I did not find it necessary to provide a guide for the vertical straw. The 
same object — adhesion to the heart — was secured by pinning the apex of the 
ventricle to a thin plate of cork inserted beneath it. The pin served also as an 
electrode, the other pole being provided by another small pin close to the first, 
and either piercing the apex or fixed in the cork close beside it. When the auricle 
was to be stimulated, a pin was passed through the left margin of the left auricle 
into the cork beneath, and another was fixed in the cork in contact with the 
auricle. The pin through the ventricular apex had in this case no electrical con- 
nection. ‘The pins can be inserted with little loss of blood. 

On the whole, I do not regard this method as equal to that of suspension; it 
wears out. the heart more quickly. Still it gave satisfactory results, and enabled 
the stimulus to be very sharply localized. 

2 MAREY: Travaux du laboratoire, 1876, pp. 73 ff. 


250 R. S. Woodworth. 


but a small fraction of a second when the stimulus was applied in 
diastole, was sometimes a whole second when the stimulus came early 
in systole. 

This extreme length of the latent period was a suspicious circum- 
stance. Almost immediately Hildebrand! suggested that the extra 
contraction that followed a strong stimulus applied to the ventricle 
during systole was not a direct response of the ventricle, but was due 
to a leakage of current to the auricle. Since, during systole of the 
ventricle, the auricle is in diastole and therefore irritable, leakage 
would, if strong enough, call forth an extra contraction of the auricle, 
and this, in the regular progress of the wave of contraction, would be 
followed by an extra systole of the ventricle. 

This explanation has so much in its favor that it has won the assent 
of many special students of the subject,? yet the description and trac- 
ings of Marey have remained classic. 

All the evidence is in favor of Hildebrand’s suggestion; it may be 
summarized as follows: 

(1) In Marey’s experiments the stimulating electrodes were appar- 
ently applied to the base of the ventricle, so that spread of current to 
the auricle was very probable. Hildebrand* and later Engelmann 4 
found that when the electrodes were applied near the apex, much 
stronger shocks than those employed by Marey were necessary in 
order to get an extra contraction by stimulation during systole. I 
have confirmed this result. 

(2) The extra contraction elicited by stimulation during systole 
comes at the same time and gives the same ventricular tracing as 
when the auricle is stimulated during ventricular systole. In other 
words, the result obtained is exactly what the hypothesis of Hilde- 
brand would require. This is strong circumstantial evidence. 

(3) The direct evidence is even clearer. On recording simulta- 
neously the movements of the auricle and of the ventricle, it is found 
that the first effect of the strong stimulus applied to the ventricle 
during its systole is actually an extra contraction of the auricle. This 
is then followed as usual, and at the usual interval, by an extra con- 
traction of the ventricle. These facts are seen in Fig. 11, which is 


' HILDEBRAND: Nordiskt medicinskt Archiv, 1877, ix. See also LovEN, 
Mittheilungen vom physiologischen Laboratorium in Stockholm, 1886, i, p. 4. 

* As, e.g. ENGELMANN: Archiv fiir die gesammte Physiologie, 1895, lix, pp. 
316-321; KAISER: Zeitschrift fiir Biologie, 1892, xxix, pp. 215, 217. 

* HILDEBRAND: OP. cit. 4 ENGELMANN: Of. cit. 


1% 


ee ee Mah 4 pa > 


Pe. 


eg es ers 


Contractions of the Heart. 233 
similar to some given by Engelmann,! but is easier to follow, because 
auricle and ventricle are recorded separately.” 

(4) It only remains to make sure that the extra contraction of the 
auricle, that occurs on exciting the ventricle during its systole with 
strong shocks, is actually due to escape of current. This is demon- 
strated by the following simple experiment. 

A frog’s heart is excised and laid on a slab of cork; pin electrodes 
are fixed in the ventricular apex. The strength of current is deter- 
mined that is just sufficient to cause an extra contraction of the 


5? <r, Sa eae eB = 

FIGURE 1].— Frog’s heart. Record of auricle above and of ventricle below. Read 
from left to right. Time in seconds. Primary current, 6 volts. Distance of sec- 
ondary coil, 0. In the record of stimuli, down means make; up, break. Stimulus 
applied to apex of ventricle. 

The first and third shocks fell during the irritable period of the ventricle, and 
called out direct ventricular response. The third fell also during the irritable period 
of the auricle, but, being a make, was not strong enough to cause by leakage an 
auricular extra contraction. The second and fourth shocks (breaks) fell during the 
refractory period of the ventricle, but called out extra contractions of the auricle, 
which were followed by extra contractions of the ventricle. 


auricle (and ventricle) when applied to the apex at any particular 
stage of auricular diastole. Then, with a sharp knife, auricle and ven- 
tricle are completely severed, but left in contact in the same position 
as before. Now the same current, applied to the apex at the same 


1 ENGELMANN: OP. cit., pp. 320, 321. . 

2 BRUNTON and CASH: (Proceedings of the Royal Society of London, 1883, 
Xxxv, p. 460) also publish a tracing, which, though not intended for this demon- 
stration, is as appropriate as could be wished. The tracings in their paper are full 
of examples of the leakage of current from one chamber of the heart to another, 
though the authors take scant notice of this source of strange results. Curiously 
enough, after they have once remarked that “these results may be due, in part 
at least, to escape of current,” they take so little further account of this possibility 
as to conclude from the occasional prompt response of the ventricle, without long 
latent period, to stimulus applied to the auricle, that the impulse to contraction 
is conducted by means of nerves, and that, therefore, besides the regular muscular 
conduction of impulses in the heart, there must be a nervous conduction. Much 
more probable, since it was only strong currents that called out the prompt response 
of distant parts of the heart, is the supposition that the conduction was purely 
electrical. 


234 R. S. Woodworth. 


stage of auricular diastole as before, calls out still an extra contraction 
of the auricle. (Naturally, no contraction of the ventricle here fol- 
lows that of the auricle.) Since in this preparation no physiological 
conduction of any sort remains between auricle and ventricle, we can 
fall back on no other hypothesis but that of escaped current. And 
since the same strength of current gives the same auricular response, 
either with or without physiological connection, we must suppose 
leakage to be the cause in both cases. 

A similar experiment shows that the disappearance of the refractory 
period that Marey observed on warming the heart, is likewise the re- 
sult of escape of current to the auricle. For on watching the auricle, 
it is seen that warming it enables it to respond toa somewhat weak cur- 
rent applied to the apex; and the same if auricle and ventricle are 
severed. What warming accomplishes is an increased irritability of 
the auricle, so that the leakage of feebler currents is effective. 

Hildebrand’s suggestion is thus demonstrated. The extra contrac- 
tion of the ventricle that is obtained by stimulating it strongly during 
systole is not a direct response to the electrical shock. There is 
nothing to indicate that the ventricle is any less than absolutely re- 
fractory during systole. The dog’s apex, as we saw, is absolutely re- 
fractory during that phase. The end of the refractory period, in the 
dog’s apex, was marked by the very summit of the myogram; and 
practically the same limit has been assigned by Engelmann! to the 
refractory period of the frog’s heart, and by Cushny and Matthews? 
to that of the intact mammalian heart. In short, cardiac muscle is, 
during systole, entirely unirritable. 

The refractory period is clearly of a piece with the law of ‘all or 
none;” both point to the same peculiarity of cardiac muscle. The 
existence of the systolic refractory period means that the spontaneous 
beats, as well as electrically excited contractions, are maximal.® The 
lobster’s heart, which does not show the law of all or none, also fails 
to show any refractory period! The existence of an absolutely re- 


1 ENGELMANN: Archiv fiir die gesammte Physiologie, 1895, lix, p. 315. 

* CusHny and MATTHEWS: Journal of physiology, 1897, xxi, p. 219. Also 
GLEY: Archives de physiologie, 1889, p. 503. 

% This view has been well expressed by KAISER (Zeitschrift fiir Biologie, 1892, 
xxix, p. 218), and by CusHNy and MATTHEWS ((oc. cit.). 

* Hunt, BOOKMAN, and TIERNEY: Centralblatt fiir Physiologie, 1897, xi, 
pp. 275 ff., especially Figs. 1 and 7. Compare also KAISER’s results with frog’s 
hearts treated with muscarin, Zeitschrift fiir Biologie, 1892, xxix, p. 219. 


"we 


ee 


Contractions of the Heart. 235 


fractory period during systole could be inferred and predicted from 
the law of maximal contraction, wherever that law holds. 

Not to be predicted from the law of “all or none,” however, is 
the fact that a stimulus applied during systole has no effect on the 
irritability or on the development of contractile energy, during the 
following diastole. As far as concerns the irritability, this statement 
is demonstrated by Fig. 12. If a shock applied during systole raised 
the irritability of the muscle during the following diastole, then a series 
of shocks begun in systole and lasting on into diastole should pro- 
duce an extra contraction earlier than it could be produced by a single 
shock applied in diastole. But no such result appears; the single 


2S aS oe Ne a or, ee 


rept woul 


FiGure 12.— Stimulus and record of frog’s ventricle. Read from right to left. ‘lime in 
seconds. Primary current, 2.2 volts. Distance of secondary coil, 100 mm. Down 
= make, up = break. Rate of faradization, 34 make-breaks per second. The upper 
curve was traced about a minute and a half after the lower. 


shock is as efficient as the series. I have varied this experiment by 
using a single strong shock during systole and following it by a weak 
shock in diastole; the former had no influence on the efficiency of the 
latter. A shock that reaches the muscle during the refractory period 
leaves no appreciable after-effect on the irritability. 

It is the same with contractility! A stimulus applied during the 
refractory period has no effect on the force of the next following con- 
traction. I have applied strong shocks to the frog’s ventricle, and 
also to the dog’s apex, during each successive systole for a number of 
beats, without affecting the height of contractions in the least. The 
development of available energy is not affected by a stimulus during 
the refractory period. Such a stimulus is in every respect as devoid 
of effect as if it had not occurred. 

The account of the irritability during the cardiac cycle would not 
be complete if we spoke of the systole as a refractory period, and op- 
posed to it diastole and pause as an irritable period. For though no 


1 Cf. ENGELMANN: Archiv fiir die gesammte Physiologie, 1896, Ixv, p. 139. 


226 R. S. Woodworth. 


2) 


portion of this latter period is absolutely refractory, yet, at its begin- 
ning, only strong stimuli take effect, whereas later on in diastole weak 
stimuli are sufficient. The irritability increases gradually from the 
beginning of diastole on. Bowditch! showed that even 3, 4, or 5 sec- 
onds after the commencement of diastole, the frog’s apex had not yet 
reached its maximum of irritability. The rapid gain during the early 
part of diastole has been demonstrated by Engelmann? in the frog, 
by McWilliam * and by Cushny and Matthews# in the intact dog’s 
heart; and I have found the same in the mammalian apex prepa- 
ration, both after spontaneous and after electrically excited contrac- 
tions. The stronger the stimulus, the earlier in diastole can it be 
successfully applied. 

It is inaccurate to speak of the refractory period as of variable 
length, extending on into diastole when the stimulus is weak. The 
refractoriness that obtains during systole, and the low irritability early 
in diastole, are two quite different phenomena. The one is absolute, 
the other relative. The one is a corollary of the law of “all or none,” 
the other has no such dependence. The low irritability during re- 
laxation appears in muscular tissue that shows no law of “all 
or none” and no refractory period during contraction, — such, for 
example, as the smooth muscle of the frog’s stomach.® 

The refractory period, being peculiar to cardiac muscle, is signifi- 
cant of its fundamental properties. But the low irritability early in 
diastole, though not specially significant in this respect, is the more 
important fact in explaining certain other peculiarities of the muscle, 
such as its response to a tetanizing current by a series of rapid beats 
(see above, page 216), or such as the compensatory pause. 


THE COMPENSATORY PAUSE. 


The first fact to be borne in mind in attempting to explain the 
compensatory pause is that such a pause does not follow an extra 
contraction of a spontancously beating piece of the heart. Whether 
the piece be from the apex or base of the dog’s ventricle, as in my 


' Bowpitcu: Arbeiten aus der physiologischen Anstalt zu Leipzig, 1871, pp. 150, 
151. This result has been confirmed by OuRWALL: Skandinavisches Archiv fiir 
Physiologie, 1898, viii, pp. 37-40. 

* ENGELMANN: Archiv ftir die gesammte Physiologie, 1895, lix, p. 315. 

* MCWILLIAM: Journal of physiology, 1888, ix, p. 170. 

* CusHny and MATTHEWS: Journal of physiology, 1897, xxi, p. 215. 


5 See WoopwortH, This journal, 1899, iii, p. 39. 


Contractions of the Fleart. 37 


experiments, or from the sinus! or great veins” of the frog, whether 
it contain ganglia or not, the result is the same: no compensatory 
pause follows the extra contraction. In some cases, the pause after 
the extra contraction is somewhat prolonged, but it has not the 
accurate compensating Jength observed in the intact heart; usually 
it is much too short to compensate, and in the perfused dog’s apex it 
averaged less than the regular pause. 

A similar fact is that a piece of the heart beating in response to a 
constant stimulus does not show the compensatory pause. This again 
is true without regard to the presence or absence of ganglia.® 

It is not safe to infer from these two facts that the internal stimulus 
producing the spontaneous contractions is constant. What the two 
cases have certainly in common is this: in both, the rhythm of a 
piece of heart is developed within that piece. 

When, on the contrary, the rhythm of a piece of heart is impressed 
upon it by rhythmic stimuli from beyond, then an extra contraction is 
followed by a compensatory pause. The same piece of muscle, or of 
muscle and ganglia, which showed no compensatory pause when beat- 
ing spontaneously or in response to a constant stimulus, will show the 
pause when beating in response to rhythmic stimuli. 

As an example of this, an experiment of Engelmann‘ may be 


1 TIGERSTEDT and STROMBERG: Mittheilungen vom physiologischen Labora- 
torium in Stockholm, 1888, pp. 31, 32. 

2 ENGELMANN: Archiv fiir die gesammte Physiologie, 1896, Ixv, p. 137, p. 140. 

3 ENGELMANN: Archiv fiir die gesammte Physiologie, 1895, lix, pp. 328, 329. 
The statement made by LANGENDORFF (Archiv fiir Physiologie, 1884, Supplement- 
Band, p. 59; and 1885, p. 285) and repeated by KAISER (Zeitschrift fiir Biologie, 
1895, xxxii, p. 457), that during constant stimulation of the entire but isolated frog’s 
ventricle an extra contraction is followed by a compensatory pause, is, I am con- 
vinced, a mistake, the cause of which is not far to seek. In order to avoid injury 
to the ganglia in the base of the ventricle, the cut or clamp by which the ventricle 
was isolated would naturally be located a little a@éove the auriculo-ventricular 
groove. From the authors’ descriptions, it is clear that this was at least some- 
times the case. The constant chemical or mechanical stimulus, which was applied 
at the base of the ventricle, acted directly on the small remnant of the auricle. 
The pulsation of the ventricle would then be aroused by the auricle, and not 
directly by the artificial stimulus. The same remark applies to the spontaneous 
contractions which the authors sometimes observed. They probably originated 
in the auricle. This interpretation has at least sufficient probability to make 
LANGENDORFF’S and KAISER’s result of small weight as compared with the mass 
of facts on the other side. 

I am under the impression of having read this criticism in one of ENGELMANN’S 
papers, but cannot now find the passage. 

4 ENGELMANN: Archiv fiir die gesammte Physiologie, 1895, lix, p. 326. 


238 R. S. Woodworth. 


quoted. He excited a quiescent frog’s apex’ to rhythmical contraction 
by means of stimuli succeeding each other at about the rate of the 
normal heart beat. He then secured an extra contraction by interpo- 
lating an extra stimulus. After the extra contraction, one of the 
periodic beats failed to appear, and the apex remained at rest until 
the next of the rhythmic stimuli reached it. 

The result of this simple experiment of Engelmann could have 
been foretold with certainty. It shows nothing new regarding the 
properties of the cardiac muscle. But it does serve to throw into 
relief the conditions necessary for the development of a compensatory 
pause. These are, first, the low irritability at the beginning of 
the diastole, which prevents a stimulus following quickly upon the 
extra contraction from taking effect, and thus cuts out one regular 
beat ; and second, the absence of spontaneous contractions or of a 
constant stimulus. 

I have repeated Engelmann’s experiment on the mammalian apex, 
with one modification which makes the conditions approach more 
nearly to those of the normally beating ventricle. I took an irritable 
but non-spontaneous piece of apex, cut it nearly in two, and tied a 
ligature, not too tightly, about the narrow bridge that connected the 
two portions. On applying a stimulus to one portion, I obtained first 
a contraction of this portion, and then, after a perceptible interval, 
a contraction of the other portion.) I then obtained a series of such 
contractions, and into this series interpolated, by stimulation of the 
second portion, an extra contraction of that portion. The next 
periodic contraction failed to spread across the bridge into the second 
portion, the latter remained quiet till the following periodic con- 
traction occurred, thus showing a compensatory pause. 

This latter experiment is a very close counterfeit of what doubtless 
occurs in the normally beating heart after an extra contraction. We 
have abundant reason, from the work of Gaskell and of Engelmann, 
to believe that the regular beats of the ventricle are aroused by as 
many waves of excitation that come down from the auricle (and 
ultimately from the sinus and great veins). The low irritability 
immediately after the extra contraction prevents the ventricle from 
responding to one auricular stimulus; and as the ventricle is not 


! So far, this is simply a repetition on the mammalian ventricle of a familiar 
experiment of GASKELL on the cardiac muscle of amphibians. See Journal of 
physiology, 1883, iv, p. 64, and ScnArer’s Text-book of Physiology, 1900, ii, pp. 
180, 182. 


Contractions of the Fleart. 239 


beating spontaneously, and is not subject to a constant stimulus, 
it remains quiet till the next auricular stimulus, and so shows a com- 
pensatory pause. 

One or two more experiments throw light on the compensatory 
pause. ‘The first concerns the irritability of the heart muscle during 
the pause. Engelmann! conjectured that the curve of returning 
irritability after the extra contraction would run about the same 
course as after a regular contraction; possibly, he thought, the 
return of irritability would be somewhat slower after the extra con- 
traction. Cushny and Matthews,” working on the intact mammalian 
heart, found this slower return of irritability present to a certain 
degree, though not toa sufficient degree to account for the full length 
of the compensatory pause. In my preparations, I have not found 
it present at all. A given strength of stimulus can elicit an extra 
contraction just as early in the diastole of an extra contraction as in 
that of a regular contraction, or even a little earlier. The easiest 
way of testing this matter is by applying a faradic current during 
a regular contraction, and continuing it during the extra contraction 
until a second extra contraction arises. The result appears in Fig. 9 
(dog’s apex), and in exactly similar tracings from the intact frog’s 
ventricle. 

The prolonged inactivity of the ventricle after an extra contraction 
is not due toa lack of contractility nor to a lack of adequate irrita- 
bility, but simply to the lack of a stimulus. The compensatory pause 
is not a period of inhibition, but a period of waiting. If a constant 
stimulus of only moderate strength were present, it would take 
effect before the end of the compensatory pause. If the ventricle 
were beating spontaneously, it would usually develop within itself 
an adequate stimulus before the end of that time. 

Another valuable experiment consists in exciting two or more extra 
contractions, one following immediately after the other, and studying 
the effect on the pause. After Engelmann’s® work on this line, there 
would be little excuse for returning to it, had not Kaiser* confused 
the matter by an incomplete and misleading account, accompa- 
nied by tracings which, taken alone, seem to support an incorrect 
view. 


ENGELMANN: Archiv fiir die gesammte Physiologie, 1895, lix, p. 328. 
CusHny and MATTHEWS: Journal of physiology, 1897, xxi, p. 220. 
ENGELMANN: Archiv fiir die gesammte Physiologie, 1895, lix, pp. 330 ff. 
KAISER: Zeitschrift fiir Biologie, 1895, xxxii, p. 455. 


me © BD 


240 R. S. Woodworth. 


Kaiser’s statement is that the length of the pause increases with 
the number of consecutive extra contractions. A glance at Figs. 13 
and 14 will show that this is at least not always true. I have never 
found it true, either in the dog’s apex or in the intact frog’s ventricle. 


ube 

FiGurE 13.— Frog’s heart. Stimulation and record of ventricle. Three-fifths the origi- 
nal size. Read from right to left. Time in seconds. Primary circuit, 1 Daniell 
Distance of secondary coil, 80 mm. 


Kaiser himself admits that he found it true only within limits, 
but he makes light of the limits. Kaiser also asserts! that the pause 
following two extra contractions will be the longer, the more quickly 
they come, because, according to his views, their inhibitory after-effect 
will thus be more completely summated. Fig. 13 puts this matter 
straight. Two extra contractions are likely to cut out two regular 
beats, and the pause after them may then be long as compared with 
the ordinary compensatory pause, and it will be longer the earlier the 
extra contractions come. But it is possible to make two extra con- 


—> 


a z ri fre Ma ae Et are Se 


= ae \ fh MP ate 2 ar te Se a x3 5 wate Wane oe A 
FIGURE 14.— Frog’s heart. Stimulus applied to ventricle. Record of auricle above, of 
ventricle below. Time in seconds. The moments of stimulation indicated by the 


electromagnet needed to be displaced to the right, and the true moments are indicated 
by short vertical scratches across the magnet tracing. 


tractions come so early that only one regular beat shall be cut out, 
the irritability being already sufficiently restored, after the second 
extra contraction, to permit the second auricular stimulus to take 
effect. In this case, the pause is even shorter than the regular pause 
between beats. 

The only rule governing the length of the pause after multiple 
extra contractions, is that given by Engelmann,” namely, that the first 


1 KAISER: Of. cit., p, 456. 2? ENGELMANN: Loc. cit. 


——s 


Contractions of the Fleart. 241 


regular beat after the pause comes at one of the regular beat periods. 
This fact, and the reason for it, are well illustrated in Fig. 14. Each 
regular beat of the ventricle is there the sequel toa beat of the auricle. 
During the stimulation and extra contractions of the ventricle, the 
auricular beat goes on undisturbed, but has no effect on the ventricle. 
When the extra contractions have ceased, the auricular beat has once 
more its ventricular sequel, and both chambers go on beating as if 
nothing had happened. 

Viewed in this way, the restoration of the regular times of beat, 
after an extra contraction, is a simple matter and not very funda- 
mental. When the times of the ventricular beat are disturbed, the 
ventricle does not of itself regain them, but waits passively on the 
auricle, whose beats, in such a case as Fig. 14, have not been dis- 
turbed. If the auricle has been disturbed, it can wait passively on 
the coming of the rhythmic stimuli from the sinus. But if the beat of 
the sinus is disturbed by an extra contraction, there is no compen- 
satory pause and no restoration of the regular times of beat;! the 
situation is here the same as in the spontaneously beating dog’s apex: 
the extra contraction affects the place of origin of the rhythmic 
stimuli and disturbs their sequence. Whenever the regular times of 
beat are restored after an extra contraction, it is because the dis- 
turbance has not affected the source of the normal rhythmic stimuli. 
The “Law of the Preservation of the Physiological Moments of 
Stimulation,’ ? as Engelmann terms it, is not, therefore, a fundamental 
property of the heart muscle or of the ganglia; it is not worthy to be 
classed with the law of maximal contractions, or with the low irrita- 
bility during diastole, or with the spread of the excitation wave from 
sinus, through auricle, to ventricle; it is simply a result of these 
fundamental properties of cardiac muscle. 

This view of the compensatory pause to which our summary of 
facts has led us is the view of Engelmann and also that of Gaskell.® 
The restoration of the regular moments of beat after one or more 
extra contractions is so readily and completely explained by Gaskell’s 
theory of the heart beat as to constitute a strong argument in favor 
of that theory. 

1 TIGERSTEDT and STROMBERG: Mittheilungen vom physiologischen Labora- 
torium in Stockholm, 1888, pp. 31, 32; ENGELMANN: Archiv fiir die gesammte 
Physiologie, 1896, Ixv, p. 137. 

2 “Gesetz der Erhaltung der physiologischen Reizperiode.” ENGELMANN: 


Archiv fiir die gesammte Physiologie, 1895, lix, p. 333- 
3 GASKELL in SCHAFER’S Text-book of physiology, 1900, ii, p. 191. 


242 R. S. Woodworth. 


A new fact in support of that theory is the absence of a compen- 
satory pause in case of a spontaneously beating piece of the ventricle, 
whether from the apex or from the base. Since ventricular muscle 
does not show the pause when its beats are certainly spontaneous, 
the probability is that in the normal heart beat, when the ventricle 
shows the compensatory pause, the reason is that the ventricle is not 
then spontaneous. 

The facts brought together in the preceding pages enable us to 
throw out at once all other vievs of the compensatory pause that have 
been suggested. 

In opposition to the view! that the ganglia are necessary for the 
production of the pause, we have the fact that, given like conditions 
of stimulation or of spontaneity, the presence or absence of ganglia 
makes no difference in the result. 

In opposition to the view that the pause is some sort of a compen- 
satory” or inhibitory ® after-effect, we have the fact that two or more 
extra contractions are not asa rule followed by a longer pause than 
one, and not necessarily by any prolonged pause at all; and also the 
fact that tests of the irritability and contractility during the pause 
betray little or no sign of inhibition. 

The theory of the pause that regards it as an inhibitory after-effect 
of the extra contraction has been elaborately worked out by Kaiser.* 
He regards the regular diastole also as an inhibitory after-effect of 
systole. The stronger the cause, his argument runs, the stronger 
must be the effect; and hence an extra contraction, following quickly 
after a regular beat, must, by summation, produce more than the 
usual diastoleand pause. An excellent opportunity of testing Kaiser’s 
theory in general is afforded by the contraction that follows the 
compensatory pause. This contraction is usually stronger than the 
regular systole, and should therefore, according to the theory, be 
followed by an especially complete diastole and a prolonged pause. 
This, however, is never the case. On the contrary, the diastole is 
practically always incomplete or the pause shortened. When the 
extra contraction is followed by an alternation of strong and weak 
beats, or in any other case of such alternation, the strong beats are 
regularly followed by less diastole and pause than the weak beats. 


' DAsTRE: Journal de Vanatomie et de la physiologie, 1882, pp. 464, 465. 
KAISER: Zeitschrift fiir Biologie, 1895, xxxii, p. 446. 

* Marey: Travaux du laboratoire, 1876, p. 74. 

8 KAISER: Loc. cit. * KAISER: Loc. cit. 


Contractions of the Freart. 243 


On the basis of my tracings (e.g. Figs. 3, 5, 13), as well as of all 
others I have seen, including those of Kaiser,! I can most definitely 
deny that the pause following a contraction shows any tendency to 
be proportional to it. The conclusion is, therefore, not only that the 
compensatory pause is no inhibitory after-effect of the extra contrac- 
tion, but that diastole, in general, is not an inhibitory after-effect of 
systole. 

To sum up: The compensatory pause is not the expression of any 
compensatory function exerted by cardiac muscle or ganglia; and it 
is not a period of inhibition; it is simply a period of waiting for a 
stimulus. The conditions necessary for its appearance are: 

1. The absence of spontaneous contractions from the portion of 
the heart that shows the pause; and the absence of any constant 
stimulus directly affecting that portion. 

2. The excitation of that portion by rhythmic stimuli coming from 
beyond itself. 

3. The refractory period occupying the systole of the extra contrac- 
tion and the low irritability early in its diastole; one or another of 
these causes prevents one of the rhythmic stimuli from taking effect 
after an extra contraction. 

4. The freedom of the sinus, or in general of the portion of the 
heart muscle in which the rhythmic stimuli originate, from disturb- 
ance by the artificial stimulus or by the extra contraction. 


THe ForRcE OF CONTRACTION. 


Two facts demand an explanation: 

I. The extra contraction is always weak in comparison with the 
regular beat, and it is weaker the more closely it follows the preced- 
ing regular beat. 

2. The beat after the extra contraction and pause is stronger than 
the regular beat. 

The first of these facts is visible in every tracing from Marey down. 
The second, for some reason, does not appear in the tracings of 
some authors. It was first noted by Langendorff? in the frog’s heart, 
and has been remarked also by Kaiser? and by Bottazzi* in batra- 


1 KAISER: Zeitschrift fiir Biologie, 1892, xxix, Taf. IV; 1895, xxxii, Taf. VI. 
2? LANGENDORFF: Archiv fiir Physiologie, 1885 p. 287. 

3 KAISER: Zeitschrift fiir Biologie, 1892, xxix, p. 216. 

* Bottazzi1: Centralblatt fiir Physiologie, 189, x, pp. 403, 404. 


244 R. S. Woodworth. 


chians, by Bottazzi! in the embryo chick, and in mammals by McWil- 
liam,? Gley,? Langendorff, and Cushny and Matthews.’ My tracings 
show it in the frog’s heart, and in the apex and base preparations 
from the dog’s ventricle. 

In regard to the weakness of the extra contraction, there can be no 
doubt of the correctness of McWilliam’s ® suggestion, that insufficient 
time has elapsed since the end of the preceding systole to permit the 
accumulation of the regular amount of contractile force. On account 
of the maximal character of the contractions of cardiac muscle, no 
contractile force, no available potential energy, is present at the com- 
mencement of diastole; and it is only gradually accumulated. It 
continues to accumulate throughout the diastole and pause, as is seen 
by the increasing height of the extra contraction. 

To explain the great force of the contraction following a compen- 
satory pause, we naturally push the preceding explanation a little fur- 
ther.’ Since the contractility accumulates throughout the regular 
pause, it would very likely continue to accumulate if the pause were 
further prolonged, as is the case after an extra contraction. And, in 
fact, the longer the compensatory pause, the stronger is the following 
contraction. 

In the earlier part of this paper, the length of the pause was proved 
to be a genuine factor in the production of a strong contraction. This 
factor alone, however, was found insufficient to account for the phe- 
nomenon in the apex and base preparations, since in them an extra 
contraction was not usually followed by a prolonged pause, and yet 
the next contraction was always exaggerated. The pause after an 
extra contraction was even found to average less than the regular 
pause, yet the following contraction averaged more than the regular 
contraction. 

Another factor also was shown to be effective in increasing the 
height of a contraction. The hastening of a contraction had a stimu- 
lating effect on the force of the next beat or of several following beats. 
This was interpreted as being a form of the “ staircase” phenomenon. 


1 Borrazzi: Archives italiennes de biologie, 1897, xxvii, pp. 121-123. 
2 McWILLIAM, Journal of physiology, 1888, ix, p. 171. 
8’ GLey: Archives de physiologie, 1889, p. 505. 
4 LANGENDORFF: Archiv fiir die gesammte Physiologie, 1895, Ixi, p. 317; and 
1898, Ixx, p. 473. 

6 Cusuny and MATTHEWs, Journal of physiology, 1897, xxi, 216. 

6 MCWILLIAM: Of. cit., p. 170. 


1 MCWILLIAM: Of. cit., p. 171. 


Contractions of the Fleart. 


The more the extra contraction was hastened, 
the stronger was the stimulating effect. 

In order to look for this factor in the beat of 
the intact frog’s heart, we must get rid of the 
compensatory pause after the hastened con- 
traction. This can be done by interpolating 
a second extra contraction. The result can be 
seen in Fig. 13, and, with a greater variety of 
cases, in Fig. 15. The effect of the hastened 
contraction, without the pause, is seen to be 
practically nothing. The only visible factor in 
determining the force of a contraction is the 
length of the preceding pause. 

Here, then, we encounter a wide divergence 
of results as between the frog’s heart and the 
perfused dog’s apex (or base). In the frog’s 
heart, the length of the pause controls the 
force of the following contraction, and the 
after-effect of a hastened contraction is not 
visible. In the dog’s apex, the pause has in- 
deed its effect, but this is small in compar- 
ison with the effect of hastening a beat. 

The cause of this difference can, I believe, be 
assigned. The dog’s apex preparation almost 
always beats at a rate slower than its “optimal 
rate,” as is proved by the fact that artificially 
hastening its rhythm, by a series of shocks, 
gives rise to a ‘‘staircase.” The intact frog’s 
ventricle, on the contrary, usually beats faster 
than its optimal rate; for artificially hastening 
its rate diminishes the force of its beat, whereas 
artificially slowing its rate increases the force. 
Artificial slowing, free from vagus inhibition, 
can be accomplished in the following way: It 
is known! that stimulation of the auricle, very 
early in its diastole, causes an extra contraction 
that does not spread to the ventricle. The 


1 DASTRE: Journal de l’anatomie et de la physiolo- 
gie, 1882, p. 464; KAISER: Zeitschrift fiir Biologie, 
1892, Xxix, p. 217. 


A ITGLSAWAGTATLUAWARLATAVATATACAATIWLPSOABLOIATALOLIAUL 


SAR ap oe am CEES Ra ao, he. 


is) 
+ 
ou 


The little elevations between 


= break. 


Down = make, up 


‘Time in seconds. 


FiGuRE 15.— Stimulation of the ventricle of the frog’s heart. 


the large ventricular curves record the beats of the auricle. 


246 Re. S. Woodworth. 


auricle, and the ventricle with it, then miss one of the regular beats. 
If the auricle is again stimulated soon after its next regular beat, and 
so on, the result is that the ventricle beats at half its former rate, 
and that its beats are stronger. In one heart on which I tried this 
experiment, this form of ventricular beat, once started, kept on of 
itself, evidently because the prolonged ventricular contractions cut 
out every second auricular excitation. In this case, too, the slowed 
beat was stronger than the regular beat. McWilliam! has found 
that the mammalian ventricle, also, beats more strongly when the 
rhythm of the heart is artificially slowed without vagus inhibition. 

Now since the ventricle of the intact heart beats faster than its 
optimal rate, the force of any contraction is more increased by pro- 
longation of the preceding pause than by hastening the preceding 
contraction. And the contrary is the case in the perfused apex, 
because its spontaneous beats are slower than their optimal rate. 

That a spontaneously beating portion of cardiac muscle should 
assume a rate slower than its optimum seems a pregnant fact; as yet, 
however, it stands too much alone to warrant an hypothesis.” 

Besides the two factors mentioned above, several others have been 
suggested by different authors as determining the force of the beat 
following the extra contraction. The view of Gley,® that the main 
factor is the distention of the ventricle by a double supply of blood, 
and the view of Bottazzi* that a positive after-effect of vagus stimula- 
tion comes into play, are neither of them applicable to the apex 
preparation. 

Langendorff ® regards the great force of the beat following the extra 
contraction as simply a compensation for the weakness of the extra 
contraction, and as having nothing to do with the length of the inter- 
vening pause. His view seems to be that the heart has some com- 
pensatory power, by virtue of which it necessarily follows up a weak 
contraction by a contraction that is correspondingly strong. He 
offers no evidence in support of his statement that the length of the 


t MCWILLIAM: OD. cié., p. 171. 

* I have used the term “optimum rate” to mean such a rate as will make the 
single beats the strongest. The sus total of force exerted by cardiac muscle 
would probably be greatest at a faster rate than this, since we found the develop- 
ment of available energy to proceed fastest after the most hastened extra con- 
tractions. 

* GLEy: Archives de physiologie, 1889, p. 505. 

* BorTazzi: Centralblatt fiir Physiologie, 1896, x, p. 403. 

5 LANGENDORFF: Loc. cit. 


Contractions of the Heart. 247 


pause has nothing to do with the matter, and that statement has been 
abundantly disproved above, both for the frog’s ventricle and for the 
dog’s apex. It was also found that in the mammalian preparations 
the effect cannot be called compensation, but is rather actual stimula- 
tion. 

Throughout this study, I have been unable to detect anything in 
the nature of an active compensatory function. All the facts, new 
and old, that have here been collected, tend to lessen the importance 
of the doctrine of compensation. Such compensation as occurs is 
merely incidental to the operation of the fundamental properties of 
cardiac muscle. 


SUMMARY. 


The perfused apex of the dog’s ventricle shows the law of maximal 
contractions, or ‘all or none.” 

It shows also the refractory period, which extends throughout the 
systole of all contractions, and is not diminished in length by greatly 
strengthening the stimulus. 

It shows also a gradual increase in irritability during diastole. 


1 LANGENDORFF'’S view leaves altogether out of account the law of “all or 
none,” as applied to the extra contraction. This mistake can be seen in the 
following statement of his view (LANGENDORFF: Archiv fiir die gesammte 
Physiologie, 1898, Ixx, pp. 480, .481): 

“Kehren wir zur compensatorischen Systole [by which name he calls the 
strengthened beat following the extra contraction] des Ventrikels zuriick. Ver- 
muthlich ist ihre gréssere Starke dadurch bedingt, dass der Herzmuskel in gleichen 
Zeiten gleiche Energiemengen ausgibt, die kiinstlich hervorgerufene Systole aber 
nur einen mehr oder minder grossen Bruchtheil der disponiblen Spannkraft in 
Anspruch genommen hat. . . . Man kann aber iiberhaupt nicht mehr die vorauf- 
gehende langere Herzpause zur Erklarung der Pulsverstarkung herbeiziehen; denn 
wir wissen jetzt, dass eine tibernormale Pause gar nicht besteht; die Herzruhe 
dauert nur so lange, als néthig ist, um die vorhergehende Verkiirzung der Pause zu 
compensiren. Wenn abnorm hohe Energiemengen sich angehauft haben, so ist 
daran nicht die Pawse Schuld, sondern die abortive, kiinstlich herbeigefuhrte 
Systole, die wegen ihrer Kleinheit nicht denjenigen Energievorrath erschépft hat, 
der einer normalen Systole zu Verfiigung gestanden hatte.” 

Since the extra contraction, however small, is, for the time when it occurs, 
maximal, it consumes all the available energy that has accumulated up to that 
time. The accumulation of energy after it begins at zero, and proceeds gradually. 
Whether the compensatory pause be called “ iibernormal ” or not, it is longer than 
the regular pause, and so allows time for the accumulation of more than the 
regular quantity of energy. It would be difficult, following LANGENDORFF’S view, 
to understand the weakness of the extra contraction. 


| 


248 Rk. S. Woodworth. 


It shows also the Bowditch “‘ staircase”’ phenomenon. 

It does not show, when beating spontaneously, a ‘‘ compensatory 
pause” after an extra contraction. The extra contraction is followed 
by a somewhat shortened but variable pause, and the earlier in diastole 
the extra contraction has intervened, the shorter is the following 
pause. 

The spontaneous beat following an extra contraction of the apex is 
much stronger than the regular beat. This results from the hasten- 
ing of the extra contraction, being an example of the “ staircase”’ 
effect. 

The height of any contraction of the apex, or of the frog’s heart, is, 
ceteris paribus, roughly proportional to the length of the preceding 
pause. 

The hastened extra contraction has in the dog’s apex an actual 
stimulating effect, which persists on the average for about eight sub- 
sequent beats. The more hastened the extra contraction, the stronger 
is its stimulating effect. Two or more extra contractions have a 
stronger stimulating effect than one. 

The optimum interval between beats is much shorter in the dog’s 
apex than in the frog’s heart, being as small as one second. 

The dog’s apex is not thrown into complete tetanus, nor into incom- 
plete tetanus with superposition of contractions, by the action of 
faradic or galvanic currents. 

A preparation similar to that from the dog’s apex, but including 
some of the base with ganglion cells, responds to electrical stimula- 
tion in the same way as the apex. 

The frog’s ventricle as well as the dog’s apex is absolutely refractory 
during systole. Hildebrand’s explanation of the apparent exceptions 
is shown to be correct. 

During the compensatory pause of the frog’s ventricle, both irri- 
tability and contractility are high; hence the pause is not a period of 
inhibition, 

Kaiser’s statement that two extra contractions of the frog’s heart 
are followed by a longer pause than one, three than two, etc., is not 
confirmed. Two or more extra contractions are often followed by 
even less than the normal length of pause. 

Engelmann’s explanation of the compensatory pause is supported 
by some new facts and is undoubtedly correct. 

The absence of the compensatory pause from a spontaneously beat- 
ing piece of the dog’s ventricle goes to prove that where the ventricle 


Contractions of the Heart. 249 


does show a compensatory pause, as in the normal heart beat, there 
the ventricle is not spontaneous. 

The great force of the beat following the extra contraction is depend- 
ent, in the frog’s ventricle, mostly on the prolonged compensatory * 
pause. It is not directly dependent, as Langendorff supposes, on 
the weakness of the extra contraction. 

No special or active compensatory function has yet been demon- 
strated in the heart; such compensation as appears after an extra 
contraction is incidental to the action of well-known properties of the 
cardiac muscle. 


: -- Vee ¥ 
| 7 7 = $5) ff ‘ q vi) ‘ wy 7 ix 
‘ tye V2 ¥ n ra 5, , 

. ¢ Le pe 7 a - 

iat) re ere wens: ¥ 

a F , @ 7 a ed he 

Ly i“ S49) See hae =i iaAae Pat a 
77 ae Laced Si, Rc Bameen were baer . 
’ 4 p ! 7 “ ‘ vy sa) 


(tiie al ebay ial re ae 

J * 1 Mati » pag A ee a 

Pere =" tix} te a vel oe 
Ree eS 

He et ia 

os Poss oe: =) rol fal | 


rts eed, alps 


THE RATE OF NERVOUS IMPULSE IN CERTAIN 
MOLLUSCS. 


By O. P. JENKINS anp A. J. CARLSON. 
From the Physiological Laboratory of Leland Stanford Junior University. 
Ly s wy 3 


'@ 1863 Fick! estimated the rate of the nervous impulse in Anodon 

as not more than one centimetre per second. He used the 
graphic method and based his estimate upon a small number of 
experiments. In more recent years the rate in three species of cephal- 
opods has been determined. Fuchs,” in 1894, found in the mantle 
nerve of Eledone moschata a rate of one metre per second. His 
determination was made by measuring the delay in the negative 
variation. Uexkiill,’ also in 1894, found in the same species, by means 
of the graphic method, a rate of one-half to one metre per second. 
Boruttau,* in 1897, determined by the method used by Fuchs, rates 
of three and one-half to five and one-half metres per second, in 
Octopus vulgaris, Octopus macropus, and Eledone moschata. So 
far as we have been able to learn, no other attempts have been made 
to determine the rate of the nervous impulse in molluscs. In this 
paper are presented the results obtained by us in determining the 
rate of the nervous impulse in the following species® of molluscs: 
Ariolimax columbianus (var.), Limax maximus, Pleurobranchza cali- 
fornica, Octopus punctatus, and Loligo pealii. 


1 Fick: Beitrage zur vergleichende Physiologie der irritablen Substanzen. 
Braunschweig, 1863. 

2 Fucus: Sitzungsbericht der kaiserlichen Akademie der Wissenschaften. 
Wien, 1894, ciii, p. 207. 

8 UEXKULL: Zeitschrift fiir Biologie, 1894, xxx, p. 317. 

4 BoruTTAu: Archiv fiir die gesammte Physiologie, 1897, Ixvi, p. 285. 

5 The identifications of the species of Ariolimax and Limax were made for us 
by Mr. H. A. Pitssry of the Academy of Natural Sciences of Philadelphia; 
those of Octopus and Loligo, by Professor H. HEATH of Stanford University. 
The sea rabbit is described in an unpublished paper by Professor F. M. 
MACFARLAND of Stanford University, as Pleurobranchea californica. 

251 


252 O. P. Jenkins and A. J. Carlson. 


Under the heading of each species are given the nerve experi- 
mented on, the dissection necessary, and the arrangement of apparatus 
used. A large number of graphic records were taken. <A represen- 
tative pair of tracings, one obtained by stimulation of the peripheral 
point, the other from the central, which show well the characters of 
the contractions, is given for each species. In Ariolimax columbianus, 
Limax maximus, and Pleurobranchza californica, the time of contrac- 
tion is of such a length that only the portions of the curves showing 
the latent periods are given. In Octopus punctatus the period of 
contraction is much shorter, and a large portion of its curve is given. 
In Loligo pealii the contraction and relaxation are prompt and quickly 
passed, producing a curve very similar to one obtained from a verte- 
brate muscle-nerve preparation. A single example of the tables of 
records obtained from each individual is also presented. 

In this table, “the total latent period” represents the time from 
the instant of stimulation to the beginning of the contraction of the 
musculature. “The length of nerve” is the distance between the 
central and the peripheral points of stimulation. The “ time of trans- 
mission” is the difference between the average of the latent periods 
obtained from the central point of stimulation and the average of 
those obtained from the peripheral point. The records from the 
central and the peripheral points were in most cases taken alter- 
nately. The summary given in the account of the experiments with 
each species brings together the rates obtained from the individuals 
of that species. 


ARIOLIMAX COLUMBIANUS. 


The giant slug, Ariolimax columbianus, first engaged our atten- 
tion. This slug is abundant in the neighborhood of Stanford 
University. It is found in greatest numbers along the banks of San 
Francisquito Creek during the rains, but can be obtained in the more 
moist places along this stream even during the dry season. Individ- 
uals have been observed ten inches in length when extended in 
crawling. Consequently these animals furnish long pedal nerves, 
easily exposed, giving, in large individuals, from ten to twelve centi- 
metres distance between the central and peripheral points of stimu- 
lation, which, together with the low rate of propagation, allows of 
fairly accurate determinations of rates. These conditions render the 
extent of the errors in measuring the rate of impulse in this nerve 
very much less than in the like determination in the nerves of ver- 


The Rate of Nervous Impulse in Certain Molluscs. 253 


tebrates where the rate is from seventy-five to one hundred times 
as great. 

The following method was found to be the most serviceable: A 
large moist chamber with the usual furnishing of electrodes was 
constructed. The body of the slug, with the exception of two to 
three centimetres of the posterior portion, which was left free to 
contract and relax, was firmly pinned, foot down, to the movable 
wooden floor of the moist chamber. The movable floor could be 
clamped in any position and thus allowed a greater range of adjust- 
ment without disturbing the animal once fixed. It was also convenient 
for the removal of the abundant mucus secreted. 

The animal was prepared by opening the body cavity by a longi- 
tudinal slit along the dorsal surface. The viscera being turned aside, 
the pedal ganglia and pedal nerves were exposed. The pedal nerves 
were sometimes cut close to the pedal ganglia, and sometimes were 
left in connection with the pedal ganglia, which were isolated from 
all other connections. All the branches of the pedal nerves were 
cut except those in the posterior portion of the foot. The tip of this 
free posterior end of the foot received a small hook, which was 
‘ attached to a thread passing over a small light pulley to a horizontal 
writing lever below. The weighting of this lever to bring it into a 
horizontal position varied very much on account of the unequal 
relaxing of the preparation. The lever magnified the motion four 
times. During most of the work Zimmermann’s universal stand was 
used to support this whole arrangement. This stand is almost 
essential in this work with this slug, since in it the unequal relaxa- 
tions after contractions make necessary repeated adjustment of the 
apparatus to the recording drum to keep the writing point of the 
lever on the drum and in a perpendicular line with the signal and the 
time marker. For much of the work Ludwig’s kymograph, Cambridge 
pattern, was used. However, Straub’s electromotor drum was found 
to be by far the most convenient and serviceable, and after its adoption 
it was used in all subsequent work. Tuning forks of fifty or one 
hundred double vibrations, either applied directly or by means of an 
electric signal, gave the time. The means for an efficient stimulus 
was obtained from induction currents, the break being employed. 

The Zimmermann vertical form of Du Bois-Reymond’s induction 
apparatus was used, containing 10328 windings in the secondary coil. 
Four Edison-Lalande cells, type S, gave the primary current. With 
the signal placed in the circuit, as was always the case, this arrange- 


254 O. P. Jenkins and A. J. Carlson. 


ment gave a current of 0.6 ampere. Platinum electrodes were used 
for contact with the nerve, one pair for each point of stimulation, and 
so connected by means of a commutator that the same direction of 
current was used in each pair of electrodes. 

The tendency of the slugs on being handled in any way, and much 
more on being injured, to contract the whole body and so remain for 
an indefinite time, introduces much difficulty in experimenting with 
them. The use of chloroform was found to be wholly impracticable, 
as it threw the animal into a strongly-contracted state in which it 
remained so long a time that further work was impossible. Ether 
proved to be serviceable, as it rendered the animal inactive, and put 
the body in a flaccid condition, allowing the preparation to be made 
successfully. This aneesthetic was used in many of the experiments. 
When used, only a sufficient amount was applied to complete the 
preparation, then a sufficient time was allowed to elapse to permit, as 


DAIWA 00 08000 WAAAY AAAAAAAAAA PAD AAAAA AAA AAD AAARAAPBADDAARS 


FIGURE 1. — Four-fifths the original size. Ariolimax columbianus. Pedal nerve. Dis- 
tance between central and peripheral electrodes: 8 cm. Rate: 40 cm. per second. 
Time: 50 d. v. per second. 


much as possible, a recovery from its effects before the experiment 
was begun. In the later experiments the anzsthetic was dispensed 
with, to avoid any possible errors which it might introduce, but no 
noticeable differences are found between results thus obtained and 
those obtained when the ether was used, followed by a delay for the 
effects to disappear. 

The nerves remain sufficiently irritable for from twelve to twenty- 
four hours to obtain muscular responses, but, as the rate of the 


tt is 


The Rate of Nervous Impulse in Certain Molluscs. 255 


nervous impulse seems to vary with the freshness of the preparation, 
only the first few tracings taken in each case are considered here. 
The rate seems also to vary with a large range of temperature, but as 
it is the purpose to consider conditions affecting the rate at another 
time, discussion of these conditions is deferred at present. Within 
the limits of temperature of the records here given, no variation in 
rate could be assigned to changes in temperature. The nerves were 
allowed, as much as possible, to lie in the blood and the abundant 
secretion of mucus, and were thus easily kept moist. The loss of 
moisture from the mucus was made up by the addition of distilled 
water to the mucus and blood, which, thus applied, seemed to act 
indifferently, while the usual physiological solutions acted as stimuli, 
and interfered with the experiments. 

Altogether fifty-four slugs were used, from which two hundred and 
fourteen records were obtained, one hundred and nine of these being 
from central points of stimulation, and one hundred and nineteen from 
peripheral points. 

Table I is given as an illustration of the data taken in a single 
individual. In Fig. 1 are given the curves obtained from another 
individual which are fairly representative. 


TABLE If. 
ARIOLIMAX COLUMBIANUS. 


Detail of record of No. 45. August 22,1901. Temperature, 17°-19° C. 


Total latent period in seconds. 


Central. Peripheral. 


0.59 0.30 
0.63 0.32 
0.61 0.35 


Average: 0.61 


Transmission time: 0.29 sec. Length of nerve: 
12cm. Rate: 41.4 cm. 


256 O. P. Jenkins and A. J. Carlson. 


Table II contains the rates obtained from the data of the fifty-four 


individuals. 
TABIGE Td: 


ARIOLIMAX COLUMBIANUS. 
Summary of rates in fifty-four individuals. 


Rate in Rate in Rate in Rate in 
cm. cm. cm. cm. 


48.14 : Biro 

36.52 
TBS) 
59.00 
AS, 
58.13 
57.26 
28.91 
22.00 
20.00 


94.64 
48.3 


87.00 


21.66 


As will be readily seen from this table, a great range, inside of 
limits, occurs in the rates obtained, and it is obviously impossible to 
fix upon any number as the rate of the nervous impulse in this slug. 
The conspicuous facts are the slowness of the rate and its variability, 
within the limits observed. 

An application of methods in use in the study of varying data to 
these rates, gives the following results. In forming the classes the 
first of each pair is not included. 


Class (rate incem.) 15-25 25-35 35-45 45-55 55-65 65-75 75-85 85-95 
Frequency 9 10 12 9 6 5 ] 2 
Of these the mean rate is 44 cm. per second. 
The standard deviation is 18. 


The coefficient of variability is 0.41. 


The Rate of Nervous Lupulse in Certain Molluscs. 257 


LIMAX MAXIMUS. 


This large slug, which is not a native of California, was obtained 
in abundance from the gardens in San José. The same apparatus 
was used in experiments with this slug that served with Ariolimax. 

Limax maximus allowed a distance of from four to six centimetres 
of nerve between the central and peripheral points of stimulation. 
The nerves and muscles are more irritable, and die more quickly than 
those of Ariolimax, and the muscle gives a more prompt reaction. 
The animal secretes less mucus. Both pedal nerves were placed on 
the electrodes, and the induced break shock was used as the stimulus. 
No anesthetics were used. On the whole, it was found much less 
difficult to obtain accurate records with this slug than with Ariolimax 


columbianus. 
TABLE III. 


LIMAX MAXIMUS. 


Detail of record of No. 12. November 29, 1901. Temperature, 15° C. 


Total latent period in seconds. Total latent period in seconds. 
ce ok eee 


Central. Peripheral. Central. Peripheral. 


0.115 0.100 0.130 0.098 


0.155 0.108 0.140 0.100 


0.140 0.100 | pare 0.099 
0.140 0.098 | 
0.150 0.098 | 

| 


0.098 


Average central: 0.138. Average peripheral: 0.099. Transmis- 
sion time: 0.039 sec. Length of nerve: 5.5 cm. Rate: 140.8 cm. 
per sec. 


Twenty individuals were used. Table III is given asa represen- 
tative of the records obtained, and Fig. 2 shows muscle curves which 
are typical of the series. In Table IV are presented the rates as 
calculated from the data obtained from the twenty slugs. They 
represent one hundred and seventy-five records, eighty-two from the 
central point of stimulation, and ninety-three from the peripheral 
point. The range of temperature was from 15° to 17° C. An 


258 O. P. Jenkins and A. /. Carlson. 


inspection of this table shows that in Limax maximus as in Ariolimax 
columbianus, there is a large coefficient of variability in the rate, the 
“mean rate, however, being nearly three times as high. 


———— FE a ee ee a ee ee eee 
LLP QL LPP PQQ QP DQQ J JPJPDLQPPQLQQJPJPJXYJWJYJJYJYJWYYYWwwWYWXwwYYYVYN*" 


FIGURE 2.— Full size. Limax maximus. Pedal nerve. Distance between central and 
peripheral electrodes: 6cm. Rate: 150 cm. per second. Time: 50 d. v. per second. 


TABLE IV. 
LIMAX MAXIMUS. 


Summary of rates of nineteen individuals. Temperature of room from 15° to 18° C. 


Rate in | No.of | Rate in Rate in Rate in | Rate in 
cm. slug. cm. - em. ||| cm. 


204.0 | 91.3 | 
90.0 On) 94.0 
83.0 4 || 15 | 1200 


103.2 


Analysis of the above table. 


Class (rate in cm.) 30-70 70-110 110-150 150-190 190-230 
Frequency 2 7 5 2 5 
Mean rate, 124 cm. per second. 
Standard deviation, 49. 


Coefficient of variability, 0.39, 


| 


The Rate of Nervous Impulse in Certain Molluscs. 259 


PLEUROBRANCH#A CALIFORNICA. 


The large sea rabbit of Monterey Bay served next in this series, 
and rates were determined from both pedal and mantle nerves. This 
species can hardly be said to be abundant at this point, since by the 
help of the local fishermen only occasional specimens were secured. 
Those obtained were brought in under our direction, and appeared to 
be uninjured and in normal condition. With but one exception 
experimentation was begun on each within a few hours after capture. 
In the one exception, the animal remained in the aquarium twenty- 
four hours before work was begun on it. This animal furnishes an 
excellent muscle-nerve preparation. 

In the aquarium the animal was made to crawl upon a board of 
suitable size, which was clamped to the floor of the moist chamber in 
the manner made use of in the work with Limax and Ariolimax. When 
approximately normally extended, two strong needles were quickly 
thrust through the mantle and foot into the board on either side, 
about five or six centimetres from the posterior end of the foot. This 
caused the animal to contract strongly, but it soon relaxed, and was 
then similarly secured anteriorly. It was then removed from the 
aquarium, and prepared for experimentation in much the same way 
as were the slugs. 

The pedal nerves run free in the body cavity for some centimetres 
from the ganglia, and then penetrate more or less deeply the mus- 
culature of the foot, then reappear on or near the surface two to three 
centimetres from the posterior end of the body cavity. The whole 
length of nerve between the central and the peripheral electrodes was 
dissected free in only two preparations. This amount of dissection 
was found to be unnecessary, and in the remaining animals the dis- 
section was reduced to the severing of the side branches of the main 
nerves where the latter reappear in the posterior part of the body 
cavity. By this means the possible injury to the nerves by handling 
in dissecting was avoided. This did not affect the accuracy of the 
records, for the posterior reacting end was so secured that contrac- 
tions in any other part of the foot did not interfere with its record. 
At the conclusion of each of the experiments, the nerves were dis- 
sected out and measured. In the experiments with the mantle nerve, 
the portion lying free in the body cavity was used, which consequently 
was not further disturbed by dissection. The writing lever was 
attached in the usual way to the mantle. 


260 O. P. Jenkins and A. J. Carlson. 


Throughout the experiment, the preparation was freely bathed with 
the fluid of the body cavity. With the same batteries and coils used 
with the slugs, strong contractions were obtained from the central 
point of stimulation with the secondary coil at fourteen centimetres 
from the primary. The irritability of the preparation decreases 
rapidly in the course of experimentation, and when but feeble con- 
tractions followed single break shocks with the secondary at 5 cm. to 
o cm. away from the secondary, strong contractions could be obtained 
with the interrupted current of from 0.02 sec. to 0.12 sec. duration 
with the secondary at from ten to twelve centimetres from the 
primary coil. 

The preparation gave good reactions for from ten to twenty hours, 
if the stimulations were not too frequent. In one case the muscle 
gave feeble response to stimulations of the nerve twenty-six hours 
after the preparation was made. 


UA eave 


PLEUROBRANCH-EA CALIFORNICA. 


Detail of record of No. 10, pedal nerve. January 10, 1902. Temperature of Aqua- 
rium, 14° C. 


Central. Peripheral. Central. Peripheral. 


0.320 0.157 0.330 0.145 
0.332 0.165 0.327 0.162 
0.332 0.163 0.332 0.170 
0.345 0.180 0.335 


Average central: 0.330. Average peripheral: 0.163. Transmission 
time: 167 sec. Length of nerve: 13 cm. Rate: 76 cm. per sec. 


After a large number of records were taken, the preparation 
usually seemed to show a marked decrease in the rate. For this 
reason, only the first few reactions were used in determining the rate. 
The two examples of the mantle nerves appear to indicate a lower 
rate than that of the pedal nerves, but a greater number of observa- 
tions might have shown no difference in the mean rates, or the lower 
rate may be due to the fact that the records were obtained after the 
considerable delay caused by the use of the preparation in the experi- 
ments on the pedal nerves. 


———— 


The Rate of Nervous Impulse in Certain Molluscs. 


Table V exhibits the full records of a single individual. 
contains typical tracings from one of the specimens. 
includes the rates from the pedal nerves of ten individuals. 


261 


Fig. 3 
Table VI 
The 


rates from the mantle nerves of two specimens are appended. 
In this mollusc it will be noted that while the rate is low, there 
is in the series used a comparatively small coefficient of variation 


in the rates. 


ne 


anannonnnnnenennnnnnnnnnnnnnnnnanneneenennnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnrnnm 


<_< —— eee 


Ficure 3.— About four-fifths the original size. Pleurobranchza californica. Pedal 
nerve. Distance between central and peripheral electrodes: 13 cm. Rate: 71.5 cm. 


per second. Time: 100 d. v. per second. 


TABLE VI. 


PLEUROBRANCHA CALIFORNICA. 


Summary of rates in the pedal nerves in ten individuals. 


DZS to S ote: 


No. of Rate in 
specimen. cm. 


No. of 
specimen. 


Temperature of aquarium, 


Rate in 
cm. 


69.4 
84.6 
65.2 
85.0 
80.9 


61.6 
77.0 
93.75 
90.9 


262 O. P. Jenkins and A. J. Carlson. 
Analysis of Table VI: 
Class (rate in cm.) 55-65 65-75 75-85 85-95 
Frequency 1 2 5 2 


Mean rate, 78 cm. per second. 
Standard deviation, 7.8. 


Coefficient of variation, .10 


The rates obtained from the mantle nerves of two individuals 
were sixty-one centimetres and fifty-four centimetres per second 
respectively. 


OcTOPUS PUNCTATUS. 


This species of Octopus is fairly common at Pacific Grove. We 
obtained records from eight individuals. Seven were studied Decem- 
ber, 1901, to January, 1902, and one in the summer of 1902. 

The same muscle-nerve preparation was used that Uexkull em- 
ployed in his experiments on Eledone. The nerve portion consisted 
of the pallial nerves, and the muscular portion of a wide strip of the 
mantle muscle around the stellate ganglion, from three to four centi- 
metres long and two centimetres wide. The muscle strip was made 
to pull either vertically or horizontally, the two methods giving the 
same results. No anesthetics were used. The nerve was not 
entirely freed from the adjacent muscular tissue, as it was found 
difficult to free it without injury. The preparation was kept moist 
by sea-water. 

With the same electrical arrangement used in the preceding 
experiments, contractions were obtained with the secondary at from 
twenty-two to eighteen centimetres from the primary. The break 
was employed throughout, and from two to three minutes were 
allowed to elapse between the stimulations. The peripheral elec- 
trodes were placed two to four millimetres from the ganglion. That 
there was no direct stimulation of the muscle by the escape of the 
current from the peripheral electrodes, seems to be proved by the 
fact that it required ten to fifteen times the intensity of stimulus 
used with the nerves to secure equal contractions when applied 
directly to the muscle. Furthermore, the low degree of the in- 
tensity of stimulus made improbable the escape of the current from 
the peripheral point to the ganglion, thus directly stimulating it. 

Table VII, giving the detail of the records, of one specimen, shows 
great uniformity in the records, although a great number of tracings 


ST 


The Rate of Nervous Impulse in Certain Molluscs. 263 


TABEE VII. 
OcTroPUS PUNCTATUS. 


Detail of record of No. 8. January 2,1902. 14°C. Total latent period in seconds. 


Central. Peripheral. | Central. Peripheral. 


0.066 0.050 0.070 0.048 
0.067 0.045 | 0.069 0.050 
0.066 0.045 | 0.075 0.054 
0.068 0.046 0.070 0.050 
0.068 0.046 ee 0.054 
0.070 0.046 | 0.070 0.051 
0.065 0.047 | 0.070 0.048 
0.070 0.070 0.049 
0.075 0.075 0.055 
0.072 re 0.070 0.053 


0.070 0.050 0.075 0.053 


Average central: 0.070. Average peripheral : 0.050. Transmission 
time: 0.02 sec. Length of nerve: 4cm. Rate: 200 cm. per sec. 


FicurE 4.— One-half the original size. Octopus punctatus. Pallial nerve. Distance 
between central and peripheral electrodes: 4 cm. Rate: 200 cm. per second. 
Time: 100 d. v. per second. 


were taken from the single individual. Fig. 4, exhibiting a pair of 
tracings, shows a nearer approach in form to the muscle curve of the 
vertebrates. Table VIII brings together the records of eight unin- 


264 O. P. Jenkins and A. J. Carlson. 


jured individuals. It will be noted that the rate is comparatively 
higher than in the preceding molluscs, and that there is also a com- 
paratively small coefficient of variability. 


TABLE VIII. 
OCTOPUS PUNCTATUS. 


Summary of rates in eight individuals. Temperature of aquarium, 12° to 14° C. 


No. of Rate in No. of Rate in 
specimen. cm. specimen. cm. 


227 
192 
192 
227 


Analysis of table. 
Class (rate in cm.) 140-170 170-200 200-230 
Frequency 1 2 5 
Mean rate, 200 cm. per second. 


Standard deviation, 20. 
Coefficient of variability, 0.10. 


LOLIGO PEALII. 


This species is found in abundance in Monterey Bay during the 
breeding season (June-July). We succeeded in keeping the speci- 
mens in good condition in the aquarium for from one to three days 
only, as they soon injured themselves by darting against the sides 
of the aquarium. 

The pallial nerve left in connection with the fin served as the 
muscle-nerve preparation. Each nerve issues through the cranial 
cartilage in two branches closely joined to one another. One branch 
enters the stellate ganglion; the other courses past it, runs free in 
the body cavity near the median line, and penetrates the mantle about 
two centimetres dorsally to the ganglion, taking a dorso-lateral direc- 
tion, and issues on the dorsal side of the mantle at the anterior edge 
of the fin, where it immediately radiates into the musculature of the 
latter. 

The animal was secured to the floor of the moist chamber, dorso- 


The Rate of Nervous Impulse in Certain Molluscs. 265 


lateral surface down, part of the mantle being removed in order to 
facilitate the freeing of the pallial nerve on one side. The two 
branches were severed near the cranial cartilage, and the branch to 
the stellate ganglion cut near the latter, as it was found difficult to 
separate the two without injury. The integument was removed from 
the anterior part of the fin and adjacent parts of the mantle, and the 
issuing branches of the nerve separated as completely as possible 
from the muscle for from five to eight millimetres. This served as 
the point of application for the peripheral electrodes. The hook, 
connected by a thread to the writing lever, was fastened to the fin at 
about its middle portion. The preparation being turned slightly to 
one side, the fin was bent dorsally, the outer edge touching the plat- 
form. The beginning of movement ventralward was communicated 
to the lever. The latter had to be very delicately adjusted because 
the power of the fin in this position is slight. 


JPY s\e oY le] BF eal LD 
LOLIGO PEALII. 


Detail of record of No. 33. Temperature, 13° C. 


Total latent period in seconds. Total latent period in seconds. 


Central. Peripheral. Central. Peripheral. 


0.043 0.027 0.043 0.028 
0.043 0.027 | 0.044 0.027 
0.040 0.025 0.044 0,028 
0.042 0.028 0.045 | 0.027 


Average central: 0.043. Average peripheral: 0.027. Transmission 
time: 0.016 sec. Length of nerve: 8cm. Rate: 500 cm. per sec. 


With the device for stimulating the same as in Octopus, maximal 
contractions were obtained with the secondary coil at twenty-six to 
twenty-one centimetres from the primary. 

The nerve dies within five to ten minutes after being prepared. 
Hence only one preparation from each animal could be used. No 
anzesthetics were employed. : 

In the summary are brought together the results obtained from 


266 O. P. Jenkins and A. J. Carlson. 


the records of thirty-four individuals, twenty-two of which were 
taken in the summer of 1901, and twelve obtained in the summer 
of 1902. 

Table IX, the detail of the records of a single individual, shows 
great uniformity throughout the experiment. Fig. 5 shows the trac- 
ings from one of the animals. They are fairly typical of the series. 
The form is notably similar to the muscle curve obtained from the 
gastrocnemius of the frog. 


FicurE 5.—Full size. Loligo pealii. Nerve to fin. Distance between central and 
peripheral electrodes: 8 cm. Rate: 444 cm. per second. ‘Time: 100 d. v. per 
second. 

TABLE X. 


LOLIGO PEALII. 
Summary of rates in thirty-four individuals. Temperature of Aquarium, 12° to 14° C. 


| 
No. of Rate in No. of Rate in No. of Rate in || No. of Rate in 
specimen. cm. specimen. cm. specimen. cm. || specimen. 


] ie 10 350.0 19 | 325.0 


2 499.8 ll 350.0 
3 


650.0 12 333.0 
650.0 || 13 312.6 
499.8 366.3 


458.1 || 325.0 


406.2 || 341.9 


392.7 


323.4 


The Rate of Nervous Impulse in Certain Molluscs. 267 


Analysis of above table. 


Class (rateincm.) 250-350 350-450 450-550 550-650 650-750 750-850 
Frequency 10 10 S, 3 1 1 


Mean rate, 435 cm. per sec. 
Standard deviation, 121. 
Coefficient of variability, 0.28. 


In Table X, are the results obtained from thirty-four individuals. 
An examination of the table shows a much higher rate than shown 
in the other molluscs experimented with. The range is also large. 
With the two, Nos. 30 and 31, which have the exceptionally high rates 
of seven hundred centimetres and eight hundred centimetres per 
second, there were no special conditions which we were able to observe 
which could account for these specially wide variations. 

For convenience of comparison, the results of the determination of 
the rates in these molluscs are brought together in Table XI: 


TABLE XI. 


No. of Mean | Standard | Coefficient 


agate rate in Se eats 
individuals. a deviation. | variability. 
| 


Species. 


Ariolimax columbianus. . | _ 3 pedal TSO) Oct! 


Limax maximus ... . pedal 49.0 0.39 


Pleurobranchzea californica pedal | 7.8 010 


Pleurobranchea californica mantle 
Octopus punctatus . . . pallial 


Molivovpealit (. . 3 << « to fin 


The muscle curves of these molluscs, compared with their rates of 
nervous impulse, show a corresponding progressive series. The lower 
rate is accompanied by a slow and long continued contraction of the 
muscle, the recovery from the contraction being also much prolonged. 

With the increase of rapidity of nervous conduction is a corres- 
ponding increase in the promptness and quickness of muscular con- 
traction and in its rapid recovery. A series could be selected from 
these molluscs which would show a marked progressive development 
in nerve and muscle, of those conditions and properties which make 


268 O. P. Jenkins and A. J.. Carlson. 


possible the prompt and accurate motions which are the marked 
characteristics of highly organized animals. 

A series based upon the rates of nervous impulse, and the compact- 
ness of the myogram in these animals, coincides with one based on 
the degree of activity shown by them in their means of locomotion. 

The slowest in motion is Ariolimax, and far surpassing it is Octopus, 
which in turn is well outstripped by Loligo. The rates of nervous 
impulse, and the forms of the myograms of the three, well express the 
differences in the degree of activity manifested by the three. It 
would seem that these molluscs offer a promising field for the inves- 
tigation of certain problems in nerve and muscle physiology from the 
very fact of these gradations in the properties of the nerve and muscle 
tissue in them. 


ON THE INFLUENCE OF CALCIUM -AND POTASSIUM 
SALTS -UPON THE- TONE OF PLAIN MUSCLE: 


BY PERCY -G. STILES. 


[From the Physiological Laboratory of the Massachusetts Institute of Technology.| 


T is a familiar fact that minute additions of a calcium salt to a cir- 
culating medium lead to an increased tone of the cardiac muscle, 
and that a potassium salt may be employed to antagonize this action. 
This antagonism, first pointed out by Ringer, and since investigated 
by Howell and many others, has come to be a matter of ready demon- 
stration in every laboratory. But it has also been noticed that the 
influence of calcium and potassium uvon the tone of the heart is not 
the same, even qualitatively, when concentrations much greater than 
the physiological ones of the blood are used. 

In studying the influence of salts upon the rhythmic property of 
the cesophagus, the writer found that for quantities of potassium and 
calcium near the amount normal in the blood, the effect of each is 
definite, and the antagonism distinct; calcium causes shortening of 
the fibres, potassium relaxation. But it became evident at that time 
that the same rule does not apply to plain muscle, any more than to 
the heart, when quantities of calcium and potassium in excess of the 
physiological are used. An investigation of the specific effects of 
calcium and potassium in higher concentrations seemed desirable, 
and the more so because contradictory statements are common in the 
recent literature of such subjects.! 

Experiments of a simple character were made upon the following 
plan. A strip from the stomach of the frog was immersed in a bath, 
and made to record its changes of extension by means of a light lever. 
The tone under observation was that of the circular musculature. 


1 For example, ZOETHOUT (This journal, 1902, vii, p. 199) finds that potassium 
salts throw skeletal muscles into strong tonic contraction, and that calcium salts do 
not produce such a degree of tone. He made use of pure solutions of these salts, 
isotonic with the blood. These results are in harmony with my own for high 
concentrations. 

269 


270 Percy G. Stiles. 


The bath, contained in a small beaker, consisted of exactly 25 c.c. of 
0.7 per cent sodium chloride solution, to which accurately measured 
additions of 1 per cent calcium chloride and 1 per cent potassium 
chloride were made from time to time. The composition of the 
mixture at any moment could be calculated. The strength of the 
solutions was such that the osmotic pressure could not have been 
significantly changed by adding them to the bath. Spontaneous con- 
tractions were often observed, as was anticipated, but no attention 
was paid to them. Observation was limited to the changing base- 
level upon which these contractions were superimposed. 

In eighteen experiments which it is needless to detail, the effects 
of various concentrations of calcium chloride and potassium chloride, 
both separately and combined, were noted. 

The physiological amounts of calcium and potassium salts are imi- 
tated in the Ringer’s solutions in common use. ‘These contain not 
far from 0.026 per cent calcium chloride, and 0.03 per cent potassium 
chloride, — the proportions recommended by Howell and Greene, and 
based upon analyses of the ash of terrapin blood. In such propor- 
tions the two salts exhibit clearly the specific influences usually 
attributed to them. A solution of sodium chloride, plus 0.026 per 
cent calcium chloride, raises the tone of a smooth muscle preparation ; 
a solution of sodium chloride, plus 0.03 per cent potassium chloride, 
depresses it. Now, what are the facts when we exceed the physio- 
logical limits in the concentration of either salt ? 

The effect of sodium chloride upon tone is a factor which ought to 
be reckoned with. Plain muscle placed in sodium chloride solution 
relaxes steadily for a long time, and probably reaches at last as great 
an extension as can be induced by any mixture of salts. So the effect 
of potassium chloride in the same direction can only be inferred from 
its power to Aasten the decline of tone, or, better, from what is seen 
when enough calcium chloride is present to neutralize the depressing 
effect of the sodium chloride. 

Making allowance for a depressing property of sodium chloride 
which was exerted at the beginning of each experiment, the following 
facts were deduced. Calcium chloride causes increase of tone, whether 
it is brought to the contractile tissue in small amounts, or in concen- 
trations as great as 0.3 per cent, — perhaps twelve times the physio- 
logical concentration. Up to this strength of solution at least, the 
rise of tone proceeds regularly and hand in hand with the increasing 
calcium content of the bath. 


On the Influence of Potassium and Calcium Salts. 271 


With potassium chloride the case is less simple. The depressing 
effect of this salt upon tone was readily confirmed for concentrations 
below 0.15 per cent. When this percentage is exceeded, the effect is 
obscure at first; but when the concentration passes above 0.2 per 
cent, the tone is clearly heightened by the potassium chloride, just as 
it is by calcium chloride. , With concentrations of 0.3 per cent or 
0.4 per cent potassium chloride, the tone is intense. Thus it appears 
that calcium and potassium have an antagonistic action when each is 
present in a low concentration, but each acts to cause tonic contrac- 
tion when a somewhat larger amount is present. 

The experiments in which both calcium chloride and potassium 
chloride were used at once in varying quantity show that the com- 
bined effect is just what would be anticipated from the facts stated 
above. The most marked depressing effect secured with potassium 
chloride is with a concentration of about 0.1 per cent. Such an 


0 01 oz <% ONS C102) 07, 


Effect of calcium chloride in increas- Effect of potassium chloride in increas- 
ing concentration upon the tone of ing concentration upon the tone of 
smooth muscle. smooth muscle. 


The dotted verticals indicate the effects of physiological percentages of each salt. 


amount of potassium chloride was found to keep the tone at a low 
level until calcium chloride to nearly twice that percentage had been 
introduced. Then arise of tone would set in, and further additions 
of potassium chloride would not check but rather promote it. The 
action of the two salts when simultaneously present is, therefore, to 
be represented, roughly at least, by the algebraic sum of their effects 
when tried separately. 

The facts can be graphically presented, but with the caution that 
these curves are merely of qualitative meaning; there is no unit of 
measurement for the tone-changes, nor are the ordinates of one curve 
exactly comparable with those of its fellow. 

There is one objection to my method of experiment which has a 
certain force and must be admitted. When calcium chloride or 
potassium chloride solutions are added to a bath which originally 
contained 0.7 per cent sodium chloride, the percentage of sodium 


372 Percy G. Steles. 


chloride is diminished by dilution, and it is natural to ask whether 
the rise of tone which was ultimately brought about, whether calcium 
chloride or potassium chloride was added, might not be due to lack 
of sodium rather than to the other salts. Asa matter of fact, the 
sodium chloride was never reduced below 0.45 per cent in experi- 
ments the results of which were considered seriously. This tissue 
does not go into tone when placed in a hypotonic sodium chloride 
solution of 0.45 per cent, nor does it when 0.7 sodium chloride is 
diluted to the same extent with isotonic dextrose solution. So it 
seems probable that the effects on tone were really due to the calcium 
and potassium salts. 


These observations do not justify the writer in entering into the 
current controversy regarding the precise rdle of sodium, calcium, and 
potassium ions in life-processes. But it may be timely to emphasize 
the importance of studying the action of these elements zz physiologi- 
cal percentages. How unsafe it is to infer anything as to the normal 
influence of a salt from its effect at an abnormal concentration is well 
shown in the case of potassium chloride which, when present in an 
amount of 0.2 per cent or more, has quite the contrary action to that 
which is properly regarded as its physiological one. | 

My acknowledgments are due to Dr. Theodore Hough, by whose 
courtesy the laboratory was opened to me. 


on DIFFERENCES IN THE DIRECTION OF ‘THE ELEC- 
2 RICAL, CONVECTION..OF CERTAIN. FREE CELLS 
au Os NUCLET. 


~Dbe MRALPH. S. LLELLE, 
[from the Physiological Laboratory of the Harvard Medical School.] 


I. INTRODUCTORY. 


T has long been known that minute particles of.various substances 
suspended in distilled water or other feebly conducting media in 
which are immersed the terminals of a battery, are set in motion and 
travel through the liquid toward one or the other of the two poles, the 
direction of migration depending partly on the chemical composition 
of the suspended particles, and partly on.the nature of the medium ; 
thus particles of silica, graphite, and sulphur travel in water toward 
the anode, but in turpentine the direction of migration is reversed in 
the case of the first two substances. A difference of electrical poten- 
tial between particle and- medium is thus indicated; the suspended 
particles are charged oppositely to the medium at the surface of con- 
tact between the two; they therefore move up or down the potential 
gradient according to the sign of the charge they carry, whether 
negative or positive respectively. 

Recently this conception has received an interesting special appli- 
cation in the case of colloidal solutions. Colloidal substances in 
solution have been shown to exhibit a tendency to collect in the 
neighborhood of the electrodes when a current is passed;? such 
solutions have furthermore been proved very clearly to consist of 
suspensions of discrete particles, typically in a very fine or even 
ultra-microscopical state of subdivision. The conclusion seems there- 


1 For an account of the phenomena of electrical endosmose and electrical con- 
vection, see WIEDEMANN: Elektricitat, second edition, 1893, i, pp. 993-1023. 

2 See PicTON and LINDER: Journal of the Chemical Society, 1892, ]xii, p. 148, 
and 1897, Ixxi, p. 568; CoEHN: Zeitschrift fiir Elektrochemie, 1897, iv, p. 63; 
Harpy: Journal of physiology, 1899, xxiv, p. 288. 


273 


274 Ralph S. Lille. 


fore inevitable that each colloidal particle carries an electrical sur- 
face-charge; in some cases this is positive, in others. negative, as 
indicated by the characteristic differences in the direction of migra- 
tion of different colloids! A partial insight is thus gained into the 
nature of certain of the most typical peculiarities of hydrosols, —in 
particular their remarkable susceptibility to the coagulative action of 
electrolytes and their indifference toward non-electrolytes. It is 
further found that different salts vary in their ability to effect precipi- 
tation of a given colloidal substance; in several instances it has been 
determined that precipitation is due chiefly to the action of ions bear- 
ing charges of opposite sign to those of the colloidal particles; the 
coagulative power of an ion is found also to be to a marked degree a 
function of the number of its charges, increasing very rapidly with 
increase in valency.? All of these peculiarities have been found sus- 
ceptible of more or less satisfactory explanation on the basis of the 
above theory.® 

The researches of Picton and Linder and Hardy ® indicate further- 
more that the sign of the charge carried by the particles in a hydrosol 
bears a definite relation to the chemical nature of the colloidal sub- 
stance, acid particles being electrically negative and basic particles 
positive. Thus the particles are positively charged in hydrosols of 
ferric hydrate, aluminium hydroxide, methyl violet, Magdala red, 
alkali-albumin in acid solution; they are negatively charged in hydro- 
sols of silica, arsenious sulphide, aniline blue, and alkali-albumin in 
alkaline solution. It seems probable that this relation, if of general 
validity, is of far-reaching physiological significance. Thus it is 
known that a large series of proteids, the nucleo-proteids, possess 
pronounced acid characteristics; also that the degree of acidity of 
these compounds varies with the proportion of nucleic acid present,” 


1 See Picron and Linper: Loc. ci¢.; also—for an excellent discussion and 
résumé of the chief literature bearing on the nature of colloidal solutions — 
BreEbIG: Anorganische Fermente, Leipzig, 1901. A résumé and literature list are 
also given by WHITNEY and OBER: Journal of the American Chemical Society, 
1900, xxii, p. 842. 

2 See especially HARDY: Proceedings of the Royal Society, 1900, Ixvi, p. 110 ; 
also Picron and LINDER: Journal of the Chemical Society, 1895, Ixvii, p. 63. 

’ For an ingenious instance, see WHETHAM, quoted in HARpy’s paper in Journal 
of physiology, 1899, xxiv, p. 288. 

4 Picron and LinperR: Journal of the Chemical Society, 1897, Ixxi, p. 568. 

5 Harpy: Journal of physiology, 1899, xxiv, p. 288. 

5 See LILIENFELD: Archiv fiir Physiologie, 1893, p. 391. 


Electrical Convection of Certain Free Cells and Nuclet. 275 


reaching a maximum in the chromatin of dividing cells and sperma- 
tozoa where the proportion of nucleic acid is highest. The chromatin 
of sperm-heads appears indeed frequently to consist chiefly of nucleic 
acid in proteid-free combination with various organic bases (protamin, 
sturin, arbacin!); many chromosomes are probably of similar com- 
position; to this corresponds the intense affinity of these structures 
for basic or nuclear dyes. On the other hand, the extra-nuclear or 
cytoplasmic proteids seem predominantly basic in character, evincing 
a special affinity for actd dyes. A contrast thus exists between the 
nuclear and the cytoplasmic colloids in respect to general chemical 
character. It follows—if the above defined relation holds true in 
this instance— that a corresponding electrical difference must accom- 
pany this chemical difference; in other words, that the chromatin 
colloidal particles are negatively, the cytoplasmic positively charged. 
On this theory a permanent though variable difference of electrical 
potential must exist between chromatin and cytoplasm in the living 
cell. 

This condition, if actual, must of necessity profoundly influence the 
mutual relations of nucleus and cytoplasm. Thus, if the measure of 
this electrical difference is indicated by the degree of contrast in the 
reactions of chromatin and cytoplasm toward the usual dyes, it is evi- 
dent that the above difference of potential must be greatest at the 
period of mitosis, since it is then that the chromatin is invariably in 
its most strongly acid and (on the above theory) electrically negative 
phase. The further consideration immediately presents itself that it 
may be this potential-difference which in itself constitutes the primary 
and determining condition of mitosis. /x a// cases the appearance of 
the cytoplasmic radiations and the formation of the mitotic figure are 
accompanied by a passage of the nuclear chromatin into a phase rich 
in nucleic acid; evidently the two parallel series of changes are inti- 
mately inter-connected. The marked resemblance between the rays 
of the mitotic figure and the electrical and magnetic lines of force — 
not to mention other typical peculiarities of mitosis — affords no little 
additional indication that the process is essentially electrical in its 
nature. Certainly, in view of our present knowledge of the physical 
chemistry of colloidal solutions —even though this is as yet far from 
complete — we must admit that electrical theories of mitosis are 


1 See MIESCHER: Gesammelte Abhandlungen, Leipzig, 1897, p. 55; KOSSEL: 
Zeitschrift fiir physiologische Chemie, 1896, xxii, p. 179; MATHEWS: /dzd., 1897, 
XXxili, p. 339. 


276 Ralph S. Lille. 


entitled to more careful consideration than they have hitherto received. 
I shall not, however, discuss these questions at length in the present 
paper; the above remarks are intended merely to indicate the nature 
of the considerations that have led to the performance of the following 
experiments. 

It may be assumed that the direction and speed of the electrical 
migration of living cells and portions of tissues are chiefly dependent 
upon the electrical characteristics of their constituent colloids. Free 
nuclei, and especially nuclei whose chromatin contains a high propor- 
tion of nucleic acid, should on this assumption migrate with the 
negative stream, while cells with voluminous cytoplasm should prove 
less strongly negative, or even in some cases exhibit a positive direc- 
tion of migration. The following experiments represent the begin- 
ning of a series designed to test these and other inferences from the 
hypothesis outlined above. 


IJ. EXPERIMENTAL. 


The methods were as follows: Finely divided tissues, in as fresh a 
condition as possible, were teased in 7 cane-sugar solution (isotonic 
with physiological salt-solution) and mounted in the same medium 
upon a specially prepared slide, so constructed that the entire prep- 
aration while under examination could at any time be exposed to the 
action of the electric current. The construction of this slide is as 
follows: a long cover-glass (50 by 25 mm.), around which pass two 
tightly drawn transverse loops of thin platinum wire about 15 mm. 
apart, is cemented by means of Canada balsam to an ordinary micro- 
scopical slide. The platinum wires are connected through a pole- 
changer and simple key to the poles of a battery; this has consisted 
usually of three storage cells with an aggregate E. M. F. of from 7 to 
7.5 volts. The tissues and cells under examination are mounted in 
sugar-solution on the slide in the space between the wires, and their 
behavior in the electric field can then be studied under high powers. 
The rate of movement is measured by means of the ocular microm- 
eter. The resistance of the sugar-solution is very high and only a 
slight current passes between the platinum wires; as a rule, however, 
a few minute gas bubbles appear on the wires after the current has 
been flowing for a few minutes. It has not been thought necessary 
to measure each time the exact strength of this feeble current, which 
would obviously be found to exhibit considerable variability, since 


Electrical Convection of Certain Free Cells and Nuclet. 277 


the diffusion of electrolytes from the tissues under examination into 
the medium must necessarily confer upon the latter a certain slight 
and variable degree of conductivity. 

The following structures have been examined by the above methods ; 
isolated nuclei containing chromatin in different states of aggregation 
(nuclei of spermatozoa, small leucocytes, nuclei from lymphoid 
tissues); also cells and tissues with relatively voluminous cytoplasm, 
such as muscle-cells, red blood-corpuscles, and the larger forms of 
leucocytes. 

When the corpuscles of freshly drawn frog’s blood are examined 
under the conditions described above, characteristic differences in 
the behavior of the different elements become at once apparent. 
The majority of the red corpuscles move slowly (at an average 
speed of approximately 120 to 130 w per minute) in the direc- 
tion of the negative stream; but a considerable number exhibit no 
decided movement in either direction, and a few are weakly positive. 
The leucocytes vary in behavior,! the minute lymphocytes (ca. 
10 mw in diameter) almost always exhibit a well-marked negativity, 
moving rapidly (at an average of speed of approximately 1,500 m per 
minute) toward the anode. Leucocytes of medium size (15 to 20 #) 
are usually slightly negative, but may be indifferent or even slightly 
positive; while the more voluminous leucocytes (25 to 30 » diameter ) 
are in almost all instances decidedly fositive, moving definitely to- 
ward the cathode, although at a moderate speed (120 to 130 » per 
minute). The same differences of behavior appear if the cells are 
obtained by fine subdivision of the spleen or thymus, voluminous 
leucocytes (25 “ and upward) being almost invariably positive, while 
the smallest leucocytes, which consist essentially of nuclei with 
merely a thin surface-film of cytoplasm, are strongly negative. 

The appearance of the microscopic field during the passage of the 
current is remarkable: although all the cells in the field are under 
identical external conditions, some are seen to travel in one direc- 
tion, while others simultaneously pursue a diametrically opposite 
course. Whenever the direction of the current is changed, each cell 

1 E. DInEuR (Bulletin de la Société belge de Microscopie, 1891-2, xxviii, 
No. 5, p. 113) placed electrodes consisting of platinum wires enclosed in capillary 
tubes within the peritoneal cavity of the frog, and found that the leucocytes tended 
to enter these tubes, especially the one connected with the positive pole of the 
battery. Under certain conditions (as in inflammation) they tended rather to 


enter the negative electrode. He attributed the movements to a special sensibility 
to electricity (“‘ galvanotaxisme ”) possessed by the leucocytes. 


278 Ralph S. Lithe. 


instantly reverses the direction of its movement and proceeds with 
its original velocity in exactly the reverse direction. We have, as it 
were, a graphic representation of the migration of ions in an electro- 
lyte-solution through which a current is passing, the lymphocytes or 
free nuclei corresponding to anions, and the large leucocytes to 
kations. The rate of migration of these cells, needless to say, greatly 
exceeds that of the ions since the frictional resistance at the surface 
of contact with the medium is relatively slight for such large bodies; 
in other respects the parallel appears remarkably close. For instance, 
if the current is allowed to flow for a considerable period in one direc- 
tion, there results a gathering of red corpuscles and smaller leucocytes 
at the anode, and of the voluminous leucocytes at the cathode. This 
difference in the direction of electrical convection presumably implies 
a corresponding difference in the sign of the aggregate electrical 
charge carried individually by cells of the above two kinds. 

From the thymus numerous small and densely staining nuclei may 
be obtained by teasing. Many of these are of almost uniform size 
(approximately 10 uw in diameter); they are always found to exhibit 
avery uniform and rapid movement (from 1.5 to 2.0 mm. per minute) 
in the direction of the negative stream. These nuclei show the most 
rapid movement of any hitherto observed, with the exception of the 
heads of spermatozoa. Treatment with methyl green demonstrates 
that they possess an unusually dense and deeply staining chromatin. 

Fine subdivision of the testis of winter-frogs under sugar-solution 
yields a milky fluid containing large numbers of isolated spermatozoa, 
together with an organic debris consisting chiefly of partially de- 
stroyed cells and cell-aggregates. The vibratile movements of the 
spermatozoa are arrested by the sugar-solution, the substance of the 
tails apparently undergoing liquefaction, so that under examination 
with high powers the spermatozoa appear finally to consist essentially 
of isolated sperm-heads. Undoubtedly, however, a thin film of cyto- 
plasmic matter remains present. Special mention should be made of 
certain frequently obtained cell-aggregates, which consist of Sertoli- 
cells with sperm-heads still attached. The behavior of these aggre- 
gates is peculiarly interesting and important from our present 
standpoint. 

The spermatozoa constantly exhibit an active migration in the 
direction of the negative stream.! The direction of this movement is 


* Other authors have incidentally observed and remarked upon the strong 
tendency of spermatozoa to travel with the negative stream under conditions 


Electrical Convection of Certain Free Cells and Nuclet. 279 


invariable and its rate is very uniform; in these respects spermatozoa 
present a marked contrast to leucocytes which, as shown above, ex- 
hibit great individual variability in respect to both direction and speed 
of migration. This constancy of behavior is evidently related to the 
constancy of organization of these cells, all normal spermatozoa being, 
so far as we can observe, almost exactly alike in size, structure, and 
chemical composition. The rate of migration is rapid (ca. 2 mm. 
per minute). This is the more remarkable when we consider that 
the shape of a sperm-head is such as to make its surface-area (and 
hence the frictional resistance to transfer through the medium) much 
larger than that of a spherical body of even considerably greater vol- 
ume. The surface-area of such a sperm-head, regarding it as a cylin- 
der with plane circular extremities and of the dimensions 30 » by 3 
# is approximately 300 square microns; its volume is approximately 
212 cubic microns. In the case of a spherical nucleus 10 yw in 
diameter (e. g. one of the thymus-nuclei considered above), surface 
and volume are respectively 314 square microns and 524 cubic 
microns. Thus for every unit of volume the sperm-head presents to 
the medium a surface of contact approximately two and one-half times 
greater than that presented by the most rapidly moving free nucleus. 
Nevertheless in actual rate of migration the two differ relatively 
slightly, their velocities being indeed very similar (1500 to 2000 pu 
per minute) under the above conditions. From this rough estimation, 
therefore, the inference seems clear that the substance of which the 
sperm-head is composed carries a negative charge of considerable 
potential. This confirms the expectation which we were led to form 
above from more general considerations. 

A few large free cells, as a rule, in a more or less injured condition, 
are usually found among the organic debris in the fluid. Their re- 
action varies; they are frequently but not invariably positive. These 
relatively voluminous cells evidently correspond to the Sertoli-cells 
or foot-cells of the testis. Frequently they are obtained with a bundle 
of spermatozoa still attached (Fig. 1); it is then found when the 
current is made that the aggregate typically assumes the position 
represented in the figure, with the spermatozoa on the side directed 
toward the anode; the whole mass is then usually dragged slowly in 
the direction of the negative stream. On reversal of the current- 
direction, the entire aggregate slowly turns about through 180° and 


similar to the above. See HERMANN: Archiv fiir die gesammte Physiologie, 
1885, xxxvii, p. 459. 


280 Ralph S. Lutte. 


adopts the reversed position. At times the movement is of such a 
nature that the Sertoli-cell moves nearer the cathode, while the bundle 
of spermatozoa approaches the cathode; such a movement is prac- 
tically one of rotation about a vertical axis, the position of the entire 
aggregate remaining almost stationary. At other times the bundle 
of sperm-heads tapers in such a way as to render the entire mass 
roughly club-shaped (Fig. 1). It is evident that such a mass, if 
under the influence of an attractive force acting equally on all its 
parts, would move through a resisting medium blunt end foremost, 
since at that region the ratio of surface to volume is least. Actually, 
however, when the aggregate travels toward the anode, the tapering 
end is always found to assume the fore- 
most position. This proves very clearly 
that the bundle of sperm-heads is much 
more strongly attracted than the body of 
the Sertoli-cell toward the anode. In 
other words, the sperm-nuclei are nega- 
tively charged relatively to the body of 
the cell. It is possible that consideration 
of this fact may throw light upon the 
nature of the peculiar relations subsist- 
ing between these two kinds of cells. 
Blood-corpuscles and spermatozoa are 
Ficure 1, — Orientation and examples of free cells that may be exam- 
direction of movement of ined in an uninjured condition. In the 
roughly club-shaped aggre- Ain : 
gate, consisting of Sertoli-cell CaS€ of the remaining tissues whose prop- 
with attached sperm-heads. erties I have examined, the histological 
elements are unavoidably more or less 
injured by the mechanical process of teasing. In all probability 
such injury produces alterations in the normal electrical behavior 
of the tissues (just as injured muscle becomes electro-negative 
towards uninjured muscle), hence relatively little stress can be laid 
on the results of the following experiments on portions of fresh tis- 
sues. The majority of such detached fragments exhibit a negative 
response: thus portions of fresh kidney tubules are found con- 
stantly negative; isolated ciliated cells of the oral epithelium are 
also negative, though slightly so; the same is true of such bodies as 
fat-globules, the majority of granules from the liver-cells, and yolk 
granules from the unripe ovary. On the other hand, fresh involun- 
tary or heart muscle gives a positive response; if a freshly teased 


} 


Electrical Convection of Certain Free Cells and Nuclet, 281 


fragment of muscle is so disposed that the direction of the fibres is 
perpendicular to the current-lines, the loose fibrils at the extremity 
of the fragment are found constantly to exhibit a movement in a pos- 
itive direction (towards the cathode) whenever the current is made 
or its direction changed. In the case of voluntary muscle, the make 
of the current is typically followed by an active contraction of the 
cathodal side of the fibre, producing a curvature towards that pole; 
this obscures whatever movement may be due to simple electrical 
convection, leaving some doubt as to the actual direction of the 
latter. Small fragments of muscle frequently exhibit a negative 
direction of movement; this, however, is in all probability due to 
post-mortem alterations of the muscle-substance. 


SUMMARY AND CONCLUSIONS. 


On generalization from the above facts (so far as this is at present 
admissible) it appears (1) that free nuclei exhibit a strong tendency 
to migrate with the negative stream, and (2) that this tendency is 
strongest in those nuclei in which the proportion of nucleic acid is 
highest. These facts confirm the inference that the colloidal particles 
composing nuclear chromatin carry negative charges. Cells with 
voluminous cytoplasm tend, on the other hand, to exhibit the reversed 
direction of migration, indicating a prevailingly opposite condition of 
electrification of the cytoplasmic colloidal particles. This difference 
in electrical behavior corresponds to the general difference in the 
chemical properties of chromatin and cytoplasm and is in agreement 
with the demands of the above theory. 

On considering the behavior of the chromatin in dividing cells, we 
meet with what must be regarded as additional evidence that this 
substance is composed of electrically charged particles. The move- 
ments of the chromosomes during mitosis are especially suggestive of 
the action of electrified particles; the relative positions assumed by 
these bodies with the cell, both while in the equatorial plate and after 
the longitudinal division is complete, seem to indicate very clearly 
the existence of a mutually repellent action between neighboring 
chromosomes similar to that which obtains between similarly charged 
bodies. The spiral form so characteristic of the chromatic filament 
before its segmentation, is open to the same interpretation; the ad- 
jacent portions of a filament composed of a linear series of similarly 


282 Ralph S. Lithe. 


charged colloidal particles! (here the “chromomeres”) must also 
repel one another; hence the entire filament, if confined within a 
limited space —as is here the case, owing to the presence of a nuclear 
membrane — will naturally tend to assume a coiled or spiral form, 
since this is precisely the form in which the adjacent portions of the 
filament are as far as possible removed from direct contact with one 
another. It is possible to proceed still further, and to demonstrate 
that such a filament will divide equally and /ongitudinally, if division 
is regarded as the result of a lowering of the surface-tension of each 
particle following upon an increase in the expansive tension of its 
surface electrical charge.2 The longitudinal division of the chromo- 
somes, hitherto interpreted on essentially teleological grounds (follow- 
ing Roux and Weismann), appears on this theory susceptible of 
simple and adequate physical explanation. To pursue these con- 
siderations in further detail is, however, out of place in the present 
paper, and a more complete discussion is deferred. It is sufficient 
for the present to point out that electrical theories of mitosis, besides 
being an almost necessary outcome of our present knowledge of the 
colloidal constitution of living matter, are accorded the strongest 
support by a consideration of many of the most significant features of 
the process itself. 


1 For an accurate study of the structure of the chromosome, and the nature of 
its longitudinal division, see A. BRAUER: Archiv fir mikroskopische Anatomie, 
1893, xlii, p. 153. Compare Wiitson: The Cell, Second Edition, p. 112, where 
figures from HERMANN and FLEMMING are also given. The chromatic filament is 
at present usually regarded as composed of numerous minute granules of chromatin, 
the chromomeres (here identified with colloidal particles), imbedded in a homo- 
geneous ground substance (linin), and serially arranged ina single row like a chain 
of beads. The longitudinal splitting of the entire chromosome is held to be the 
end-result of the fission of each granule along a single division plane which lies 
parallel to the long axis of the chromosome. Brauer finds in the spermatocytes of 
Ascaris megalocephala a double longitudinal splitting of the chromatic filament, 
each granule dividing into four (probably by two successive divisions) which 
always lie ina single plane perpendicular to the length of the filament; they are 
of equal size and arranged to form the corners of a square (::). This behavior is 
strongly suggestive of the division of a colloidal particle under the influence of its 
surface electrical charge and of the effects of mutual repulsion in keeping the 
products of division apart. 

* For a discussion of the influence of the surface-tension of the colloidal 
particles and its alteration by electrical influences in determining the state of sub- 
division, see BREDIG: Loc. cit., pp. 15-16. 


Electrical Convection of Certain Free Cells and Nuclet. 283 


SUMMARY. 


1. Isolated cells and nuclei suspended in cane-sugar solution, 
through which an electrical current is passed, migrate in some cases 
with the negative stream, in others with the positive stream. 

2. The majority of such structures migrate with the negative 
stream; this tendency is especially strong in free nuclei and struct- 
ures consisting chiefly of nuclear matter. 

3. The speed of migrations of sperm-heads, thymus-nuclei, and 
lymphocytes decreases in the order named, showing a parallelism 
with the decrease in staining property or degree of acidity of the 
chromatin. 

4. Cells with voluminous cytoplasm — large leucocytes, many red 
blood-corpuscles, involuntary muscle-cells, and in some cases Sertoli- 
cells — tend to move with the positive stream. 

5. The possibility is pointed out that this difference may be due to 
a general contrast in the electrical properties of the colloids com- 
posing nuclear chromatin and cytoplasm respectively; a few of the 
possible consequences of such a condition are briefly indicated. 


AN EXPERIMENTAL “STUDY OF THE CHEMIC Am 
PRODUCTS OF BACILLUS COLIZFCOMMUENTS 
AND BACILLUS LACTIS AEROGENES.! 


By LEO, ff, VRETTGER: 


(Research Scholar of the Rockefeller Institute.) 


[From the Sheffield Laboratory of Physiological Chemistry, Vale University. | 


HE colon bacillus and bacillus lactis aerogenes have been found 

to be of common occurrence in the human intestine. Accord- 

ing to Escherich ? and others, the bacillus lactis aerogenes predomi- 
nates in the upper end of the small intestine in persons living on a 
milk diet, and especially in infants; in the lower end, on the other 
hand, it is outnumbered by the colon bacillus. Further, a disappear- 
ance of milk sugar from the contents of the intestine is accompanied 
by a disappearance of the lactis aerogenes organisms, or a decrease 
in their number. At the same time, there is found a relative increase 
in the number of colon bacilli. As the latter organisms are univer- 
sally present in larger or smaller numbers in the human intestine, 
their mere presence may not be of much significance. But the ques- 
tion arises: May not these organisms, by virtue of their excessively 
great numbers, or of a peculiar kind or degree of activity, be the cause 
of certain disturbances in the animal body? If there is such a 
causal relationship between the colon bacillus and certain diseases, 
it is probable, in the light of recent investigations, that the immedi- 
ate cause of the disturbances must be sought in the chemical pro- 
ducts of the micro-organisms. Whether such a relationship exists is 
very difficult to demonstrate, and is as yet undetermined. It is of 
interest, however, to ascertain as far as possible the different pro- 
ducts of decomposition, the relative quantities in which they arise, 
and the conditions of formation. Lastly, since the two micro-organ- 


isms under discussion are so closely related in their morphological 
' This research was carried on with the aid of an appropriation from the 
Rockefeller Institute for Medical Research. 
* ESCHERICH: Fortschritte der Medicin, 1885, iii, Nos. 16 and 17, pp. 515 and 
547. (Abstracted in MALy’s Jahresbericht, 1885, xv, p. 513.) 
284 


——— + 


Bacillus Coli Communis and Bacillus Lactis Aerogenes. 285 


characteristics and distribution, a comparative study of the chemical 
products which they elaborate seems desirable. 

Organisms employed. — The colon bacillus! which was chiefly used 
in this work was isolated from the intestinal contents of a patient 
suffering from pernicious anzmia, where it was found to be the pre- 
dominating organism. For comparison, an ordinary stock culture of 
bacillus coli communis was employed. The bacillus lactis aerogenes ” 
was obtained directly from the intestine of an infant. 

Culture media.— The culture media, with the exception of the 
crystallized egg-albumin, were prepared in the laboratory of Dr. E. K. 
Dunham of New York. They were: 

a. Dunham’s peptone solution (1 per cent Witte’s peptone and 5 
per cent common salt). 

b. Solutions of crystallized egg-albumin® with varying amounts of 
different salts (sodium chloride, sodium hydrogen phosphate, mag- 
nesium sulphate, sodium aspartate, etc.). 

c. Mixtures of finely chopped meat and coagulated egg-white sus- 
pended in water. They were prepared as follows: One pound of lean 
chopped beef and the whites of six eggs were employed for each litre 
of water used. The beef was mixed with half of the water, and the 
mixture heated to boiling, with sufficient stirring to prevent burning. 
The egg-whites were mixed with the other 500 c.c. of water and 
slowly added (with constant stirring) to the boiling meat mixture. 
After a brief period of thorough heating, the whole was made faintly 
alkaline to litmus paper, diluted with water to the original volume 
and sterilized. 

Bacterial products. — The bacterial products sought for were albu- 
moses and peptone, tyrosin, leucin, tryptophan,* indol, skatol, phenols, 
aromatic oxy-acids, skatol-carbonic acid, hydrogen sulphide, mercap- 
tan, and diamines (putrescin and cadaverin). 

Conditions of growth and analysis. —In order to imitate to a certain 
degree the conditions which exist in the intestines, the culture 
materials were kept in an atmosphere of hydrogen, by allowing a 
slow current of the gas to pass through them. The decomposition 
was allowed to take place at incubator (body) temperature. 

1 Isolated by Dr. E. K. DUNHAM of New York. 

2 Likewise isolated by Dr. DUNHAM. 

8 Pure solutions of egg-albumin were found to form very little precipitate on 
sterilization with steam. For methods of preparing crystallized egg-albumin, see 
Hopkins and Pinkus: Journal of physiology, 1898-99, xxiii, p. 130. 

4 See Hopkins and Cote: Journal of physiology, 1901-02, xxvii, p. 426. 


286 Leo F. Rettger. 


The egg-meat mixture was the most favorable of the culture media 
employed. In this medium decomposition was rapid and complete. 
In the peptone-salt solution the chemical changes were very slow, 
although the bacteria seemed to multiply fairly rapidly. In the 
solutions of pure egg-albumin and salts not even bacterial develop- 
ment was apparent to any marked degree, and this medium had to be 
rejected. Similar results were obtained with solutions of sodium 
caseinogenate.} 

After definite periods of incubation further growth was checked. 
The bouillon cultures were filtered into sterile bottles through a 
Chamberland filter, and the filtrates mixed with small quantities of 
chloroform. The cultures on solid media were kept in a cool place, 
and subjected to analysis as soon as possible. In some instances 
small quantities of chloroform were added. For analysis, comparable 
quantities of material were taken. The egg-meat cultures were not 
filtered, but were thoroughly shaken immediately before examination, 
in order to obtain a homogeneous suspension. Further details re- 
garding the methods of analysis are given in the appendix. 

While the results obtained in this study are essentially qualita- 
tive, an attempt was made to obtain an approximate quantitative idea 
regarding the occurrence of certain of the decomposition products of 
the organisms, under varying conditions. In the case of indol and 
skatol, comparative coloration and dilution methods were used. By 
employing definite quantities of distillate, and diluting with known 
quantities of water, regular gradations were obtained in the colors 
which are produced by nitric acid (indol) and sulphuric acid (skatol). 
For the sake of convenience and uniformity, the phenols were esti- 
mated in an analogous manner (Millon’s reagent). As no satis- 
factory methods exist for the determination of aromatic oxy-acids 
and skatol-carbonic acid, a coloration method was likewise employed 
for them. 

In every trial with the liquid media (bouillon and egg-albumin solu- 
tions) the amount of decomposition which occurred was slight; and 
with the exception of the formation of indol, skatol, and aromatic oxy- 
acids, the chemical changes were so insignificant as to be regarded 
within the limits of error. Small quantities of indol, skatol, and 


' TAYLOR has recently shown that Bacillus coli communis fails to produce 
extensive decomposition in casein, while Proteus vulgaris, on the other hand, 
decomposes it energetically. Zeitschrift fiir physiologische Chemie, 1902, xxxvi, 
p- 487. 


Bacillus Coli Communis and Bacillus Lactis Aerogenes. 287 


aromatic oxy-acids were present, however. Cultures of varying ages 
(one to four weeks) were employed. The following discussion will 
therefore be confined to the products occurring in the egg-meat 
mixtures. 


GENERAL RESULTS. 


Bacillus coli communis. — Within two to three weeks the egg-meat 
mixtures had undergone very marked transformation. The mass of 
solid matter was extensively disintegrated, leaving only a_ small 
residue of coarse particles. A very strong and disagreeable odor, 
more intense than in any of the older cultures, was emitted on open- 
ing the flasks. 

Indol was present in large quantity; 10 c.c. of the distillate 
(500 c.c.) gave a deep red or cherry-red color with nitric acid. This 
color was still perceptible after dilution with ten volumes of water. 
Only a small amount of skatol was found, the distillate (500 c.c.) 
giving only a pink color with sulphuric acid. Phenols were present 
in considerable amounts, giving a distinct color reaction after five 
dilutions of the distillate (500 c.c.). The ether extracts of the 
decomposition fluids had a very strong odor of aromatic oxy-acids. 
On chemical examination, a pronounced reaction was obtained in the 
first extraction fluid (250 c.c.), while in the second fluid (500 c.c.), 
the reaction was still quite.marked. (See appendix, p. 291). Onlya 
small percentage of the acids appeared to be para-oxy-pheny] acetic, 
the remainder being largely para-oxy-phenyl propionic acid. 

Skatol-carbonic acid was present in large quantities, as indicated 
by the strong reactions with ferric chloride and hydrochloric acid. 
The reaction was marked even in the second extraction fluid 
WEOO ¢.c:). 

Two crops of leucin aud tyrosin crystals were obtained from the 
liquid. The first (principally tyrosin) weighed 3 grams; the second 
(principally leucin), 4 grams. On further purification, the character- 
istic crystals of each were obtained separately. 

Albumoses and peptone occurred in very small amount, and only 
slight reactions could be obtained for them. 

Tryptophan was very easily detected, and considerable purple pre- 
cipitate was formed with bromine water. An alcoholic extract of 
this precipitate possessed the characteristic color. 

Hydrogen sulphide and mercaptan were present in relatively large 


288 Leo fF. Retiger. 


quantity. The combined precipitates of mercury mercaptid obtained 
from 300 c.c. decomposition mixture weighed 0.2 gram. 

By the method described for diamines in the appendix, a crystal- 
line precipitate of about 0.3 gram was obtained ; it consisted of small 
fine needles of a light yellow color, melting at 60° C. Since putrescin 
and cadaverin both have a melting point of over 100° C., it was evi- 
dent that they did not constitute this precipitate, and were presum- 
ably absent from the putrefaction material. 

As the egg-meat cultures grew older, the transformation was still 
more marked. Flasks which had been incubated three to four weeks 
contained less of the coarse material than the preceding. No indi- 
cation of the presence of albumoses and peptone could be obtained. 
The amount of mercaptan was small, yielding a precipitate of only 
0.01! gram of mercury mercaptid. There was a decrease in the 
quantity of all the intermediate products, except tyrosin, leucin, and 
indol. 

In cultures still older, the disappearance of the molecularly large 
decomposition products was very conspicuous. In flasks which had 
been incubated ten weeks, a moderate amount of indol was present; 
no skatol or phenols, and only a trace of mercaptan was observed. 
The odor of the mixture was only slightly offensive. While tyrosin 
was still present in considerable quantity, very little leucin was ob- 
tained. There was no indication of albumoses or peptone. Aro- 
matic oxy-acids and skatol-carbonic acid were present in moderate 
quantities only, as compared with cultures two to three weeks old. 
No diamines were found. 

A decomposition mixture which was four months old was found to 
be thoroughly disintegrated, and of fluid consistency. The odor was 
only slightly offensive. While no skatol or phenols could be detected, 
the amount of indol in this instance was very large. Not a trace of 
mercaptan could be detected. A moderate amount of aromatic oxy- 
acids and of skatol-carbonic acid was still present. No albumoses or 
peptone were obtained, and leucin and tyrosin were absent, or present 
in traces only. On final concentration of the culture liquid a residue 
was obtained which contained comparatively little organic material. 
Evidently the decomposition in this instance was very complete. 

A comparison of the products of this isolated colon bacillus with 
those of another (stock) culture, grown in the same media, and under 
the same conditions, showed a striking similarity in the mode of 
decomposition inaugurated by the two organisms. The transformation 


Bacillus Coli Communis and Bacillus Lactis Aerogenes. 289 


of the culture media, however, was much slower when the latter 
organism was employed, and consequently the amounts of the differ- 
ent products in the early stages of decomposition were smaller than 
when the cleavages were brought about by the organism recently 
obtained from the intestinal tract. 

Bacillus lactis aerogenes. — The production of mercaptan, skatol, 
phenols, aromatic oxy-acids and skatol-carbonic acid was much 
slower and less pronounced than in corresponding cultures of the 
colon bacillus. This was not due to any failure of the organism to 
decompose the proteid material present, as the amounts of tyrosin 
and leucin, and even of indol, in such mixtures were large, and the 
mixtures themselves had undergone disintegration. 

In cultures two to three weeks old, no mercaptan could be detected. 
The odor was not at all offensive as compared with corresponding 
colon cultures. Very little skatol and no phenols were observed. 
Indol was detected in considerable quantity, while the quantities of 
aromatic oxy-acids and skatol-carbonic acid were very smal]. Leucin 
and tyrosin were quite abundant. The presence of albumoses and 
peptone was indicated, although in small amounts. 

As decomposition continued the quantities of the different pro- 
ducts gradually increased. In cultures eight to ten weeks old a 
small amount of mercaptan was observed. There was a trace of 
skatol and phenols, and a moderate quantity of aromatic oxy-acids 
present. Indol was very abundant; and a large mass of tyrosin 
crystals with only little leucin was obtained. Albumoses and peptone 
were present in very small amount, while the presence of tryptophan 
was readily demonstrated. 


SUMMARY. 


1. Bacillus coli communis and Bacillus lactis aerogenes fail to 
bring about very marked decomposition in peptone-bouillon. On the 
other hand, an egg-meat mixture undergoes rapid and extensive 
transformation. 

2. Common products of the colon bacillus are indol, skatol, phe- 
nols, aromatic oxy-acids, skatol-carbonic acid, hydrogen sulphide, 
mercaptan, tyrosin, leucin, and tryptophan. Albumoses and peptone 
are present in very small amounts. This fact is contrary to general 
belief. It has been assumed that albumoses and peptone occur 
abundantly among putrefaction products. Fermi and Pampersi’ 


1 FERMI and PAMPERSI: Jahresbericht fiir Thierchemie, 1897, xxvii, p. 827. 


290 Leo Ff. Rettger. 


go to the other extreme, however, and contend that bacteria are 
unable to peptonize proteids. A more plausible idea presents 
itself: namely, that bacteria do peptonize proteids, but that the 
albumoses and peptone formed are immediately broken up further 
by the organisms or their enzymes, and therefore are detected with 
great difficulty. Diamines were not found in the putrefaction mix- 
tures. This fact is in harmony with the observations of Garcia! and 
others who failed to find diamines in ordinary intestinal putrefac- 
tion. 

3. Bacillus coli communis causes more rapid and more profound 
decomposition than Bacillus lactis aerogenes. While the former pro- 
duces its optimum results within two to three weeks (under favorable 
conditions), the lactis aerogenes organism requires eight to ten weeks 
to bring about comparable results. Indol is a product of both organ- 
isms, while the colon bacillus alone produces mercaptan and phenols 
during the first few weeks. 

4. When the bacterial digestion progresses beyond a certain point, 
the intermediate products (peptone, amido-acids, indol, etc.) gradu- 
ally disappear from the mixtures. Indol may, however, persist for a 
long time. This disappearance of the more intermediate products is 
evidently due to their further cleavage and the formation of still 
simpler bodies, yielding ultimately carbon dioxide, water, methane, 
etc. 

It may be objected that in the intestine the period of decomposi- 
tion is much shorter than in these experiments, and that therefore 
the results here obtained do not illustrate what actually occurs there. 
This is in large part true; it is not feasible, however, to examine by 
chemical means putrefaction materials which are only twenty-four to 
forty-eight hours old, since the various characteristic products occur 
in such small amounts as to escape detection. In the intestine, on 
the other hand, the conditions are all favorable to rapid decomposi- 
tion. Here putrefaction is doubtless facilitated by the presence of 
digestive juices and their products, by the peculiar character of the 
intestinal contents, and by the movements of the alimentary canal. 


I am indebted to Professors C. A. Herter, E. K. Dunham, and 
L. B. Mendel for the valuable assistance which they have given me 


in this investigation. 


1 GARCIA: Zeitschrift fiir physiologische Chemie, 1893, xvii, p. 588. 


— 


val 


Bacillus Coli Communts and Bacillus Lactis Aerogenes. 291 


APPENDIX. — METHODS OF ANALYSIS. 


a. Liquid media. — Zofal acidity. — Titration with #% KOH (phenolphtha- 
lein, indicator). 

Volatile acids. — Distillation with phosphoric acid, and titration of 
distillate. 

Ammonia. — Distillation with magnesium oxide, and titration of dis- 
tillate with #4; H.SO,. 

Albumoses. — Precipitation with zinc sulphate,' and determination of 
nitrogen in precipitate by the Kjeldahl-Gunning method. 

Peptone. — Precipitation of the albumose-free solution with bromine,” 
and determination of nitrogen by the Kjeldahl-Gunning method. 

Indol, skatol, phenols, and aromatic oxy-acids. — Distillation with steam, 
etc., as described for solid media. 

b. Solid media.— Three different portions from each culture flask are em- 
ployed: one for the isolation and detection of indol, skatol, phenols, 
aromatic oxy-acids, skatol-carbonic acid, tyrosin, leucin, albumoses and 
peptone, and tryptophan — (1) ; another for the study of hydrogen sul- 
phide and mercaptan — (II); and the third for the isolation and identifi- 
cation of diamines — (III). 

I. Jndol, skatol, phenols, etc. — Portions of 400 c.c. of the decomposi- 
tion mixture are diluted with an equal volume of water, and after the 
addition of 5 c.c. of dilute sulphuric acid, are distilled with steam until 
600-700 c.c. of distillate have formed. ‘The distillate is made alkaline 
with potassium hydrate, and again distilled with steam until 500 c.c. of 
liquid have collected — (A). Indol and skatol are in the distillate. The 
last distillation residue is saturated with carbon dioxide gas,’ and dis- 
tilled until 500 c.c. of liquid have collected — (B). Phenols are in this 
distillate. 

For the estimation of indol, definite quantities of distillate A are diluted 
with known quantities of water until such a degree of dilution is obtained 
that the color reaction with nitric acid is only slight* The amount of 
dilution required is an index to the approximate quantity of indol present. 
Skatol gives no color reaction with nitric acid; it is estimated in a manner 


1 Method of BOMER. See ZuNz: Zeitschrift fiir physiologische Chemie, 
1899, xxvii, p. 220. 

2 ALLEN: Commercial organic analysis, 1898, iv, p. 320. 

8 Hopre-SEYLER und THIERFELDER: Handbuch der physiologisch- und 
pathologisch-chemischen Analyse, 1893, p. 158. 

4 For color reactions of indol, skatol, phenols, aromatic oxy-acids, and skatol- 
carbonic acid, see SALKOWSKI’s Practikum der physiologischen und_patho- 
logischen Chemie, 1900. 


292 


Leo F.. Rettger. 


analogous to that of indol, sulphuric acid being used instead of nitric to 
produce the color reaction. 

For the estimation of phenols, distillate B is diluted in the same manner 
as A, and the color reaction is brought about by Millon’s reagent. 

Aromatic oxy-acids and skatol-carbonic acid. —'The liquid remaining 
after the first distillation is filtered, concentrated, and again filtered ; it is 
then extracted with ether, to remove the aromatic oxy-acids and skatol- 
carbonic acid. ‘The combined ether extracts are filtered, and the ether 
removed by evaporation. The oily residue is extracted with warm 
water, and the solution filtered and distilled with steam.’ The distillation 
residue is again concentrated and extracted with ether. After evaporation 
of the ether, the residue is extracted twice with warm water — 250 c.c. 
and 500 c.c. Both of the aqueous extracts are examined for aromatic 
oxy-acids and skatol-carbonic acid. 

Zyrosin, leucin, albumose, peptone, and tryptophan. — The fluid residue 
remaining after the removal of the aromatic oxy-acids and skatol-carbonic 
acid is concentrated to the point of crystallization. After cooling, the 
crystalline mass is examined for tyrosin and leucin.? A second crop of 
crystals is obtained in the same manner. 

After the removal of tyrosin and leucin, the syrup is treated with four 
to five volumes of 95 per cent alcohol; albumoses and peptone are 
precipitated, along with certain salts. The precipitate is washed with 
alcohol, dissolved in water, and the solution tested for albumoses and 
peptone.’ 

For the isolation and detection of tryptophan, the alcoholic filtrate and 
washings from the last precipitate are concentrated in order to remove 
the alcohol, acidified with acetic acid, and treated with 1o-12 c.c. of 
bromine water. Tryptophan is precipitated as a purple bromine com- 
pound.' which dissolves in alcohol, imparting the characteristic color to 
the solution. 

Il. Hydrogen sulphide and mercaptan.—In order to eliminate any 
error arising from the presence of hydrogen sulphide formed in the 
sterilization,® a current of air or hydrogen should be passed through the 
culture medium before inoculation, and while the medium is being 
warmed at 50°-60° C. 


' To remove any traces of phenol that may still be present; see SALKOWSKI: 
Zeitschrift fiir physiologische Chemie, 1879, ix, p. 8. 


9 


* See HAMMARSTEN: Textbook of physiological chemistry (translated by 


MANDEL), 1900, p. 62. 


3 


4 


5 


HAMMARSTEN: Loc. cit., p. 33. 
KURAJEFF : Zeitschrift fiir physiologische Chemie, 1898-99, xxvi, p. 501. 
RETTGER: This journal, 1902, vi, p. 450. 


Bacillus Coli Communts and Bacillus Lactis Aerogenes. 293 


Mercury cyanide precipitate. An equal volume of water is added to 
300 c.c. of the decomposition mixture in a litre flask. The flask is 
connected, on the one hand, with a small precipitating bottle containing 
a 3 per cent solution of mercury cyanide, and, on the other, with a wash 
bottle partly filled with permanganate solution. The decomposition 
material is then acidified with 10 grams of oxalic acid, and warmed for 
two to three hours at 40°-50° C. During the warming a current of air 
is slowly drawn through the flasks. Hydrogen sulphide comes over first, 
and produces a black precipitate in the cyanide solution. Finally the 
mercaptan passes over, and its presence is indicated by a gray to light- 
green precipitate in the cyanide bottle. By replacing the mercury cyanide 
with fresh solutions, the precipitates with these two volatile bodies are 
obtained separately. They are washed, dried, and weighed.’ 

Mercaptan-isatin reaction.? — A small quantity of isatin is dissolved in 
concentrated sulphuric acid. On passing mercaptan gas through this 
solution it is colored grass-green. The reaction is a very delicate one, 
and is in no way interfered with by any of the other ordinary decompo- 
sition products. The method of detection is somewhat similar to the 
cyanide method ; the isatin solution is substituted for the cyanide, and 
calcium chloride drying tubes must be employed to prevent any mois- 
ture from entering the isatin solution. 

III. Diamines. — The benzoylation method, as described by Garcia,* 
was employed. 


ADDITIONAL REFERENCES. 


BAUMANN and BrieEGER: Zeitschrift fiir physiologische Chemie, 1879, iii, 
p- 149. 

BRIEGER: Zeitschrift fiir physiologische Chemie, 1882-83, vii, p. 274; 1883-84, 
viii, p. 306; 1885, ix, p. I. 

SALKOWSKI: Zeitschrift fiir physiologische Chemie, 1883-84, viii, p. 417; 1885, 
ix, p: 5. : 

HirScHLER: Zeitschrift fiir physiologische Chemie, 1886, x, p. 306. 

NENCKI: Monatshefte fiir Chemie, 1889, x, p. 506. 

SELITRENNY: Monatshefte fiir Chemie, 1889, x, p. 908. 

BLUMENTHAL: Zeitschrift fiir klinische Medizin, 1895, xxviii, p. 223. 

EMMERLING: Berichte der deutschen chemischen Gesellschaft, 1897, xxx, 
p- 1863. 

Morris: Archiv fiir Hygiene, 1897, xxx, p. 304. 


1 NENCKI and SIEBER: Monatshefte fiir Chemie, 1889, p. 526. 

2 NIEMANN: Archiv fiir Hygiene, 1893, xix, p. 126; also BAUER: Zeitschrift 
fiir physiologische Chemie, 1902, xxxv, p. 346. 

3 GarcIA: Zeitschrift fiir physiologische Chemie, 1893, xvii, p. 571. 


ELECTRIGAL POLARITY IN THE HYDRO: 


By ALBERT P. MATHEWS. 
[From the Hull Physiological Laboratories of the University of Chicago.) 


T has long been known that many animals and plants possess a 
marked physiological polarity, but the cause of that polarity is 
still undetermined. As an extended discussion of this subject is to 
be found in Morgan’s treatise on ‘“‘ Regeneration,” it is unnecessary 
to enter into detail here, but the following particular examples may 
be mentioned as more especially bearing on my own work. The 
point of most rapid regeneration in pieces of Planaria and other 
organism appears to be determined in part by the shape of the piece, 
being sometimes in the centre of the cut surface, sometimes at 
one side of the centre; and the differentiation of the new tissues 
appears to depend, in part at least, on the position of the surface of 
regeneration in relation to the whole piece. The same tissues will, 
if that surface be anterior, regenerate a head; if posterior, a tail. In 
the hydroid, Parypha, a piece of the stem which shows no optical 
difference between the two ends, nevertheless regenerates the polyp 
more rapidly from the polyp end of the piece, than from the stolon 
end. In grafting many plants, it has been found far more difficult to 
graft two stem surfaces together than stem with root. 

The marked resemblance of animal and plant polarity to that of a 
magnet suggested that it might be possible to detect and measure 
the amount of this polarity by purely physical means, on the assump- 
tion that the polarity was electrical in nature, or gave rise to electrical 
differences. This hypothesis was supported by several facts. It is 
well known that electrical differences exist between two portions of 
an organ, or animal, which are unequally active. Such electrical 
differences have been observed in glands, in epithelia, in nerves, in 
muscles, and in other tissues. If there is a difference in the rate of 
chemical or physical change between the two ends of a hydroid, it 
should be possible to detect it electrically. Furthermore, in plants 


' MorGAN: Regeneration, Macmillan and Co., New York, 1901. 
294 


Electrical Polarity in the Hydrords. 295 


similar electrical differences have been observed between the mid- 
rib and the edges, and between the upper and lower surfaces of leaves, 
and have been studied by Burdon-Sanderson! and others. In the 
roots of the spinal nerves, Mendelssohn ? has discovered that a small 
but definite difference in potential exists between the two cut trans- 
verse surfaces of a piece of nerve. Ina piece of the anterior roots, 
the peripheral surface is always negative to the central; while in the 
posterior roots, the peripheral surface is positive toward the central 
surface. In other words the spinal nerves are definitely polarized, and 
although the two surfaces look alike, they differ electrically. Finally 
Du Bois-Raymond? long ago showed that in a prism of muscle the 
point of greatest negativity was determined by the shape of the piece 
of muscle, as will be pointed out farther on. These facts are appar- 
ently so strikingly parallel to the facts of regeneration, and the physi- 
ological effects of such small differences in electrical potential are so 
well recognized, that it seemed not impossible that animals might 
show an electrical, corresponding to and causing their physiological, 
polarity. 

Observations were made chiefly on the tubularian hydroid, Parypha, 
though several other hydroids were examined. The hydroids are 
very suitable for such observations, owing to their well marked polarity 
and their lack of motility. The electrodes used were zinc-zinc sul- 
phate, clay electrodes of Du Bois-Raymond, the clay being moistened 
with 3 sodium chloride solution, making them isosmotic with the 
sea-water. The electrodes were isoelectric. The current: was de- 
tected by means of a delicate capillary electrometer designed by 
Porter. 

When a piece of the stem of Parypha one half or one inch long was 
placed with its cut surfaces on the two electrodes, it was found invari- 
ably that the polyp surface was electro-negative to the stolon surface. 
A current passed through the electrometer from the stolon toward 
the polyp end. The current was always in this direction in all stems 
examined, and was invariably reversed through the electrometer 
when the piece of stem was reversed on the electrodes. The amount 
of this current is variable, depending on the age of the piece and the 


1 BURDON-SANDERSON: Cited from MENDELSSOHN, Dictionnaire de physi- 
ologie, edited by RICHET, vol. v, p. 386. 

2 MENDELSSOHN: Dictionnaire de physiologie, edited by RICHET, vol. v, 
P> 335- 

8 Du Bols-RAYMOND: Cited from MENDELSSOHN: Loc. cit., Pp. 325. 


296 Albert P. Mathews. 


position it occupies in the stem. In pieces of old stem taken near 
the stolon, hardly any difference is to be detected. The difference 
is greater the fresher the hydroid. It is also greater in pieces of 
young stem taken with one surface just behind the polyp. If the 
polyp itself is crushed against one electrode, the cut surface being 
on the other, a maximum difference is obtained, the polyp being 
very strongly negative. The movement of the meniscus in the 
electrometer was measured by a micrometer eye-piece and was, in the 
most favorable instances, in pieces one inch long, about one-third 
the maximum movement, which the same electrometer showed when 
the electrodes were on the cut and longitudinal surface of the frog’s 
sciatic nerve. The current was probably about 0.005 volt. 

Other hydroids, z. ¢., Pinnaria and Campanularia, showed the same 
negativity of the polyp surface, but in these forms the amount of 
difference was a good deal less that in Parypha. In the hydroids, 
therefore, the more rapid regeneration occurs at the point of greatest 
negativity. 

Morgan observed that in Planaria or Fundulus the point of most 
rapid regeneration was apparently determined by the shape of the 
piece, or of the cut surface. If, for example, a planarian be cut at 
right angles to the long axis of the body, the most rapid growth 
occurs at the middle of the cut surface; if it be cut obliquely, the 
most rapid growth occurs at the acute angle. Similar facts are 
observed in the regeneration of the tail of Fundulus. These facts 
recall Du Bois-Raymond’s observations on the electrical relationships 
in a piece of muscle. Ifa muscle is cut perpendicularly to its longi- 
tudinal surface, the point of greatest negativity is to be found in the 
centre of the cut surface. If the muscle is cut obliquely, the point of 
greatest negativity is at the acute angle. 

I think the similarity of these observations to Morgan’s will be 
apparent. While it was impossible, owing to the small size of Pla- 
naria, to prove that the acute angle was negative to the obtuse in 
pieces cut obliquely, and unsatisfactory results were obtained also in 
Cerebratulus, I found that such electrical differences existed in oblique 
sections of the tail of Fundulus. Although not detected in Planaria, 
I believe it to be probable that the body of the Planarian does not 
differ essentially in this respect from that of Fundulus and the 
muscle prism, particularly as the regeneration phenomena resemble 
each other in the two cases. In Fundulus at any rate, and probably 
in Planaria, we find hence the same law as in the hydroids, ¢.¢., re- 


——— 


Electrical Polarity in the Flydrovds. 297 


generation takes place with greatest rapidity at the point of most 
marked negativity; and furthermore, the point of most marked nega- 
tivity is determined in part by the shape of the piece. 

I endeavored to test experimentally the conclusion that regeneration 
takes place most rapidly at the most negative point, by placing pieces 
of hydroid stems in sea-water and passing an electrical current through 
the water, with some of the stems turned with the polyp end toward 
the current, others turned with the stolon end toward the current, in 
the hopes of neutralizing or increasing the negativity of the polyp 
end. The experiments were interrupted by other work, and I cannot 
speak positively without further tests, which it is hoped may be tried 
next summer, but in many cases pronounced inhibition of develop- 
ment was obtained when the polyp end was turned toward the 
current. 

It may be asked how the fact that one point is negative toward 
another can influence the rate of regeneration. It is possible the 
process may be as follows: The original difference of potential be- 
tween the cut surface and the uninjured portion of the protoplasm is 
caused partly by the shock of the cut, partly by the new environment 
for the protoplasm of the cut surface. Just which point of the cut 
surface shall be most negative is determined by the position, of the 
point relative to the mass of the piece and the totality of uninjured 
and injured points. The electrical differences thus set up cause cur- 
rents which are closed through the surrounding cells, and these 
currents, though small, profoundly influence all the processes in these 
cells. 

It is not probable that the electrical disturbances here described 
are the accompaniment of injury only. On the contrary, it has been 
shown that they exist also in the intact organism. Every excess of 
action, every change in physical state of the protoplasm of any organ, 
or of any area in the embryo or in the egg produces, it is believed, an 
electrical disturbance. Electrical tensions are set up in the adult or 
throughout the embryo. This electrical disturbance, rendering one 
part of the animal negative or positive to another part, must cause 
electrical currents, z. ¢., the movements of ions in the surrounding 
cells or protoplasm. In my opinion, the importance of the electrical 
currents thus set up physiologically in the normal animal by varying 
activities in its organs, has not been properly recognized. These cur- 
rents probably play a larger part in the determination of rates of 
growth, in the orientation or polarization of the cells, and the differ- 


298 Albert P. Mathews. 


entiation of the organism, in its polarity, in other words, than has 
been supposed. That electrical currents do have such a profound 
effect on the state of protoplasm is shown by the well-known electro- 
tonic effects in muscle and nerve. A momentary exposure of a sym- 
pathetic nerve to the anode will often block conductivity at this point 
for minutes, or even hours. The directive action of a current on in- 
fusoria is known to all. The amount of current necessary to produce 
electrotonic effects is very small. Mendelssohn found that in nerve, 
0.0001 milliampere was sufficient. 

Bearing these facts in mind and remembering the undoubted 
existence of an electrical polarity in the hydroids, is it not possible 
that the directive ‘“ pull” exerted by one part of an animal on an- 
other, of which Morgan speaks, is really due to these differences of 
electrical potential between the parts of the organism? From this 
point of view, the observations of Roux on the varying degree of 
polarization in the different cells of the frog’s egg, and the trans- 
portation of material in the cells by electrical currents, derive 
additional importance. 

Finally, if the rate of regeneration is thus conditioned by the state 
of the protoplasm in the neighborhood of the injury, it ought to be 
possible to alter that state at will, and hence increase or decrease 
regeneration. The control of this process offers one of the most at- 
tractive fields of physiological research. 

The observations recorded in this paper were made at the Marine 
Biological Laboratory at Wood’s Holl. 


SUMMARY. 


1. A difference of electrical potential exists between the cut-sur- 
faces of the stems of the hydroids Parypha, Campanularia, and Pin- 
naria, the head, or polyp surface, being always negative to the stolon 
surface. The amount of this difference is variable, at its maximum 
being approximately one-third the current of injury in the frog’s 
sciatic nerve. 

2. Physiological polarity of these hydroids is hence accompanied 
by an electrical polarity, and may be quantitatively measured by 
physical means. 

3- The amount of the current depends on the position of the stem, 
whether the piece examined comes from the neighborhood of the 
polyp, or the stolon; whether old or growing; whether fresh or 


Electrical Polarity in the FHydrovds. 299 


dying; being at its maximum in fresh, growing stems when one sur- 
face is taken near the polyp. 

4. The point of most rapid regeneration corresponds here to the 
point of maximum negativity. 

5. In Fundulus, and probably also in Planaria, the point of greatest 
negativity also corresponds with the point of greatest regenerative 
power. The most negative point is determined, among other factors, 
by the shape of the piece, or the relation of the cut surface to the 
whole animal. Probably the point of greatest negativity determines 
the point of most rapid regeneration, and thus explains the relation- 
ship of regeneration to the shape of the piece as noted by Morgan. 

6. These facts indicate, in the author’s opinion, that the so-called 
physiological polarity of the embryo or adult is due, in a measure 
at least, to the electrical differences or currents set up by unequal 
degrees of activity in the protoplasm at different regions. These 
currents traverse the surrounding protoplasm or cells, and, like any 
constant current applied from outside, polarize the protoplasm or cells 
in a definite way, causing alterations in their metabolism, and in the 
distribution of the cell contents analogous to the electrotonic effects 
in muscle, nerve, and infusoria. 


THE IMPORTANCE OF MECHANICAL SHOCK IN 
PROTOPLASMIC:-ACTIVIERY. 


By A. P. MATHEWS anv B. R. WHITCHER. 
[From the Hull Physiological Laboratory of the University of Chicago.] 


VE shocks caused by the beating of the heart, mus- 

cular movements, the jar of walking, and the vibrations of the 
floors of buildings and vehicles of transportation are among the most 
constant stimuli of the body cells, and it is an interesting question to 
what extent they influence cell life. Meltzer! has especially discussed 
this question and contributed to our knowledge of the influence of 
agitation on animal cells and bacteria. A summary of previous work 
will be found in his paper. 

It might at first be thought that such small shocks as those caused 
by the beating heart would be of no importance in the life-processes of 
the body cells, but that even very small shocks may profoundly influ- 
ence cell metabolism is shown by the observations of one of us? on 
the star-fish egg, in which the most careful transfer of the eggs from 
one dish to another is often sufficient to set up artificial partheno- 
genesis in a certain proportion of eggs and by similar observations by 
Fischer® on the eggs of Amphitrite. That the unfertilized egg is in 
respect to its susceptibility to mechanical disturbance in a particularly 
unstable condition is perhaps true, but the following observations on 
the small and hardy eggs of Arbacia show that even after fertilization 
an exceedingly small mechanical disturbance is often sufficient pro- 
foundly to change the rate of development and the physical charac- 
ters of the embryos. As a result of our work we have become 
convinced of the probable truth of Meltzer’s opinion concerning the 
importance of mechanical shock in the life history of body and other 
cells. 

The question thus raised is of considerable practical importance. 
For example, what effect has the constant vibration of the floors of 


1 MELTZER: Zeitschrift fiir Biologie, 1894, xxx, p. 3. 
? MATHEWS: This journal, Igot, vi, p. 142. 
* FISCHER: This journal, 1902, vii, p. 301. 

300 


Mechanica Shock in Protoplasmic Activity. 301 


- 


mills on the length of life, the vital resistance, and physiological func- 
tions of mill-operatives? How far will mechanical jarring account 
for the digestive and vasomotor disturbances many suffer in railway 
travel? Are the motor-men and conductors of street-railways influ- 
enced by the repeated and violent shocks to which they are constantly 
subjected ? 


I. Tue EFFrect oF A SINGLE SMALL OR LARGE SHOCK ON THE 
DEVELOPMENT OF ARBACIA EGGS. 


The eggs were transferred from one dish to another both before 
and after fertilization, and were either squirted vigorously with a 
medicine dropper, or they were dropped into water from a height of 
from one inch to three feet, or they were carried carefully from dish 
to dish with a large-mouthed pipette, the end of which was placed 
under water before discharging the eggs. 

The general result obtained in a large number of experiments was 
as follows: In certain conditions of the eggs, particularly at the 
beginning and end of the egg-laying season, and apparently when the 
temperature was low, 7. ¢., 15°-19° C., the act of transfer, when per- 
formed even with the greatest care, will prevent practically the whole 
of the eggs from forming plutei. They develop into abnormal, opaque 
gastrule, often evaginated, often flat and with crenated edges, live for 
several days, swim, as a rule, at the bottom of the dish, and then die, 
generally developing no skeleton at all, or at the best but irregular 
spicules. At other times, however, particularly during the middle of 
the season, or when the temperature is high, z.¢., from 20°-25° C., 
the act of transfer, even when roughly performed, has no effect at all, 
or it may produce an acceleration of development. In other words, 
the eggs are more susceptible at some times than at others, and are 
apparently rendered more sensitive by an exposure to a temperature 
of 18°-20°. The exact relationship to temperature changes we hope 
to work out next summer, our results in this respect being still 
incomplete. 

The increased susceptibility of the eggs brought about by a low 
temperature confirms Greely’s! observations on Asterias, where he 
found that mechanical shock and cold together were more efficacious 
in causing artificial parthenogenesis than cold or shock alone. Our 
results parallel, also, observations made on motor nerves, which are 


1 GREELY: Personal communication, 1902. 


302 A. P. Mathews and B. R. Whitcher. 


similarly increased in their excitability by a low temperature. The 
_ most probable explanation of the action of mechanical shock on the 
egg substance we believe to be that already offered by one of us for 
nerves,! 7. ¢., that a single shock causes a partial gelation of the col- 
loids of the egg substance. It produces the same effect on the pro- 
toplasm as cold, and the two processes accordingly supplement each 
other. This conclusion is, we believe, strengthened by Mrs. Andrews’? 
observations on living protoplasm. She observed that mechanical 
shock caused a distinct change in the viscidity of the protoplasm of 
the choano-flagellates, a very small jar causing the collar to become 
rigid. I believe Mrs. Andrews has seen and described better than 
any one else the changes in the state of /zvzxg protoplasm, and has 
painted a more faithful picture of its extraordinary changes in vis- 
cosity, and of its constant flux. 

The following experiments will illustrate the general character of 
the results obtained on Arbacia: 


Experiment [/.— August 17. Temperature, 19° C. Arbacia eggs were fer- 
tilized at 2 P.M. 

Lot A. The eggs were left in the dish in which they were fertilized. 
They develop into normal gastrulae and plutei. 

Lot B. These were transferred, and squirted from a pipette into a 
second dish of water five minutes after fertilization. The early develop- 
opment corresponds to A; but they form gastrulz of irregular shape, 
become crenated on the edges, and form no skeletons except in a few 
isolated instances. 

Experiment X/X.— August 27. Temperature, 26° C. Arbacia eggs fertilized 
at 3.08 P.M. 

Lot A. Not transterred. Form normal plutei. 

Lot B.. Transferred carefully with a large pipette at 3.10 P.M. Form 
large normal plutei. 

Lot C. Dropped two inches into sea-water at 3.10. Many form 
abnormal gastrulze and plutei, but the majority are normal. 

Lot D. Squirted violently with a small pipette into sea-water. Form 
normal plutei like control. 

Experiment 1//.— August 17. Temperature, 19° C. A different female than 
Experiment II. Conditions and results the same as in II. 


' MATHEWS: Science, N. S., 1902, xv, pp. 492-408. 
* ANDREWS: Journal of morphology (supplement), 1897, xii, p. 112. 


Mechanical Shock in Protoplasmic Activity. 303 


II. THe EFrrect oF REPEATED SHOCKS AND SHAKING. 


The effects of vigorously shaking the fertilized or unfertilized eggs 
will, of course, depend on the violence and length of time they are 
shaken. No accurate machine for shaking was at our disposal, so that 
we had to shake with the hand. The eggs were brought into a test 
tube half-filled with sea-water, and this was shaken briskly back and 
forth from ten to twenty times. The results obtained were variable, 
like those of the preceding section, depending, so far as we could dis- 
cover, in part on the temperature and in part on some unknown con- 
dition of the egg substance. Certain results indicated that possibly 
the degree of aeration which the sea-urchins had had before being 
used was one of the determining factors. A further examination of 
this possibility we hope to make next summer. 

In general, our results may be summarized as follows: 

1. The unfertilized eggs are far more easily disorganized by shak- 
ing than the fertilized. This result was constant. If two lots of eggs 
were taken from the same female, one lot unfertilized, the other ferti- 
lized at least ten minutes earlier, the unfertilized eggs could be shaken 
to pieces before the fertilized were markedly affected. 

2. Indications, in many cases very striking, were obtained that 
there is a definite rhythm in the resistance of the eggs to shaking 
after fertilization. Immediately after fertilization, z ¢., thirty seconds 
to three minutes, the eggs were shaken to pieces even more readily 
than unfertilized eggs. Their resistance then enormously increased, 
and at ten minutes they were very resistant. At twenty or thirty 
minutes their susceptibility had again in part returned, but it never 
reached the susceptibility of unfertilized eggs. 

The increased resistance of the fertilized eggs was also observed by 
Meltzer! independently of us. He suggested that the unfertilized 
eggs were more brittle. One of us has already remarked on the ease 
with which matured star-fish eggs may be disorganized by shaking, 
while the immature are very resistant. The means of controlling 
the shaking accurately were wanting, so that we are unable to speak 
with certainty concerning the rhythm described above. We believe, 
however, that the explanation of the greater susceptibility of the un- 
fertilized eggs is due to the greater viscidity of its protoplasm, and 
that the alterations in susceptibility described above were probably 


1 MELTZER: Personal communication. 


304 A. P. Mathews and B. R. Whitcher. 


due to the rhythmic change in the viscosity of the protoplasm as de- 
_ scribed by Mrs. Andrews.! Her observations apparently coincide 
with ours, though it was impossible, owing to her failure to give the 
time data, to compare in detail our results with hers. It would be 
surprising if such a variation in susceptibility during the process of 
fertilization did not exist, since not only has Mrs. Andrews described 
histological changes in the texture and viscosity of the protoplasm at 
this time, but the experiments of Lyon? show that there is such a 
rhythmic variation in the susceptibility of the eggs to lack of oxygen 
and potassium cyanide, and Spaulding ® has recently shown the same 
rhythmic susceptibility to acids and the anesthetics. The method of 
mechanical shaking may thus throw light on variations in the physi- 
cal consistence of the egg protoplasm. 

3. Besides thus establishing variations in the resistance of the 
eggs to shaking before and after fertilization, we obtained well- 
marked results on the speed and course of development. Morgan # 
some years ago observed that star-fish eggs which were shaken be- 
fore fertilization developed more rapidly than those fertilized without 
shaking. We have confirmed this observation in Arbacia in some 
cases. Shaking these eggs ten to twenty times briskly, either be- 
fore, or just after fertilization, may produce either of two results, 
depending apparently on the variable factors already discussed in 
Part I. The eggs are either accelerated in their development and 
form plutei larger than normal, or highly abnormal gastrule are 
obtained which never form plutei. In some cases, of which the 
following experiment is an example, a small shock prevented the 
pluteus development of nearly all the eggs, while a brisk shaking, not 
severe enough to destroy them, caused them to develop better than 
the control. This was true whether the eggs were shaken before or 
after fertilization. 


Experiment XXIII. — September 1. Temperature of room, 24° C.  Fertilized 
rr A.M. All eggs from the same sea-urchin. ‘Temperature of water 
during transfer, 19°. 

1. Control. Not transferred. Eggs taken with great care from sea-urchin. 
September 3. Nearly all are normal plutei. 

2. ‘Transferred carefully defore fertilization. 
' ANDREWS: Journal of morphology (supplement), 1897, xii, p. 80. 

* Lyon: This journal, 1902, vii, p. 56. 

® SPAULDING: Personal communication; results not yet published. 

* MorGAN: Anatomischer Anzeiger, 1893, ix, p. 141. 


= 


Mechanical Shock in Protoplasmic Activity. 305 


September 3. Only eight perycent are plutei. Remainder are irregu- 
lar and evaginated gastrule. 
3. Squirted violently from one dish to another defore fertilization. 
September 3. Practically all are plutei larger than control. 
4. Shaken twenty-five times briskly defore fertilization. 
September 3. Some disintegrate, but remnant all form plutei still 
larger than 3. 
Eggs dropped one inch into a dish of water defore fertilization. 
September 3. No plutei are formed. Irregular gastrulz of peculiar, 
shape. 
6. Dropped six inches into water defore fertilization. 
September 3. A few irregular plutei. The rest are gastrulz. 
7. Dropped three feet into water defore fertilization. 
September 3. Result same as 6. 
8. Shaken briskly just after fertilization. 
September 3. Some eggs disintegrate. Remnant form large plutei. 
g. Transferred carefully one minute afer fertilization. 
September 3. No plutei formed. All are pointed gastrulz. 
to. Dropped one inch two minutes after fertilization. 
September 3. About one percent are small plutei. Rest are like 9. 


on 


The foregoing observations establishing the remarkable conse- 
quences of very small mechanical shocks on the development of 
Arbacia support, as already pointed out, the contention of Meltzer 
concerning the fundamental importance of agitation, or mechanical 
jarring, in the physiology of protoplasm in general. The many an- 
alogies already shown to exist between developmental and other physi- 
ological processes enable us to apply with some certainty to other 
cells the conclusions deduced from the egg cell. The importance 
of such shock, or of its absence, in considering pathological as well as 
physiological processes must not be lost sight of. Furthermore, al- 
though that shock may, as in the egg-cell, produce at the time no 
obvious histological change, the effect makes itself manifest in a 
derangement many cell-generations later. 

These observations emphasize the necessity of exercising extreme 
caution in dealing with the action of chemicals, of drugs, or physical 
agencies on protoplasm. The observations of last year on the in- 
fluence of pilocarpine and atropine on development required repetition, 
since in them the controls were not always transferred. A repeti- 
tion, however, has confirmed the results already obtained. It may 
be mentioned that the possibility that the glass of the finger-bowls 
used in the experiments introduced a source of error, was dismissed 


306 A. P. Mathews and B. R. Whitcher. 


by the use of paraffined glass and Jena beakers, which did not materi- 
ally change the result. All precautions were taken to guard against 
the possibility of being misled by polyspermy, controls having been 
carried out with so few sperm that only a portion of the eggs were 
fertilized. 

The results obtained, together with those on parthenogenesis by 
mechanical agitation, suggest, as pointed out to us by Meltzer, that 
the mechanical stimulation produced by the sperm, as it enters the 
egg, may be a matter of far greater importance in fertilization than 
hitherto considered. The point of the sperm enters the egg and is 
then violently moved from side to side by the lashing tail. The 
changes in the egg protoplasm which arise at this point may be the 
result of this motion and not of a chemical stimulation by the sperm 
head, as has been supposed to be the case. Perhaps the failure to 
extract a fertilizing substance from the sperm may be thus explained? 

Finally, the effects of mechanical agitation on fertilized and unfer- 
tilized eggs demonstrate, we believe, the fact that however various 
the stimuli may be in influencing protoplasmic activity, they produce 
their effect by altering the state of the protoplasm. This change in 
state, which may be a change in the state of aggregation of the col- 
loidal particles in the egg, may be set up in various ways, by tempera- 
ture changes, by the action of ions, by mechanical shock, by drugs 
or other organic substances, by the influence of oxygen or by its ab- 
sence. This change in state, however produced, is followed by defi- 
nite metabolic and structural changes. For artificial parthenogenesis 
nothing else is necessary than that this change be produced. I have 
no doubt that electric currents will suffice also, if properly applied, 
since in all other cells electricity affords us one of the easiest ways of 
producing such changes. In other words, we are in accord with 
Morgan’s criticism that it is impossible to assume that the stimulus 
given by the spermatozoon is of the nature of any of the methods so 
far found for producing parthenogenesis. It is not that the egg 
lacks specific ions or specific substances that it does not develop 
parthenogenetically, but only because in the conditions in which it 
normally finds itself, it cannot unaided bring about the necessary 
change in staté of its protoplasm sufficiently abruptly to cause its 
development. 

Our observations were made at the Marine Biological Laboratory 
at Wood’s Holl. 


1 GiEs: This journal, rgo1, vi, p- 53- 


FIVE TYPES OF EYE MOVEMENT IN THE HORIZONTAL 
MERIDIAN PLANE OF THE FIELD OF REGARD. 


By RAYMOND DODGE. 


[From the Laboratory of Psychology in Wesleyan University.]} 


OTWITHSTANDING the large amount of literature on the 
movements of the eyes, exact quantitative knowledge of the 
varieties of movement by which they respond to different external 
circumstances is comparatively limited. Some few types have indeed 
found general recognition; but they have been isolated either on ana- 
tomical grounds or on account of their relation to special problems. 
Such, for instance, are the torsion and convergence movements of the 
eyes. But the list of well-defined varieties is short, and the majority 
of the eye movements have no better classification than their direction. 
The well-known, persistent untrustworthiness of introspective data 
regarding our eye movements, together with the difficulty of harness- 
ing the eyes to physiological registering apparatus, readily accounts 
for this lack of quantitative differentiation. But the consequent con- 
fusion has seriously handicapped more than one study in physiological 
optics, and has been especially unfortunate where the burden of theo- 
retical importance that the eye movements have been compelled to 
bear has been greatest: I mean in the theory of the visual perception 
of space. 

The present analysis of the forms of eye movement in the horizon- 
tal meridian plane of the field of regard was occasioned by some 
perplexing optical phenomena, noted during the course of an experi- 
mental study of the visual perception of motion. The limitations of 
available apparatus have necessarily restricted the scope of the inves- 
tigation; but it seemed to me that enough of general physiological 
interest had been obtained to warrant its publication as a contribu- 
tion to the classification of the eye movements. 


1 In the experimental work involved in this paper, the author gratefully 
acknowledges the faithful help of his former pupil, Mr. J. J. Cogan. 
3°7 


308 Raymond Dodge. 


METHODS AND APPARATUS. 


Until recently, practically the only source of accurate quantitative 
data concerning the eye movements has been some use of the after- 
image. Some of these after-image methods have been surprisingly 
satisfactory. They have given us all that we know of the eye torsion, 
and of the movement and position of the point of regard during 
normal vision; and in the hands of Guillery ! and Briickner? modifica- 
tions of the well-known Lamansky method have given unexpectedly 
good results in measuring the angle velocity of the eye. But valuable 
as these methods are, they have narrow limitations, both in scope and 
in accuracy, of which no one has been more sensible than those who 
have used them. 

Among the objective methods of studying the eye movements, the 
oldest and most fundamental, though at the same time one of the 
most difficult, is direct observation, either with the naked eye, or by 
the aid of optical apparatus. Direct observation has discovered most 
of what we know concerning the involuntary eye movements, as well 
as the alternation of eye movement and fixation pause in the ordinary 
use of the eyes. But it can never yield any accurate measurements 
of their velocity, nor can it follow the longer excursions of the eyes 
with accuracy. 

The first really important contribution to our knowledge of the eye 
movements by means of direct attachment to the eye was made by 
E. B. Huey? at Clark University. With a Dellabarre eye-cup and a 
delicate registering device, direct kymographic records were obtained, 
showing the eye movements during reading. At the time of their 
publication these records gave the most nearly accurate measure- 
ments of the relation of reading pause to eye movement; moreover, 
they gave an unmistakably clear picture of that relation, which, 
doubtless, more than any other one presentation, has given the recent 
discoveries in the physiology of the eyes in reading a wide-spread 
acceptance. Unfortunately, the weight and friction of the register- 
ing apparatus, small as it was, seriously interfered with the normal 
functioning of the eye muscles. 

In the Psychological Review for March, 1901, in collaboration with 


' GuILLERY: Archiv fiir die gesammte Physiologie, 1896, Ixxiii, p. 87. 
* BRUCKNER: Archiv fiir die gesammte Physiologie, 1902, xc, p. 73. 
8 Huey: The American journal of psychology, 1900, xi, p. 283. 


a 


.} Five Types of Eye Movement. 309 


one of my students, Mr. T. S. Cline, I published the details of a 
method and apparatus designed to record movements of the eyes 
photographically. The record is a continuous photograph of a line 
across the eye in its horizontal meridian, taken on a slowly falling 
photographic plate of extreme sensitiveness. As the eye moves, the 


Via, 
=aeen 
ZL, 


FicurE 1 is a plan of the apparatus reproduced from the original paper by Dodge and 
Cline. The camera (C) rests ona heavy horizontal perimeter (2 4), which is fitted 
with a solid head-rest (4), and securely fastened to a heavy table. The arrangement 
for securing an evenly moving photographic plate is shown in Fig. 1 (2), with the 
top removed; and again, in side elevation, with the side removed, ona larger scale, 
in Fig. 2. 


inequalities of light along that line, corresponding to the transition 
from sclerotic to iris on either side of the pupil, change their relative 
position, marking a curve on the negative, corresponding in ampli- 
tude and form with the extent and velocity of the eye movement. 
Since, however, in most eyes the lines of demarcation between the 


310 Raymond Dodge. 


pigmented and unpigmented parts are not sufficiently well defined to 
produce negatives that could be read with accuracy, a modification of 
the above process was instituted, on which rests the method’s real — 
claim to accuracy. That modification is to photograph the move- 
ments of a sharply defined reflection from the eccentric surface of the 
cornea. This limits the scope of the method somewhat, in that the arc 
of eye movement can no longer be determined with accuracy from the 
records. But in practice the arc of move- 
ment is usually a predetermined experimental 
condition, and the real desideratum is real- 
ized: namely, the exact registration of the 
direction and duration of the eye movements, 
as well as their relation to the moments of 
rest. 

Only one modification of any importance 
has been made in the apparatus since its 
formal description. The regulation of the fall 
of the sensitive plate has been made more 
exact. Its fall is now controlled by the escape 


YWUUdéééllae-GY,y 


SS 


N 


MMMM, 


VM 


yy Z 
Y$)J«: 


MQ Hn MAMA 
W 


YUléla 
S 
YY 


il \ of oil instead of air. In the original apparatus, 
~ the plate-holder (4) was attached to the rod of a 


piston playing in acylindrical air compressor (@); 
the plate could fall only as the air escaped from 
the cylinder, and the velocity of the fall was 
regulated by the size of the opening through 
which the air escaped. Slight oscillations in 
the velocity of the fall, owing to the elasticity 
of the air, suggested the use of a fluid resist- 
ance. The loss of head which would naturally 
result as the liquid was forced out of the 
cylinder, was obviated by leading the vent pipe 
FIGURE 2. into the top of the cylinder, and thus returning 
the oil as fast as it was forced out, maintaining 
a constant volume in the cylinder, and rendering the whole device 
automatic and clean. 
This apparatus is probably the most satisfactory method yet devised 
for recording the movements of the eyes. It operates under normal 
conditions of binocular vision, is capable of registering both eyes 


MMMMMM@@Mlllllll 


simultaneously, has a unit of measurement less than Io, while a ray 
of light is an ideal registering medium, having neither momentum 


— 


five Types of Eye Movement. 311 


nor inertia. Its chief defects age that it is not easily adjusted to 
other than horizontal eye movements, and that it gives no indication 
of eye torsion. These limitations have forcibly determined the plane 
of experimentation, and have compelled me to abstract entirely from 
torsion movements. 


Type I. 


Those movements of the eye in which the point of regard wanders 
over any relatively fixed section of the field of vision are doubtless the 
most numerous, and at the same time the best understood of all the 
eye movements. In considering the various forms of eye movements 
in any direction, these would, therefore, naturally constitute the first 
type. 

The most important differentiating characteristic of the first type 
may be summed up as follows: The duration of eye movements of the 
first type is less than that of all other movements of the eye. It 
varies directly with the angle of displacement, but is approximately 
constant for each individual under the same conditions of fatigue of 
the eye muscles, of original orientation, and of the direction and angle 
of eye movement. 

If we were dependent on subjective data alone, every one would say 
without hesitation that he could move his eyes across the field of 
vision rapidly or slowly at will. This is, however, an illusion. The 
effort to move the eyes slowly from one point of regard to another 
always results in one or more complete stops, of which, however, the 
subject is almost never directly conscious. The simplest method of 
convincing one’s self of this fact is the method of Brown.! If the 
attempt be made to move the eyes slowly along a line which passes 
through a bright light, on closing the eyes a number of well-defined 
after-images of the light will be observed, clearly indicating that the 
eye rested at corresponding points along its path. More satisfactory 
is the evidence obtained by direct observation of another’s eyes. If 
one is careful not to look directly at the moving eye, but rather at 
some point on the eyelid, the alternation of movements and stops, as 
the subject attempts to move his eye slowly, will be clearly distin- 
guished. The kinetograms show that these pauses are of varying 
length, the shortest being slightly less than 0.2”. 


1 Brown: Nature, 1895, lii, p. 184. 


312 Raymond Dodge. 


It is unnecessary to repeat here a critical résumé of the earlier 
attempts to measure the duration of the eye movements. 

The following table, showing the results of measurements with the 
kinetograph, is reprinted from the paper by Dodge and Cline, on the 
angle velocity of the eye movements. 


General 
average. 


55 b= 
10° 5-5 
15° 10-5 


20°, 10-10 


30° 15-15 


40° 20-20 


The table gives the mean duration of the eye movements of three 
subjects through angles varying from 5°-40°. The column at the 
extreme left gives the angular lateral displacement of the line of 
regard, together with an indication of the orientation of the lateral 
displacement with relation to the primary line of regard. M. signifies 
the mean value in terms of thousandths of a second; M. V., the mean 
variation ; and No., the number of records from which the mean is 
reckoned. At the extreme right is given the general average for all 
three subjects. 

The results given in the table indicate that the duration of the 
movements of any individual eye through a given angle tends to re- 
main constant within the limits of a relatively small variation from 
the mean. The larger mean variation for the angular movements 
above 15° is due in part to the differences found to exist between the 
adductive and abductive movements of the eye. 

The table shows further that the duration of eye movement in- 
creases in direct ratio with the angle. Taking the general average of 
all three subjects as a basis for calculation, it would appear that for 
each 5° added to the amplitude of the eye movement between 5° and 
40, about 10a is added to the duration of movement. But the ap- 
parent implication of a fixed maximum velocity of 100 for each 5° is 


Five Types of Eye Movement. 313 


false. The experiments of Guillery and of Briickner, as well as my 
own experiments by the Lamansky method,! all show that the max- 
imum velocity of the eye during movements of large amplitude is 
greater than the maximum velocity during movements of small am- 
plitude. Unfortunately the general conditions of our experimentation 
preclude an exact analysis of the record curves. But some of their 
important characteristics are evident to the most casual observer. 


Reem ne cetpemmemesemen pecan ene en pesca m mans 
rj ry 
CORE er ese eee ae 


i 
! 
1 
1 
‘ 
' 
' 
' 
' 
. 
— 
. 
' 
. 
' 
' 
' 
' 
' 
‘ 
‘ 


10° 20° 30° 40° 

FIGURE 3 is an enlarged drawing from four kinetograms, subject 4A. Each section 
represents the record of one movement of the right eye, towards the right, through 

an angle of 10°, 20°, 30°, and_ 40°, respectively. Each vertical division of the paper 
corresponds to 125 (0.125”). The vertical height of the oblique lines gives the 

. duration of the eye movement, in each case; while the attached vertical lines repre- 
sent fragments of the fixation pauses separating the eye movements. The average 
duration of the fixation pauses partially represented in these drawings was about 

O52 


The record of every eye movement, of the first type between 5° and 
40° shows three distinct phases. Cf Fig. 3. The first phase con- 
sists of a positive acceleration tothe maximum velocity; this is main- 
tained for a considerable angle of movement, and constitutes the 
second phase, giving place in turn to a negative acceleration phase 
as the eye comes torest. The relation of these phases is not constant. 
In the shortest excursions measured, the second phase is very short, 
while in the longest excursions, with the exception of a peculiar modi- 
fication in the abductive movements to be described later, the second 
phase is by far the most conspicuous. Moreover, if one superimpose 


1 ERDMANN and DopcE: Psychologische Untersuchungen iiber das Lesen, 
Halle, 1898, Anhang, pp. 346-360. 


314 Raymond Dodge. 


a curve for a movement of 15° on a curve fora movement of 4o’, the 
second phase of the latter record will be found to incline slightly more 
to the horizontal. This confirms the law that the maximum as well 
as the average velocity increases in direct ratio with the angle of 
movement; while, in opposition to the findings of Bruckner, the 
kinetograms indicate that the increase in velocity is confined to the 
second phase of movement, which in Brickner’s experiments could 
not be dissociated from the first phase. 

Guillery observed a decided difference between the velocity of the 
eye at the beginning and at the end of an eye movement; but his 
experimental method involved two conditions that tend to distort that 
relation. In the first place his eye movements were uniformly ex- 
treme, and involved considerably more effort and muscle strain than 
the more natural excursions measured by Dodge and Cline, which 
never exceeded 20° from the primary position of the eye. Even more 
important is the fact that I have found it absolutely impossible, even 
under the most favorable conditions, to secure a series of simple, 
direct movements of the eyes from one fixation point to another 
which was more than 40° distant. The distance is persistently under- 
estimated; and the initial long movement of the first type is suc- 
ceeded by a shorter corrective movement of the same type. Since 
Guillery’s eye movements were all 40° and over, it seems probable 
that his attempt to measure the velocity of the end of the eye move- 
ments was confused by the small corrective movements whose 
average velocity is comparatively low. In the abductive movements 
alone, the kinetograms show a marked difference between the veloc- 
ity of corresponding portions of the first and the third phases. 
This peculiarity of the third phase is sufficient to account for the 
longer duration of the abductive movements, as remarked indepen- 
dently both by Guillery, and by Dodge and Cline.! 

The question naturally arises whether the eye movements which 
occurred under the more or less artificial conditions of our experi- 
mentation are of the same kind as those which occur during the 
unconstrained changes in the line of regard in normal vision. The 
problem of producing thoroughly unconstrained, horizontal eye move- 
ments, under the rather exacting conditions of kinetographic regis- 


1 Briickner finds the relation reversed in his own case. The adductive move- 
ments are longer than the abductive. And this leads him to conclude that the 
differences are mere personal peculiarities, rather than permanent differences of 


the eye movements in the two directions. 


Five Types of Eye Movement. 315 


tration, might seem at first rathermdifficult. But such movements are 
found in reading. As is now well known, since the researches of 
Erdmann and Dodge,! and of Huey, the reading of a line of print 
involves a succession of alternating fixation pauses and eye move- 
ments. These vary in number for any one person with the length of 
line, the print, and the difficulty of the matter read. The complete 
lack of introspective data regarding this alternation is sufficient 
sponsor for the lack of constraint. Yet measurements of the eye 
movements during reading are altogether congruent with the meas- 
urements given in Table I.’ 

The second general characteristic of eye movements of the first 
type relates to their function in the process of visual perception. 
Those experiments in physiological optics which depend on the varia- 
tion in appearance of luminous points and lines during eye movement 
demonstrate that the normal functions of the retina are not sus- 
pended. But the uniformity of the eye movements of Type I, and 
their relatively short duration as compared with the fixation pauses, 
naturally raises the question how far normal vision depends upon 
them. The layman will confidently assert his belief that he can see 
while his eyes are moving from place to place. Even among physi- 
ologists and psychologists it seems to have been taken for granted 
that we could see during eye movement from one point of regard to 
another. But wherever found, the belief that we can see the succes- 
sion of points of regard clearly during eye movements of the first type, 
is an illusion of introspection. The changes which a luminous point 
undergoes during eye movement, as shown, for instance, in the 
Lamansky experiments, are significant. But the matter may be 
proved by direct experiment. An object placed midway between two 
arbitrary fixation points, say 30° apart, cannot be seen more clearly 
during an unbroken eye movement from one fixation point to the 
other, than it is seen from either fixation point, although the line of 
regard moves directly through it. Moreover, it can be demonstrated 
that, under ordinary conditions of illumination by diffused daylight, 
the object cannot be seen at all, if it is exposed only during eye move- 
ment. Ifacone of paper with a base sufficiently large to cover the 
object of regard, and an aperture at the apex the size of the pupil, be 


1 ERDMANN and DopGE: Psychologische Untersuchungen iiber das Lesen. 
Halle a. d. Saale, 1898. 

2 Huey: American journal of psychology, 1900, xi, pp. 283-302. 

8 DopGE and CLINE: Physiological review, 1901, viii, p. 153. 


316 Raymond Dodge. 


placed between the object of regard and the eye, the object will be 
visible only as long as the pupil stands directly over the aperture in 
the apex of the cone. If the eye move, as above, through an angle 
large enough, so that the object will be completely hidden by the 
sides of the cone at both termini of the movement, it will be found 
that, unless interruptions occur in the movement, absolutely nothing 
is seen of the object. If artificial illumination be used in a darkened 
room, it will be found that the conditions of fusion with a moving 
retina are approximately the same as the conditions of fusion when 
the object moves.! 

The problem why, under ordinary circumstances, we do not per- 
ceive the fusion of the field of view which should theoretically occur 
during each eye movement of the first type, has been solved only 
hypothetically. But the fact that even with careful attention, scarcely 
a trace of the theoretical fusion of the field of view during eye move- 
ments of lesser excursion is perceptible, combined with the fact that 
during longer movements, say from an extreme inner to an extreme 
outer position of the eye, some trace of the fusion may be noted, gives 
a clue to the advantages of rapid movement, uncontrolled by voluntary 
effort. It would be a serious disturbance to clear vision if the whole 
field of view melted into an even gray at every eye movement, four or 
five times a second. And it would be an even more serious disturb- 
ance if, with still further voluntary reduction of the velocity, the whole 
field of view should appear to move about, as the characteristics of 
the fifth type would indicate.” 

The general characteristics of the first type may be summed up 
as follows: 

1. Eye movements of the first type are fundamentally reactions to 
eccentric retinal stimulation, and are dependent on the tendency, de- 
veloped during the first month of infancy, to move the eyes so that 
the point of interest will be seen with the visual centre of the retina. 

2. Their velocity is practically uninfluenced by voluntary effort. 
While their duration shows a slight individual variation under sim- 
ilar circumstances, it varies in direct proportion with the angle of 
movement, 

3. They are primarily not periods of perception, but rather inter- 
ruptions of vision, whose sole function is to move the line of regard 
to an eccentric point of interest. 

' For a more detailed account of these phenomena, see DonGe, Visual percep- 
tion during eye movement. Psychological review, 1900, vii, pp- 454-465. 

* See page 328. 


Five Types of Eye Movement. 317 


Three subtypes appear to conform to the general characteristics of 
the first type, without being true reactions of the eye to eccentric 
visual stimuli. These are : 

1. Those voluntary and arbitrary movements of the eyes which 
may occur with the eyelids open or closed, without real reference to 
any object of interest in the field of vision. 

2. The involuntary movements which persist even after all ex- 
ternal stimuli have been excluded from the eye. 

3. Co-ordinate movements of one eye as the other eye reacts to an 
eccentric stimulus, unseen by the former. 


Type II. 


Eye movements of the first type, by which we look towards an 
eccentric object of interest, have been shown to be conditioned by 
the previous position of the eye, and are involved in each new act of 
vision only indirectly as a necessary precondition. If we wish to see 
a moving object, on the contrary, a more or less continuous move- 
ment of the eyes will be necessary in order to keep the line of regard 
congruent with the line of interest. The eye movements in this case 
are no longer mere episodes, dependent on the previous position of 
the eyes; they are the moments of fixation, the immediate condition 
of clear vision, and the direct analogues of the fixation pauses which 
separate movements of the first type. This modification of the 
function of the eye movements, as might be expected theoretically, 
corresponds with a modification of the character of the eye move- 
ments, constituting a new type. 

We may define this second type as those eye movements in which 
the line of regard follows an object moving across the field of vision. 
And the eye movements of the second type may be designated pz- 
suit movements. They occur in early infancy, and are so persistent 
in adult life that even the trained psychologist finds it difficult to 
keep the eyes fixed in their orbits when the object of regard moves. 

An analysis of the conditions under which a pursuit movement 
may occur shows that every group of pursuit movements must begin 
as a reaction to eccentric stimulation. Until the object of interest 
moves, there is no occasion for eye movements ; after it moves, how- 
ever, a measurable time interval must elapse before there can be a 
muscular response to the perception of motion. The available data 
concerning the reaction time of the eye indicate that this interval 
is relatively long. 


318 Raymond Dodge. 


The first published measurements of the reaction time of the eye 
were made by the writer in collaboration with Professor Benno 
Erdmann in 1897.1. The method used was the blind spot method, but 
the available apparatus was crude, and the results were published as 
mere approximations. These experiments indicated that the eye 
reaction to eccentric visual stimuli involved a reaction interval less 
than 230o¢ and more than 1800 for both subjects. In 1899 it was 
possible for me to measure the reaction of the eye by the same 
method with apparatus especially designed for the purpose. Both 
method and apparatus are fully described elsewhere,” and it will be 
necessary at this time merely to outline their general principle. A 
primary fixation mark was suddenly superseded by an eccentric mark, 
towards which the subject was to look immediately on its appearance. 
The duration of the reaction interval was given directly in the length 
of time which a bright light must persist, after the stimulus for re- 
action had been given, in order that it might emerge from the blind 
spot, on which it rested during the primary fixation, and thus become 
visible through the reactive movement of the eye itself. The meas- 
urements were made on the right eyes of two subjects, and after 
correction for the slight movement of the eyes, indicated reaction 
intervals respectively of 170¢0 and 162¢. In 1900 Huey® tested the 
earlier approximations of Erdmann and Dodge by his method of direct 
attachment, and obtained a corrected mean reaction time for two 
persons of 171.7¢0 and 196.ga, respectively. These results by totally 
different methods are fairly congruent, and show us clearly that if the 
first type of movement were the only type, it would never be possible 
for us to see a moving object clearly. The attempt to follow the 
moving object would keep the point of regard at least one full reaction 
interval behind the point of interest; moreover, since the eye move- 
ments of the first type effectually prevent clear vision during the 
movement, the best view we could possibly get of a moving object 
would be given by the confused, and more or less eccentric image of 
the moving object while the eye was motionless. That we are not 
confined to such unsatisfactory conditions of observation is evident 
from the most obvious data of introspection. 

Direct observation of an eye, following a uniformly moving object, 


' ERDMANN and DopGeE: Psychologische Untersuchungen iiber das Lesen, 
1898, pp. 116 and following. 
* DonGE: Psychological review, 1899, vi, pp. 477-483. 


’ Hury: American journal of psychology, 1900, xi, pp. 294, 295. 


five Types of Eye Movement. 319 


discloses a relatively complex phenomenon, which apparently includes 
at least two distinct kinds of eye movements. A succession of rapid, 
jerk-like movements are separated by what appear to be longer regular 
movements of less velocity. The exact relation of these components, 
however, cannot be made out by mere observation. In order to 
adapt the kinetograph to this new problem, a moving object must 
take the place of the fixation marks. For this purpose we used a 
continuous belt of white paper about one inch wide, marked with a 
succession of black crosses. This ran around a polished vertical rod 
at one side of the table, and around a vertical spool driven by clock 
work at the other. Screens were then adjusted on the perimeter, to 
shut off all but twenty degrees of the belt, and the clock work was 
regulated so that a point on the belt traversed the entire twenty 
degrees in about one second. 

Kinetograms of the eye movements made while the belt was in 
motion, in the endeavor to follow the cross fixated as long as it was 
visible, and then to transfer the fixation immediately to another just 
appearing, show the following details: 

I. The attempt to fixate a cross after the signal is given, results, 
first, in two or three rapid movements of the eye, separated by 
moments of rest. This first phase of pursuit differs from the kineto- 
grams of reaction movements of the first type, only in the shortening 
of the fixation pauses. 

2. After the first two or three rapid movements there are no 
further moments of rest during the entire course of a group of meas- 
urements, about eighteen seconds. Instead of the moments of rest 
appear moments of relatively slow eye movement, approximating in 
angle velocity the movement of the crosses on the belt. 

3. Several rapid movements in the direction of pursuit are always 
found to interrupt each pursuit sweep. These vary from the rapid 
movements in the opposite direction, which separate the pursuit 
sweeps, not only in number and amplitude, but also in their apparent 
function. The latter are usually simple movements of the first type 
through the entire arc of movement, and indicate the most rapid 
transition from the end of one pursuit sweep to the beginning of a 
new one, resembling in every way the return sweep of the eye in 
reading. The rapid movements in the direction of the pursuit, on 
the other hand, interrupt each pursuit sweep of twenty degrees from 
three to six times, three-fourths of such interruptions occurring in the 
first half of each sweep. 


‘ 
| eatiesteetiesteetientienti alent ee 


1 
i 
! 
" 
| 
! 
‘ 


1 
t--------|------- 
H 


See eer ee 


20° 


FIGURE 4 is an en- 

larged reproduction 
of a kinetogram of 
one entire pursuit 
sweep of the eye 
from left to right, 
bounded by the re- 
turn sweeps from 
right to left, and 
broken by three 
short auxiliary 
jerks of the first 
type. 


Raymond Dodge. 


Since the rapid movements in either direction 
differ in no appreciable way from the movements 
of the first type, it is evident that we must dis- 
tinguish between the slower true pursuit move- 
ments of the second type, and the pursuit sweeps 
which include both the true pursuit movements 
and the frequent additional movements of the first 
type. In the attempt to free the true pursuit 
movements from the little auxiliary jerks, and thus 
study them in a pure form, we tried to reduce the 
velocity of the belt until it should exactly coincide 
with the true pursuit movements. But the plates 
showed that we had misapprehended the meaning 
of the auxiliary jerks. For even in slow move- 
ments of the object of regard, in which the twenty 
degrees was covered in about three seconds, the 
little jerks still persisted, though they were of 
extremely small amplitude. Since the velocity of 
the true pursuit movements constantly decreased 
with the velocity of the object of regard, it seems 
probable that we must regard the auxiliary jerks 
of the first type as constant accompaniments of 
the pursuit movements; and since they always 
appear in the direction of the pursuit, they indi- 
cate that the true pursuit movement tends to lag 
a little, and is supplemented from time to time by 
movements of the first type. 

The most important feature of the records, from 
the standpoint of classification, is the fact that, 
after the first few moments of rest, which separate 
the first two or three auxiliary jerks at the begin- 
ning of each series of measurements, the pursuit 
movements assume the character of a continuous 
series of easily modified habitual movements. The 
second pursuit movement does not begin after the 
elapse of a reaction interval, but immediately on 
the return of the eye to the beginning of the arc 
of movement, the new pursuit sweep begins as 
though it were the release of a spring. Our inter- 
pretation of the pursuit movements as a continuous 


five Types of “Lye Movement. Se1 


series of rapidly developed habitual responses to a general condition, 
rather than as individual reactions to specific stimuli, is corroborated 
by the peculiar action of the eyes of a person who, having watched 
a moving object for some time, suddenly tries to fixate an unmoved 
object. In such cases, notwithstanding the absence of any true 
stimulus for the pursuit movements, and even in opposition to the 
most strenuous effort to fixate the object at rest, the pursuit move- 
ments persist for an appreciable length of time, and give rise to a 
well-known illusion of motion. The objects which are really at rest 
seem to move in a direction opposite to that in which the real move- 
ment occurred. In experiments with a pendulum in place of the 
endless belt, the records show that the habitual response may involve 
considerable variation in velocity in its different phases. 

In the traditional accounts of the visual perception of motion, the 
perception of the eye movements by which the object is followed is 
always mentioned as an important factor. Without discussing the 
general question in this place, it may be remarked that since the 
pursuit movements invariably lag, they alone would give very erro- 
neous data concerning the velocity of the object. Moreover, since 
they are initiated as reactions, it would seem that the movement of 
the object must already be apprehended before the pursuit can begin, 
and that its velocity must have been estimated with considerable 
accuracy before the pursuit movements could fulfil their function 
sufficiently well to give any reliable data. 

The chief characteristies of the second type of eye movements may 
be summed up as follows: 

1. The velocity of the pursuit movements has no normal value, but 
varies with the apparent velocity of the object as it moves across the 
field of vision. The line of regard appears to lag behind the line of 
interest, and to overtake it, from time to time, by short eye move- 
ments of the first type. 

2. Unlike movements of the first type, the pursuit movements are 
moments of clear vision. In fact it is only during pursuit movements 
that clear vision of a moving object is possible. 

3. In further distinction from movements of the first type, which 
are fundamentally reactions to specific eccentric stimuli, the pursuit 
movements assume the character of habitual movements, and may 
persist after the occasion for them has ceased. 

4. Finally it is to be noted that, whereas movements of the first 
type are separated by relatively long periods of rest, movements of the 


222 Raymond Dodge. 


second type show no periods of rest whatsoever, and must, therefore, 
involve a more continuous activity of the eye muscles. 

A form of eye movement, not objectively different from the pursuit 
movements, occurs during passive movements of the entire body, as 
when one rides in a car or is revolved steadily in a chair, while the 
environment remains stationary. This form, however, is complicated 
by tendencies to movement which belong to the fourth type.! 


Type III. 


Similar in some respects to the pursuit movements, yet essentially 
a new type, are those which we may call after Lotze, compensatory eye 
movements. In order, however, to differentiate them from allied 
movements of a fourth type, we must designate them co-ordinate com- 
pensatory eye movements. This third type may be defined as those 
movements of the eyes by which the constant fixation of an unmoved 
object of interest is maintained during rotation of the head. 

Obviously in every movement of the eyeball which corresponds to 
the above conditions, there are two distinct factors. The first factor 
depends on the optical eccentricity of the axis of the head. Since 
the vertical axis of the head lies behind the eyes, they will be carried 
with their orbits in the direction of the head movement. Besides the 
passive movement around the axis of the head, every compensatory 
movement involves a movement of the eyeball in its orbit. The 
entire process has sometimes been regarded by those with whom I 
have spoken, as a mere matter of momentum, by which the eye 
remained relatively fixed, while the head revolved around it. This 
conception is, of course, as thoughtless as it is misleading. It would 
be possible only if the axis of rotation of the eye were identical with 
that of the head, while the eye hung free in a frictionless orbit. Since 
neither of these conditions is fulfilled, we must conclude, on theo- 
retical grounds, even if we had no direct experimental proof, that 
the compensatory eye movements have their direct antecedents in 
the more or less continuous activity of the orbital muscles. 

The quantitative investigation of the third type presents some pecu- 
liar difficulties. Since both the head and the eyes move simultane- 
ously, it is obviously impossible to use any of the ordinary head-rests. 
Without them, however, the probability of securing satisfactory nega- 


1 See page 327. 


Five Types of Eye Movement. 225 


tives is very small, since a chamge of one-sixteenth of an inch in 
the elevation of the head is sufficient to throw the reflection from 
the cornea out of the exposure slit. In order to offset their loss, 
a sort of peep sight was used to give the head the right elevation. 
About half-way between the eye and the constant fixation mark, a fine 
silk strand was stretched across a concave support, at such a height 
that its ‘upper edge appeared to touch the fixation mark, when the 
head was at the required elevation. The chief theoretical objections 
to this arrangement are two: Interest is divided during the experi- 
ment, between the true fixation mark and the thread; and this might 
easily occasion involuntary changes in the line of regard, which would be 
overlooked by the subject. The second objection involves the almost 
inevitable involuntary changes in the elevation of the head. Such 
changes, after they once occur, must disturb the free movements of the 
head until they are corrected. On the other hand, this means of pre- 
serving the elevation of the head is preferable to any mechanical de- 
vice attached to the top of the head, since the latter would inevitably 
modify both the direction and the velocity of the normal fulfilment of 
a volitional impulse; and this would destroy an essential condition 
of the experiment. The resulting kinetograms show that the head of 
the subject never kept within the limits of permissible variation from 
a constant elevation for more than four or five movements. But even 
the faulty records are not worthless, since, once understood, their 
obscurer lines are sufficiently distinct to control the perfect records. 

In order sharply to differentiate head lines from eye lines on the 
kinetograms, a black cardboard square with a heavy white line was 
fastened to the nose. The eye record was made by the movement of 
the corneal reflection, as in the previous types. 

The chief characteristic of the records is the fact that not once is 
there a break in the regular compensation of the eye to movements 
of the head. Nowhere is there discoverable even a hint of an eye 
movement of the first type, either at the beginning of a series of head 
movements, or during their progress. This justifies our emphasis on 
the co-ordinate character of the eye movements of the third type. 
That they do not begin as reflex movements, like the second type, is 
shown by the entire absence of the reaction interval, as well as by 
the absence of secondary corrective movements of the first type. 
These two characteristics also show that the form of compensatory 
eye movements under discussion cannot be a response to the afferent 
impulses from the semi-circular canals, as was proposed by Mach and 


324 Raymond Dodge. 


Brown. There are, indeed, such responses, but they belong to a 
_ different class of compensatory eye movements. 

Since there is no control for a primary position of the head, and 
the angle which the axis of the eye and the nose mark subtend to 
the centre of the camera lens, varies constantly during the rotation 
of the head, the important question whether the compensatory eye 
movements follow variations in the velocity of the head movement 
accurately or only approximately, is not answered by the kine- 
tograms. Recourse was consequently had to an experiment of the 
Lamansky type which serves at once as a demonstration of the most 
important characteristics of the third type, and an effectual control of 
the kinetographic records. Ina darkened room, a black pasteboard 
screen, perforated near the centre by a narrow slit made by the thin 
point of a penknife, was interposed between the eye and a bright 
gas flame. Directly in front of the screen, a revolving disk was 
adjusted so that perforations on its periphery exposed the open slit 
about a hundred times a second. As long as the slit was fixated by 
the observer, it appeared as a narrow line of light. Whenever the line 
of regard wandered, the slit appeared to broaden, and if the angle of 
eye movement exceeded half a degree, it appeared double. If the slit 
retained its normal shape while it was fixated during head movement, 
it was conclusive evidence that the eye movement completely com- 
pensated for the head movement, not only at the end of the head 
movement, but also at every moment of that movement. Our experi- 
ments showed that compensation is accurate only for head movements 
of moderate velocity and excursion. When the head moves with 
extreme velocity, the eye lags. Since it is a well-known fact that 
the head can move more rapidly than the eye, this insufficiency of 
the compensating movements under extreme conditions was to be 
expected, and in no way discounts their general accuracy during mod- 
erate head movements. <A quantitative determination of the limits of 
exact ocular compensation to head movements was prevented by the 
fact that when the head moves back and forth as rapidly as possible, 
considerable confusion results in the field of view, leading to dizzi- 
ness and even to nausea. In relation to our specific problems in the 
visual perception of motion, it is to be noted that as long as the 
compensatory movements were accurate, the point of light appeared 
stationary ; on the other hand, when the head was moved as rapidly as 
possible, z. ¢., under conditions which have been shown to preclude 
accurate compensation, the point of light appeared to move. 


five Types of Eye Movement. 325 


Apart from its relation to the problems of the visual perception of 
motion, which need not be exploited in this place, this third type 
seems to me to be of some general physiological interest. In the first 
place we are compelled to postulate a very complex anatomical basis of 
co-ordination. This must not only suffice to explain the compensating 
movements of the eyes when the head rotates, but similar phenomena 
which I have found to accompany changes in the elevation of the 
head in walking and running. An instructive instance of a lack of 
co-ordinate ocular compensation is found if the head remains fixed 
with relation to the shoulders, while the trunk rotates back and forth 
on a vertical axis at the hips. The attempt to fixate a given object 
under these conditions leads to apparent movements of the field of 
view, and even to dizziness, contrasting sharply with the experience 
during mere head movements. Furthermore, this delicate preadjus- 
ment of the eye movements with head and leg movements, not only 
with respect to their extent, but also with respect to their velocity, 
seems to me to represent an important and hitherto unexploited 
link between tactual space and visual space. 

This side of the problem naturally leads to the further question 
whether the compensatory eye movements are wholly dependent on 
central co-ordination, or in part also on some control from periph- 
eral stimulation. In all subjects tested, I have found compensatory 
eye movements to occur whenever the head was rotated voluntarily, 
even if the eyelids were held tightly closed. These eye movements 
may be shown to exist in spite of all efforts to prevent them, by direct 
observation, or by allowing the finger-tips to rest lightly on the closed 
eyelids as one rotates one’s head from side to side. The accuracy of 
such compensation, however, can be determined only by a special ex- 
perimental study, for which we have at present no available recording 
apparatus. 

The close relation between the rotation of the head and the com- 
pensatory eye movements suggests the question how the head and 
eyes function together, when both unite in bringing an object of 
interest into the centre of the field of regard. My own observation 
may be summed up briefly as follows: The eye never remains fixed 
in its orbit when the head rotates, unless the point of interest 
happens to have the same angle velocity as the head, as, for 
instance, when we fixate the end of the nose. As a rule, when the 
subject acts naturally and without effort, the eye reacts first to an 
eccentric point of interest with a movement of the first type, and 


326 Raymond Dodge. 


the consequent slower movement of the head, to equalize the strain 
_on the eye muscles by keeping the point of regard approximately at 
the centre of the field of regard, is accompanied by the regular com- 
pensatory movements of the eyes. Under the influence of introspec- 
tion this order may be inverted, but I have never found any one who 
could hold his eyes fixed in their orbits, during head movements, 
unless, as mentioned above, the end of the nose or some point moving 
with the same angle velocity was fixated. Through the kindness of 
Dr. A. R. Defendorf, Pathologist of the Connecticut State Hospital 
for the Insane, I have been able to examine the eyes of patients suffer- 
ing from various forms of insanity, with respect to this peculiarity ; 
and it has appeared to me that some cases of the katatonic form of 
dementia przcox, as well as cases of dementia paralytica, presented 
exceptions to the normal compensatory eye movements, which might, 
perhaps, become of diagnostic value. These abnormalities, however, 
are to be understood as pathological conditions, and not as exceptions 
to the general rule. The law of the normal interaction of head and 
eye is doubtless to be explained in part by the overpowering tendency 
to compensatory eye movements, and in part by the fact that, for rea- 
sons previously explained, the reactive eye movements of the first 
type are relatively constant in velocity, while the angle velocity of 
the head movements may be varied voluntarily within wide limits. 
Casual observation of the relation between the head and eye move- 
ments of the lower mammalia shows a decided deviation from the 
rule observed in man. 

The general character of the third type of eye movements naturally 
suggests the cerebellum as the anatomical centre, and it seems not 
improbable that we have come upon a physiological explanation of 
those sagittal fibres of the middle peduncle which appear to connect 
the cerebellum directly with the nuclei of the third, fourth, and sixth 
cranial nerves. 

The essential characteristics of the third type of eye movements 
may be summed up as follows: The co-ordinate compensatory move- 
ments of the eyes are not reactions to external stimuli. They adjust 
themselves to all changes in the angle velocity of the head move- 
ments, except movements of extreme velocity, immediately, without 
the elapse of a reaction interval, and without measurable error. 


Five Types of Eye Movement. 


Oo 
to 
= | 


Type IV. 


A type. of compensatory eye movements hitherto undifferentiated 
from the third type, but really distinct from it both in its origin and 
in its development, consists of those eye movements which occur in 
reaction to the organic sensations of passive movement. In contra- 
distinction from the third type, these may be called reactive compensa- 
tory movements. 

If the subject be rotated in a revolving chair, first in one direction, 
and then in another, while his finger-tips rest lightly on his closed 
eyelids, he will feel a number of jerky compensatory movements of 
the eye. These jerks are not reactions to the apparent movement of 
the environment like Subtype I of Type II, since the eyes are closed: 
neither can they be due to a co-ordinate efferent impulse like Type 
III, since the movement is passive. Unfortunately, the essential 
condition of the phenomena in pure form is that the eyes be closed. 
This, however, precludes a quantitative investigation by any of the 
methods at hand. Even their function is obscure, though they 
probably co-operate with movements of the second type during the 
rotation of the body on its vertical axis. 

According to the observations of Brown and Mach, they cease 
during a long continued rotation in one direction, and begin again 
only when the rotation ceases. If this is found to be a general 
characteristic, their persistence after rotation ceases must be due to 
entirely different causes from those conditioning the habitual per- 
sistence of the simple pursuit movements after the objective occasion 
ceases. 


Type V. 


The fifth type is in several respects unique. It consists, primarily, 
of movements of convergence and divergence. Like the first type, it 
is originally a reaction to eccentric stimulation, but the stimulus falls 
on disparate points of the two retinz, and the movements of the 
two eyes are consequently not in the same but in opposite directions. 
The most conspicuous of the differentia: of this type are the relative 
slowness of the movements, and the fact that, notwithstanding their 
reactive character, they permit a more or less clear perception of the 
field of vision during eye movement. 


328 Raymond Dodge. 


Whereas movements of the first type through 10° occupy about 40a; 
_ eye movements of the fifth type, of which we have records, occupy 
about 4000; and I have noted movements subjectively that occupied 
a full second. The physiological grounds of this lessened velocity we 
have some data for explaining. If two needles are placed, respec- 
tively, one and two feet distant from the head, in such positions that 
when the right eye is at its primary position, both needles lie in its 
line of regard, it will be found that, in looking from one to the other, 
the movement of the left eye is of the fifth type, while the right eye, 
which should theoretically remain stationary, makes jerky little excur- 
sions out of the primary position, which it, in turn, afterwards corrects 
by movements of the fifth type. Kinetographic records of the little 
abductive jerks show that they correspond in velocity to movements 
of the first type. We may conjecture that they are occasioned by 
the same tendency to binocular co-ordination of the eye movements, 
that normally causes one eye to move in the same direction and ap- 
proximately through the same angle with the other, when the latter 
responds to an eccentric stimulus, even though the former eye be 
closed, or the operative stimulus completely screened from its field of 
view. It seems not improbable that in all convergence movements 
of the eyes, each eye is retarded more or less by the tendency to 
move sympathetically in the same direction with the other eye. 

The second characteristic of the fifth type may be observed under 
almost any condition of convergence as soon as one is attentive to it. 
It is particularly noticeable when fixation marks, which are approxi- 
mately in line with the nose, and clearly outlined against a plain white 
background, are alternately regarded. In each change of the line of 
regard, the double images of the fixation mark, towards which the 
lines of regard are moving, may be seen to approach each other, while 
the single image of the mark previously fixated resolves itself into two, 
which gradually separate to a maximum distance as the former two 
coalesce. These phenomena are even more pronounced if the con- 
vergence of the eyes is increased or relaxed artificially. The ap- 
parent movement of two stereoscopic diagrams which are united 
without apparatus by voluntary divergence is a familiar example. 
Similar movements may be noticed when one eye, which has been 
closed for some time, while the other eye has been functioning nor- 
mally, corrects its faulty fixation. 

Physiologically, this dual perception of the fields of vision moving 
over each other is interesting as an even more decided case of dis- 


five Types of Eye Movement. 329 


parate images from identical retinal points than the well-known 
Wheatstone phenomenon. Psychologically, it appears as a constant 
though relatively unexploited datum in the differentiation of the 
crossed and uncrossed retinal images in the perception of the third 


dimension. 


ON THE QUANTITATIVE DETERMINATION OF 
AMMONIA IN URINE. 


BY PHILIP SHARPER: 
[From the Chemical Laboratory of the McLean Hospital for the Insane, Waverley, Mass.| 


CONTENTS. 

Page 
I. Introduction. . . aS Rola ee MEE eee. ee 1 4 580) 
II. The Schlésing method seeks, Pie os é nel a) noe es Seeger 
Conditions governing the SEnGone Satiod oi Siw 15) sac a Oe ea 
The Schlésing method as described in textbooks . . . ..... . 333 
The Schlésing method as used by previous investigators . . . . . . 335 
The alkali used to liberate the ammonia . . . . 6:5 «Deco 
The effect of temperature upon the Schloésing wictiod oe os 2 eS 
The Schlésing method as best carried out... “» . . . . = | |) == 
Tables of results by the Schlosing; method\e fe) esa) eset teers 
IIT. Vacuum distillation method according to Wurster . . . 2 ke 2s tee 
The modifications by Nencki and Zaleski, and by Séldner i: 2 (> SSS 
hv. ‘DheFolinsmethods! = -)-).) < F eee ES 
V. Vacuum distillation method according to *Roussinaantt i be. a ts 
VI. The vacuum distillation method as best carried out. . . . : 348 

Tables of results by the new Folin method and by the peciaed vacuum 
distillation methods: 2%. rads) 2 a 
VIL. ‘Summaryinc 665 joe Oe Ne LS GI eae aie ec mer 


I. INTRODUCTION. 


HEN one compares the amount of nitrogen excreted from the 

body as ammonia, with that excreted as uric acid and the 

other nitrogenous substances existing in urine in less quantities than 
urea, and when one considers the large part played by ammonia in 
metabolism, and the consequent importance of the variations which 
occur in the amount of this substance excreted under different physi- 
ologic and pathologic conditions, it seems strange that the methods 
proposed for its estimation have not been worked out toa greater 
degree of efficiency and accuracy. The determination of ammonia 
has assumed an increased importance in connection with a work on 


metabolism now being pursued in this laboratory, and it has seemed 
330 


2 ee 


Quantitative Determination of Ammonia in Urine. 331 


advisable to examine critically the better adapted and more generally 
used methods to determine their accuracy and reliability. 

The better known methods are: The Schlésing method for the 
absorption of the ammonia by acid on standing in a closed vessel, 
and the several methods by distillation in a vacuum. For one of the 
latter I shall propose several modifications which make it highly 
accurate and very satisfactory. The methods based upon the precip- 
itation of the ammonia as a platinum double salt involve, according to 
Huppert,! the error of having an admixture of organic-platinum pre- 
cipitates, and furthermore are too tedious and expensive for constant 
use. Such processes were rejected for these reasons. 

Two methods by Otto Folin, one of which has not yet been pub- 
lished, have been included in the investigation. 


II. Tue ScuLiosinc METHOD. 


This method was first described in a paper by Schlosing? and was 
used to determine the ammonia in tobacco. The first application to 
urine seems to have been made by Neubauer in the first edition of 
his “ Anleitung fiir Harn Analyse.” Neubauer had not at that time 
made any critical test of the reliability of the method, but in 1855 he 
published a paper ® stating the details of the process and giving the 
results of a number of ammonia determinations in normal urines. 
He stated that the method gave very accurate results. 

As used by Neubauer, the Schlésing method was as follows: To 
10 c.c. of urine * was added magnesia or milk of lime, and this mixture 
was placed in a flat dish ten or twelve centimetres wide; over this dish 
and supported by a triangle was another dish containing the acid; 
all was enclosed by a tightly fitted bell-jar, and allowed to stand forty- 
eight hours, at the end of which time, according to both Schlosing 
and Neubauer, all the ammonia had been liberated and driven off from 
the liquid by the alkali and absorbed by the acid above. That this 


1 NEUBAUER and VOGEL: Anleitung zur Analyse des Harns, 1898, p. 745, 
ed. (TO. 

2 ScHLOsING: Annales de chimie et de physique, 1851, xxxi, p. 153. This 
paper contains a cut of the apparatus. The paper was translated in full into the 
German and again published in the Journal fiir praktische Chemie, 1851, lii, 
P- 372. 

3 NEUBAUER: Journal fiir praktische Chemie, 1855, lxiv, p. 177. 

4 In a later paper (Journal fiir praktische Chemie, 1855, Ixiv, p. 278), 
NEUBAUER proposed to use 20 c.c. of urine. 


342 Philip Shaffer. 


is incorrect will be shown below. The authors were led into the 
error by using standard acid and alkali of such a strength as to make 
it impossible to measure the small quantities of ammonia given off on 
standing longer than the first forty-eight hours. Neubauer used to 
absorb the ammonia, 10 c.c. of sulphuric acid of such a strength as to 
demand for saturation 225.42 mg. ammonia, or the acid was about 
33 per cent normal. The alkali was of such a strength that I c.c. 
was equivalent to 14.7 mg. ammonia, or was almost 87 per cent 
normal. I have found cases where IO c.c. urine, under the most 
favorable conditions for the Schlésing method, gave off more than 0.5 
mg. ammonia in addition to that given off during the first forty-eight 
hours. This would make a difference of 50 mg. per litre, which in 
many instances would be one-sixth of the total amount in a twenty- 
four hour quantity of urine. This 0.5 mg. ammonia would be equiva- 
lent to only about 3'5 of a cubic centimetre of the alkali used by 
Neubauer. Thus may be explained Neubauer’s and Schlosing’s 
‘“complete results after forty-eight hours.” 

Conditions governing the Schlésing method. — To get even approxi- 
mate results with this method, it is necessary to fulfil certain 
conditions, few of which have apparently been considered by any 
investigator using or describing the method in recent years. Both 
Schlésing and Neubauer understood the more important of these 
conditions, and mention them in a general way in their papers, but 
nowhere else do I find any exact description either of the most suit- 
able apparatus, or of the proper manner of carrying out the operations. 
In most textbooks, the only method given for the determination of 
ammonia in urine is the Schlésing method, and with one exception 
the directions are either too inexact to be of value, or are entirely 
incorrect. Asa result, the method has been used frequently under 
the most unfavorable conditions, and often has given undoubtedly 
very erroneous results. In view of the absence of an exact under- 
standing of this method, it seems worth while to discuss it in some 
detail. 

In order to judge of the method’s real value, it was first necessary 
to learn the conditions under which it could be worked most success- 
fully! The Schlosing method depends upon the fact that free ammo- 


' It was only after learning the difficulties in carrying out the method according 
to the directions given in textbooks, and when I had already discovered the 
source of the error as thus worked, that the very general descriptions contained in 
the papers of SCHLOSING and NEUBAUER were fully appreciated. ScuLOsiNnG did 


Quantitative Determination of Ammonia in Urine. 333 


nia is given off from a liquid containing it on being exposed to the 
air, and that from the air it will be absorbed by an acid. This trans- 
fer will take place completely, if carried out in a small closed vessel 
for a sufficient length of time. The time necessary for all the ammo- 
nia to be given off from a liquid and absorbed by actd depends on the 
depth of the liquid. The following experiment with urine will illustrate 
this fact. The urine and milk of lime was let stand for seven days 
in a small desiccator (6.7 cm. diameter) containing a titrated amount 
of acid. At the end of the time, the acid in each was titrated, and 
the ammonia calculated for one litre of urine. In the first was 
50 c.c. urine + 5 c.c. milk of lime; and in the second, 10 c.c. urine + 
3 c.c. milk of lime. 


Volume of liquid. | Depth of liquid. N Hs in litre. 
| 


16 mm. 581 mg. 


4 mm. 714 mg. 


From the fact illustrated in this experiment I have concluded that 
the layer of liquid should not be deeper than 2 mm. This would 
allow the use of about 25 c.c. of solution in a dish 12 cm. in diameter. 
Schlésing used 35 c.c. in a dish of this size. This consideration of 
the depth of the liquid is not mentioned in books describing the 
method, and has been disregarded by all recent investigators, as is 
shown by the fact that where they do give any of the details of their 
manner of carrying out the method, they state the amount of urine 
used, but say nothing of the diameter of the dish containing it. 

The Schlosing method as described in textbooks: 


Huppert’ directs one to use 25 c.c. urine + to c.c. milk of lime, but does 
not mention the size of dish or beli-jar. He suggests the use of ‘ acid 
dishes” to contain the urine. These are as a rule about 10 cm. in 
diameter, making the layer of liquid about 5 mm. deep. He lets the 
urine stand in the apparatus from three to four days, and claims that in 
that time almost all the ammonia will be given off. 


indeed have an accurate understanding of these conditions, but NEUBAUER placed 
so little emphasis upon them in the application of the method to urine that 
it is not surprising that the points have escaped the attention of succeeding 
investigators. 

1 NEUBAUER and VOGEL: Anleitung zur Analyse des Harns, 1898, toth ed., 
p- 742. 


334 Philip Shaffer. 


Salkowski! uses a crystallizing dish to hold the liquid containing the am- 
monia. To 25 c.c. of filtered urine’ he adds the same volume of milk 
of lime, places the mixture under a bell-jar with % or /%, acid,® and 
lets it stand two or three days. Salkowski likewise does not mention 
the size of the apparatus. 

Hopkins‘ directs one to use 25 c.c. urine + 20 c.c. milk of lime in a 
“basin with vertical sides’ and to let stand three days, when all the 
ammonia will be absorbed by the acid. 


Halliburton ® advises the use of 20 c.c. urine + to c.c. milk of lime placed 


in a beaker and covered with a bell-jar. In forty-eight hours, according to 
this author, all the ammonia will be driven off and absorbed by the acid. 
Halliburton makes a further direction which is, to say the least, surprising. 
He proposes to make a control determination by letting the same amount 
of urine stand in a similar apparatus, without lime, his idea being to sub- 
tract the amount of ammonia formed by putrefaction from that obtained 
in the first estimation. Without a preservative, a bacterial decomposition 


will take place in urine, and even after forty-eight hours in many cases, an‘ 


ammoniacal odor will be noticeable, and the urine will be alkaline. 


1 SALKOWSKI: Practicum der physiologischen und pathologischen Chemie, 
1900, 2d €d., p.. 252. 

2 In most cases it will be found advisable to filter the urine from the pus, etc., 
which is usually present, thus lessening the tendency toward decomposition. 
Most authors describing the method prescribe the use of filtered urine. 

8 The idea seems to exist that the rate of absorption of the ammonia depends 
upon the strength of the acid. This is probably the reason why the earlier 
workers used such unusually strong solutions for the purpose. Ina recent paper 
by NENCKI and ZALESKI (Zeitschrift ftir physiologie Chemie, 1901, xxxiii, p. 203), 
they state that in using their own method (discussed later), with urine an acid 
stronger than /% is necessary. This idea is not correct. So long as the acid 
remains in moderate excess, it will absorb the ammonia as rapidly as it is liberated 
from the liquid containing it. ‘The following determinations were carried out under 
identical conditions in an ammonium chloride solution. The volume of the liquid 
was in each case 30 c.c., and the time seventy-four hours. The results are given in 
c.c. % NH,. 


10 


Amount of acid used, With MgO. With Ca(OH ),. 


10 c.c. HySOy = 78.07 cc. ys 22.97 


10 


35 cc. ty HeSO,y 


' Hopkins: ‘* The chemistry of the urine,” in SCHAFER’S Textbook of Physi- 
ology, 1900, i, p- 586. 
® HALLIBURTON : Chemical physiology and pathology, 1891, p. 806. 


Quantitative Determination of Ammonia in Urine. 335 

Where this occurs, the ammonia formed from the decomposition of the 

urea will obviously first tend to neutralize the acidity of the urine, and the 

excess will be given off. It is this excess only which would be contained 

in the correction. It has furthermore been shown that the addition of 

milk of lime prevents the bacterial decomposition which would otherwise 
occur." 

Fresenius” gives the correct description of the Schlosing method as pro- 
posed in its author’s original communication. Fresenius used 35 c.c. or 
less of liquid in a flat dish 10 cm. to 12 cm. in diameter, and let it stand 
under a bell-jar forty-eight hours, when he tested the completeness of the 
expulsion of the ammonia and of its absorption by introducing into the 
jar a moistened strip of red litmus paper. If the paper changed in color, 
the apparatus was allowed to stand longer. Even before Neubauer had 
used the method, Fresenius had quoted this principle direct from 
Schlésing’s paper, in an early edition of his ‘‘ Quantitative Analyse.” 
His early description has probably not been altered, thus explaining its 
correctness. Evidently this author has not been consulted on this point 
by writers of books describing methods for urine analysis. 


The Schlosing method as used by previous investigators. — For the 
reason that they have given no consideration to this condition of 
the depth of the layer of liquid, it is impossible to form any correct 
estimate of the accuracy of the results of the many workers who have 
used this method.* They worked little if any with ammonium salt 
solutions of known content, and had no reliable method to use as 
a check upon their experiments. 


Bohland proposed a modification of the Schlosing method, in first creating a 
vacuum in the bell-jar, thus hastening the expulsion of the ammonia, so 
that he claimed it complete in forty-eight hours. 


1 SALKOWSKI: VIRCHOW’S Archiv, 1873, lviii, p. 486, and Centralblatt fiir die 
medicinischen Wissenschaften, 1880, xxxviii, p. 690; SALOMEN: VIRCHOW’S 
Archiv, 1884, xcvii, p. 150; HALLERVORDEN: Archiv fiir experimentelle Path- 
ologie und Pharmakologie, 1880, xii, p. 237; NEUBAUER originally proposed this 
correction in his paper, Journal fiir praktische Chemie, lxiv, p. 278, but its error 
was recognized and pointed out by the authors quoted above. It seems remark- 
able to find such a procedure prescribed in a book so recent as HALLIBURTON’S. 

2 FRESENIUS: Quantitative Analysis, 1899, p. 220. 

8 MUNK: VIRCHOW’s Archiv, 1877, ]xix, p. 365; HALLERVORDEN: Archiv fiir 
experimentelle Pathologie und Pharmakologie, 1880, xii, p. 237; BOHLAND: Archiv 
fiir die gesammte Physiologie, 1891, xlviii, p. 32; CAMERER: Zeitschrift fiir Biol- 
ogie, 1899, Xxxviii, p. 237. 


336 Philip Shaffer. 


Hammarsten! states that correct results can be obtained with the Schlosing 
method only with this modification. My own experience has been that 
although the vacuum does materially aid the expulsion of the ammonia, 
this advantage is more than offset by the difficulty in keeping the vacuum. 
Obviously the vacuum must not be made again after the alkali has been 
added to the urine. 

Camerer quotes a series of determinations according to Schlosing by himself, 
and a series in the same urines according to the same method by 
Soldner. The figures compare very poorly, and make an excellent exam- 
ple of the untrustworthiness of results by this method as ordinarily carried 
out. The average difference between the published results of these 
workers (in the same urines) is to mg. in too c.c. or fully one-seventh 
of the total amount. 

Camerer got higher results in nearly every case and in less time, and con- 
cluded that the reason was that he used a smaller apparatus than Soldner. 
During the present work it has been learned from experiment that the 
size of the bell-jar has no effect upon the absorption of the ammonia. 


The alkali used to liberate the ammonia in the Schlosing method. — 
The alkali used to liberate the ammonia from its salts determines to 
some extent the rate of its expulsion from a liquid, presumably in that 
on it depends its solubility. Munk? used sodium hydrate, but this on 
standing decomposes some substance in the urine and gives too high 
results. Milk of lime or magnesia has been most commonly used 
for the purpose. By magnesia the ammonia is given off very slowly, 
and, according to Soldner, not completely on account of the formation 
of ammonium-magnesium phosphate crystals.® Milk of lime gives 
off the ammonia more rapidly, but causes decomposition on long 
standing, especially in pathological urines.* 

The following results ®° show the rapidity of the expulsion by lime 
as compared with magnesia. The determinations were carried out 
under the same conditions, the only difference being in the alkali 
used. The figures show also the high results with lime and the 
incomplete results with magnesia. The temperature was about 24° 


1 HAMMARSTEN: Lehrbuch der physiologischen Chemie, 1899, p. 482. 
2 MuNK: Loe, cit. 

8 See CAMERER: Loc. ct¢. Some of my own results with magnesia have been 
too low after standing four hundred and fifty-six hours, or nineteen days. 

4 HALLERVORDEN: Loc. cit. 

® Iach experiment and result quoted in this paper has been chosen as repre- 


senting a fair average of a large number of similar operations. 


Quantitative Determination of Ammonia in Urine. 337 


C. Sodium carbonate prevents putrefaction in urines as does lime 
(see page 335). 


50 c.c. urine contained 39 mg. NH3. 


50 c.c. urine. With Ca(OH )s. 


48 hours 
96 hours 


144 hours 


Where absolute figures are given in this paper, they have 
been obtained by one or both of the accurate methods described 
below (pages 344 and 348). 


In my experiments I have used sodium carbonate with greater suc- 
cess than with either lime or magnesia. It drives off the ammonia 
much more rapidly than does magnesia, and does not cause decom- 
position as does lime. In order to test this point, some experiments 
were performed with a 4 per cent solution of peptone which was free 
from ammonia and contained chloroform as a preservative. In 25 c.c. 
of this solution standing forty-eight hours at 35° C., 


Lime gave off 4.5 gm. NHs, 
Magnesia gave off 0.6 mg. NHs, 
Sodium carbonate gave off 0.85 mg. NH;3. 


At 20° C., the decomposition was somewhat less in each case, but 
bore the same relation, the decomposition from lime being much 
greater than from either magnesia or sodium carbonate. The follow- 
ing shows the relative speed with which the ammonia is given off by 
sodium carbonate, and again indicates the decomposition by lime: 


50 c.c. urine contained 19.72 mg. NHs3. 


Time. With Ca(OH)>. With Na,CO,.2 


48 hours 19.89 mg. 19.04 mg. 


96 hours 21.08 mg. 19.72 mg. 


1 About 0.5 gm. to 1 gm. dry NasCOs for each 25 c.c. urine. 


a3 Philip Shaffer. 


Between the lowest and the highest of these figures there is a 
_ difference of 2.04 mg. which would amount to 40.8 mg. per litre. 
With none of the generally used methods for determining ammonia 
could a greater constant correspondence be obtained, and such results 
have accordingly been considered satisfactory. By the use of meth- 
ods which will be described further on, a degree of accuracy within 
5 mg. to 10 mg. tn the litre is attained without difficulty. 

The effect of temperature upon the Schlosing method. — Temperature 
also has a great influence upon the rapidity with which the ammonia 
is driven off. Its effect is shown by the following experiment. Four 
determinations in urine were carried out under the most favorable 
conditions, but at different temperatures. The alkali used was 0.5 
gram dry sodium carbonate for each 25 c.c. urine. The results are 
given in milligrams of ammonia for 50 c.c. urine. 


50 c.c. of the urine contained 19.89 mg. NHs3. 


Time. 16° C. 


48 hours | 17.17 
108 hours ). 18.96 


156 hours aperiee 18.96 


? In the residues standing at 12° C. and at 16° C., there was found 0.51 mg. N Hs. 
This, added to the amount absorbed by the acid, will scarcely equal the correct 
result. The slight difference was probably lost in opening the apparatus at the 
end of each period to titrate the acid. 

2 A decomposition is to be observed with Na,COg at 38° C. even in normal urines. 


At 12° and 16° C. we see from the figures above how slowly it 
comes off. At such temperatures the expulsion is probably not com- 
plete even aftera week. I have tried letting the desiccators stand in 
an incubator at 38° C., and with some success, but at this tempera- 
ture, even with magnesia, there is a slight decomposition which 
renders the results too inaccurate for scientific work. 

The Schlosing method as best carried out.— The most accurate 
results with the Schlésing method may be obtained as follows. To 
filtered urine add sodium carbonate and let stand with + acid in a 


10 
desiccator or under a tightly fitting bell-jar. With 25 c.c. urine, about 


a 


Se 


Quantitative Determination of Ammonia in Urine. 339 


half gram sodium carbonate, and an excess of sodium chloride,’ 
standing in a dish from 15 cm. to 17 cm. in diameter, at 20 C. or 
higher, the expulsion of the ammonia will be almost complete in three 
to four days. The length of the operations may be reduced to forty- 
eight hours by letting the apparatus stand at 38° C. On longer 
standing at such a temperature the ammonia from decomposition 
becomes considerable,” and even after the first forty-eight hours the 
results are not always reliable. When a smaller dish is used to hold 
the urine, a correspondingly smaller amount must be used, the best 
conditions being, as has already been mentioned, to have the depth of 
the liquid not more than 2mm. For the same amount of urine ¢/e 
wider the dish the more rapid will be the expulsion of the ammonta. 

In order that the liquid may cover the entire surface of the dish, 
and that the layer may not be thicker in some parts than in others, 
the dish should have a flat bottom. Crystallizing dishes or the 
bottom of large desiccators are adapted to the purpose. The 
Schlésing method as thus described may be considered satisfactory 
in many respects for clinical purposes; but it can of course be used 
only where one can wait two days or longer for the results. The 
uncertainty of the method, however, renders it quite unsuitable for 
more accurate work. 

Tables of results by Schlosing method. — The following tables con- 
tain results of determinations in urines according to the Schlosing 
method. The conditions of the operations are stated in each case, 
except that the depth of the liquid has always been about 2 mm. 
The ammonia has been determined also by one or both of the accurate 
methods described further on, and it is the average of these results 
which is given as “ammonia actually present.” The results con- 
tained in the tables represent the greatest accuracy that may be 
obtained with the Schlosing method. 


1 Enough sodium chloride to make a saturated solution prevents to some extent 
any decomposition which may otherwise take place, presumably by lessening the 
amount of dissociation of the alkali. 

2 In each case a few drops of chloroform have been added to prevent bacterial 
growth. 


40 


Philip Shaffer. 


Urine No. 1. 


50 c.c urine. 


SUrGG: 


Cc 


“cc 


“cc 


. urine 


Normal. 


Ammonia actually present 


Sysio) (Ge 


sis} (C- 


96 hours 
27 days 


192 hours 


Bie 
44 
44“ 

56. 

pat 
100 

56 

a ae 

43. 
Lobe 

S4 

43. 

4 


MgO 


“e 


Urine No. 2. Normal. 


Ammonia actually present 


g0c1G. 


48 hours 


120 
LOO 
48 
100. 
48 ‘ 
+56 


Na,CO,; 


“e 


Ca(OH). 


MgO 


Gm. NHg 
per litre. 


0.490 
0.459 
0.527 
0.489 
0.506 


0.467 
0.498 
0.476 
0.467 
0.484 


| Difference. 


Mg. 


Quantitative Determination of Ammonia in Urine. 341 


Urine No. 3. 


Normal. 


Gm. NH, | Difference. 


per litre. Mg. 
Ammonia actually present . . . 0.742 
25 c.c. urine 22°C. 117 hours MgO 0.727 =15 
After standing 117 hours the am- 
monia remaining in the liquid 
was determined according to 
Folin (page 344). This amount 
added to that above gave. . . 0.741 
Urine No. 4. Normal. 
Ammonia actually present . . . 0.328 
25 c.c. urine 22°C. l1v7hours MgO 0.234 as 
After standing 117 hours the am- 
monia remaining in the liquid 
was determined according to 
Folin (page 344). This amount 
added to that above gave. . . 0.326 — 2 
Urine No. 5. Normal. 
Ammonia actually present . . . | 0.398 
25 c.c. urine 38°C. 48hours NasCO; 0.394 ee 


Zpec. BSC son ej 0.440 42 


Urine No. 6. Normal. 


sade Ammonia actually present . . . 0.394 


25 c.c. urine 22°C. 96hours Ca(OH), 0.421 +27 


POCe:, - eo Geae 9G): Na,CO3 0.394 0 


Urine No. 7. Contained large quantity of blood. 


Ammonia actually present 


25 c.c. urine 22°C. 134hours Ca(OH), 0.462 +69 


342 Philip Shaffer. 


IlI. VAcuuM DISTILLATION METHOD ACCORDING TO WURSTER. 


In 1887 Wurster! proposed what is generally considered the 
original method for determining ammonia by distillation in a vacuum 
at low temperatures. He proposed to distil at about 15 mm. pres- 
sure, and 50° C., 5, 10, or 20 c.c. urine with baryta water, magnesia, 
or lime, and claimed the operation complete after five, ten, or fifteen 
minutes, depending upon the amount of urine used. The use of such 
small quantities of urine for the determination greatly lessens the 
accuracy of the method by multiplying any existing error. 

The ammonia is not completely given off under these conditions in 
the short time named by Wurster, as the following will show : 


20 c.c. of urine was distilled in a round bottom litre flask with 15 c.c. of a cold 
saturated solution of barium hydrate? at 15 mm. pressure and 50° C. 
After fifteen minutes from the beginning of the boiling, the operation was 
stopped and the acid titrated. The distillation was then continued for 
two similar periods. 


15 minutes gave per litre 2.057 gm. NH; 
30 ce 6" Ee OS 2-662. fim IN Eds 


45 ee eS 2 CS OsO MN ele 


These results would have been considered satisfactory by Wurster, 
but a higher degree of accuracy has been desired in the present work, 
and it is with such a standard in view that the methods are criticised. 

The greatest objection to the method, as used by Wurster, is, how- 
ever, the practical difficulty. His apparatus was bulky and of an 
inconvenient arrangement. A troublesome foaming which always 
occurs with either of the alkalies used by Wurster frequently ruins 
the operation, — by the foam getting into the acid, — necessitating 
many repetitions. Both magnesia and lime cause more foaming than 
does baryta. 

To avoid this Wurster and others® used different arrangements of 


' WoursTER : Centralblatt fiir Physiologie, 1887, i, p. 485; also Berichte der 
deutschen chemischen Gesellschaft, 1889, p. 1903; see the BouSSINGAULT method, 
p- 345. 

* With both lime and magnesia the operation is still less complete in the same 
time. 

’ NENCKI und ZALESKI: Archiv fiir experimentelle Pathologie und Pharmak- 
ologie, 1895, xxxvi, p. 385; SCHWARTZ: Wiener medicinische Wochenschrift, 
1898, iii, p. 98; SOLDNER: In a paper by CAMERER: Zeitschrift fiir Biologie, 
1899, xxxvili, p. 236. 


eee EE 


Quantitative Determination of Ammonia in Urine. 343 


flasks, etc., intended to catch the foam, which complicate the appa- 
ratus and rarely accomplish the purpose. 

The modifications by Nencki and Zaleski, and by Soldner. — Nencki 
and Zaleski modified the method of Wurster by heating slowly to 
35. C., and then distilling at this temperature for three hours or 
longer. Correct results may be obtained with this modification, but 
the unusual length of the operation, and the complexity of the appa- 
ratus, do not recommend it for general use.' Soldner distils in a 
partial vacuum (about 15 mm.) at 50° C., but renders the apparatus 
still more bulky by introducing two extra Woulf’s flasks and a Liebig 
condenser. The condenser is used both to lessen the foaming and 
to prevent the boiling of the acid. With a properly arranged wash 
bottle, to contain the acid, its boiling will cause no trouble, and such 
a contrivance is not necessary. To absorb the ammonia completely, 
a much more effective arrangement is that of two wash bottles shown 
in figure, page 348. 

Séldner uses 50 c.c. urine + 10 c.c. milk of lime? and distils off 
about one-fifth of the liquid. The time will vary from thirty to sixty 
minutes. During such a time a slight decomposition sometimes 
takes place, as may readily be shown by continuing the distillation 
for a second sixty minutes with new acid. 


IV. THE Foutin METHODS. 


Folin * described a method based upon the distillation of the urine 
with magnesia in two periods of the same length. The difference 
between the ammonia received during the second period (which was 
formed by the decomposition of the urea) and the total obtained 
during the first period was thought to be the preformed ammonia. 
On continuing the distillation for four or more successive periods of 
forty-five minutes each, the amount of ammonia given off in each 
period after the first was, if worked carefully, remarkably constant. 
The results by this method were repeatedly lower than those accord- 


1 Each determination takes about four hours. This method has been used 
largely for determining ammenia in blood. 

Two parallel determinations in dog’s urine by these authors have a difference of 
88 mg. ammonia per litre. The method can, however, be worked more accurately 
than this would indicate. 

2 As stated on page 336, SOLDNER claims that with magnesia the results are 
always low. 

3 FoLin: Zeitschrift fiir physiologische Chemie, 1go!, xxxii, p. 515. 


344 Philip Shaffer. 


ing to Schlésing however, and by working with known artificial mix- 
tures of urea and ammonia were found to be in every case too low by 
as much as 5 mg. for the amount of solution used (25 c.c. or 50 c.C), 
or as much as 200 mg. per litre. This is explained by the fact that 
on boiling an aqueous solution of urea, no point can be reached, so 
long as undecomposed urea remains, at which the boiling liquid will 
not contain a considerable amount of ammonia,! whereas before 
boiling the liquid is free from ammonia. Consequently the amount 
of ammonia given off from the liquid is less, by the amount retained, 
than that actually formed. Likewise on boiling urine with an alkali, 
all of the ammonia formed from decomposition of the urea during the 
first period is not given off. The preformed ammonia is of course 
all given off during the first period. 

At the beginning of the second and following periods the liquid 
will contain its maximum of ammonia from the decomposition, and 
consequently all formed during these periods will be given off. The 
error being dependent upon the amount of urea in solution was not 
constant in different urines, and so destroyed the usefulness of the 
method. This I reported to Dr. Folin and in conjunction with him 
tested it further and arrived at the above conclusion as to its cause. 

Later Dr. Folin devised a method for the determination of ammonia 
in other animal fluids, as well as in urine, which he will shortly de- 
scribe in detail in Hoppe-Seyler’s Zeitschrift fir physiologische 
Chemie. The reliability of this method for urines is established. 
It has been used in this work and its results compared with those of 
the other methods. Tables of results will be found at the close of 
this paper. The method is carried out as follows: 

To 25 c.c. or 50 c.c. urine (or solution of ammonium salt), placed 
in an areometer cylinder about 45 cm. high and 5 cm. in diameter, is 
added 1 gram or 2 grams sodium carbonate and 8 grams or 16 grams 
sodium chloride. A current of air is driven through the urine, 
carrying off the ammonia set free by the sodium carbonate, and 
allowing it to be absorbed by passing through a definite amount of 
ig acid, the excess of which is afterward titrated and the ammonia 
calculated. The length of time necessary for the completion of the 


1 , a ae e es . . . . . 
About 2.5 c.c. #4 in 200 c.c. liquid. After reaching about this amount, the 


ammonia in the solution at any one time will remain practically constant. 
* The sodium chloride is added here, as in the SCHLOSING method, to prevent 
decomposition, The amount of sodium carbonate and of sodium chloride depends 


upon the amount ol urine used. 


Quantitative Determination of Ammonia in Urine. 345 


operation will depend on the amount of air passing through the liquid, 
as well as upon the temperature. At from 22° to 25° C,, and with an 
air current of approximately 700 litres per hour, the ammonia will be 
completely driven off from 25 c.c. of urine in an hour and a quarter, 
or from 50 c.c. in an hour and ahalf. With a larger volume of urine, 
a smaller air current, or lower temperature of the urine, a longer time 
will be necessary. The requisite time must be learned for each air 
current. This can easily be done by continuing the operation until 
in an hour there is no ammonia given off. It is necessary to pass 
the air through a tuft of absorbent cotton or closely packed glass 
wool, after it leaves the urine and before it passes through the acid, 
in order to prevent alkali being mechanically carried over. The air 
current should be uniform and constant. Folin has used a specially 
constructed tube for the absorption of the ammonia by the acid. 
Where this is not done, it will usually be found necessary to pass the 
air through two successive portions of acid to prevent any loss of 
ammonia. The application of this method to pathological urines and 
to dog’s urine will be discussed in comparing it with the other accurate 
method yet to be described. 


V. THE VAcuuM DISTILLATION METHODS ACCORDING TO 
~ BouSSINGAULT. 


Boussingault in 1850 published’ a paper in which he described a 
method for determining ammonia in solution of its salts and in urine 
by distilling to dryness in a vacuum with milk of lime at 40° to 
50°C. The distillation lasted in most cases about onehour. Sodium 
carbonate and bicarbonate were also used in place of lime. The 
ammonia was absorbed by standard acid and titrated. This is un- 
doubtedly the original application of vacuum distillation to the deter- 
mination of ammonia ; but the paper received little attention and is 
apparently not known to workers in this line at the present time. 
Wurster and those who have modified his method have evidently 
not been aware either of the work of Boussingault or of the criticism 
it received from Lehman, Neubauer, and Hoppe-Seyler. These 


1 BoussINGAULT: Annales de chimie et de physique, 1850, xxix, p. 472. 
Translated and published in the Journal fiir praktische Chemie, 1850, li, p. 281. 
Also in BoUSSINGAULT: Mémoires de chimie agricole et de physiologie, p. 285. 


346 Philip Shaffer. 


authors passed adverse criticism upon the method, and declared in 
favor of the less reliable procedure of Schlosing. 

The Boussingault method has been carefully examined during 
this investigation, and it has been concluded that with simple modi- 
fication, which will be given, very accurate results are obtained. 
Even as originally described by Boussingault this method is not 
only superior to the Schlésing method, but is to be preferred to any 
of the forms of the same method proposed by Wurster, Nencki, or 
Soldner. 

Boussingault claimed that it was necessary to evaporate the liquid 
to dryness in order that all the ammonia be given off, and quoted an 
experiment to prove the point. That this is not the fact one can 
readily see from the manner of carrying out the method given farther 
on. Boussingault did not mention foaming, but with lime he un- 
doubtedly experienced this trouble. He used sodium carbonate very 
largely, however, and with this there is no foaming. . 

Boussingault found no decomposition on distilling urea solutions 
in a vacuum with either lime, sodium carbonate, or bicarbonate at 
50 C., and therefore concluded that there was no decomposition in 
urine under the same conditions. He records, however, no experi- 
ments in which he continued the distillation of urine for a second 
period to discover any decomposition. 

The following experiment confirms the opinion of Boussingault 
that the expulsion of the ammonia from solutions is complete, and 
also that there is no decomposition of urea: 


50 c.c. of an ammonium chloride solution containing about 25 per cent urea, 
+ 2 gms. sodium carbonate was distilled in a vacuum at 48° to 50° C. for 
thirty minutes. The ammonia obtained was 29.07 mg. The theoretical 
ammonia in 50 c.c. of solution was 29.04 mg. 20 c.c. water was then 
added and the liquid distilled a second thirty minutes at the same 
temperature. 


This distillation gave absolutely no ammonia. 


The following experiment, selected from a large number of similar 
ones, proves the reliability of the method for many normal urines: 


50 c.c. urine + 2 gms. NagCO,. 
rst distillation gave 11 c.c. 7% NHs, 


2d distillation, 70 min. at 50° C., gave o. 


—— oe 


Quantitative Determination of Ammonia in Urine. 347 


That there is a slight decomposition in some urines, however, I 
have proved by repeatedly continuing the distillation for a second 
and third period after the ammonia had been driven off (as in the 
above experiments). In many cases the second period gives a small 
amount of ammonia, which remains constant in suceeeding periods, 
thus proving the decomposition continuous. In many pathological 
urines this will amount to as much as 30 mg. in the litre.! 

To shorten the period of distillation, without using too small 
amounts of urine, to a point where this slight decomposition might 
be disregarded, seemed desirable. Lessening the time needed to 
make an ammonia determination would of course render the method 
still more valuable. This was done by adding to the urine a quantity 
of methyl alcohol, thus lowering the boiling point of the mixture so 
that a yet more rapid ebullition took place, and more of the liquid 
distilled over, carrying with it all of the ammonia. 

The time of the distillation may in this way be reduced to fifteen 
minutes, and the entire operation may, after some little experience, be 
completed in less than half an hour. During the fifteen minutes of 
the actual distillation, the decomposition is so slight that it will rarely 
amount to more than 3 mg. in a litre. In pathological urines 
containing large quantities of albumin the error may be somewhat 
greater, but in such cases the addition of an excess of sodium chloride 
(enough to form a saturated solution) will prevent the decomposition 
being more than the amount just stated. 

Sodium carbonate is preferable to the other alkalies used, because 
with it, as has been said, there is no foaming, and the operation may 
be carried out without difficulty. With sodium carbonate the de- 
composition is less than with lime, and no greater than with mag- 
nesia (see page 337), and the expulsion of the ammonia is as rapid 
when liberated by sodium carbonate from its salts in solution as it is 
from a solution of free ammonia. The exact amount of sodium car- 
bonate added is not important. The decomposition from even a large 
excess is inconsiderable. One gram for 50 c.c. urine is always suf- 
ficient, and with this amount the decomposition in fifteen minutes is 
practically nothing. 


As will be seen, this amount is far within the boundaries of error in duplicate 
analyses by the SCHLOSING method. (See tables of SCHLOSING results.) 


348 Philip Shaffer. 


VI. THe Vacuum-DISTILLATION METHOD AS BEST CARRIED OUT. 


The most satisfactory manner of determining ammonia in urine by 
vacuum distillation is as follows: The arrangement of apparatus is 
shown in the figure. 

To 50c.c. urine in flask A add an excess (15 or 20 grams) of sodium 
chloride, and about 50 c.c. methyl alcohol. In bottle 4 place 25 or 
50 c.c. 7% acid and in B’, 10 c.c. {4 acid, diluted in each case with a 
small amount of water. If too much water is added, there will be 
danger of loss of acid by jumping over during the violent commotion 
which is set up in the acid by the rapid passage of the steam. If 
such a loss should occur, the acid can always be recovered by rinsing 


>. 


Description of the Apparatus. — A is a round bottom litre flask having a two-hole rubber 
stopper with the two tubes. Tube a has a stop-cock to admit air at the end of the 
operation. This tube does not run below the liquid, as is usual in vacuum distilla- 
tions, but ends as shown in figure. £ and 4’ are areometer cylinders with rubber 
stoppers. They contain the acid to absorb the ammonia. Any wash bottles would 
answer the purpose. C is an ordinary filter flask which is connected with the 
vacuum pump. In the vessel 2, surrounding the flask, is water at 50° C., and in 
is placed ice and water. The cooling is not necessary, however. The glass tubes 
are connected at // with rubber tubing in order that the apparatus may be readily 
disconnected for titrating the acid. 


out the filter flask C. When the apparatus is ready, about 1 gram 
dry sodium carbonate is added to the liquid in flask A, the stopper 
quickly put in place, and the suction started. With a good pump the 
pressure will be reduced to about 10 mm. in two or three minutes, 
when, the liquid surrounding A being at 50° C., a rapid boiling will 
begin. The temperature is maintained, and the boiling allowed to 
continue for fifteen minutes. At the end of that time the ammonia 
will in all cases have been completely given off, and the operation 
may be stopped by slowly letting in air at the stop-cock in tube a. 
The acid in # and B’ is titrated and the ammonia calculated. 
Alizarin red (1 per cent aqueous solution) has been used as the 
indicator for the titrations. This substance is not affected by the 


Quantitative Determination of Ammonia in Urine. 349 


presence of ammonium salts, and is very little sensitive to carbonic 
acid. The end point of the titration is sharp. This indicator is used 
in this laboratory in all operations where solutions of ammonium salts 
are titrated. 

With this method, as well as in the Folin method already described, 
the ammonia may be determined with as great accuracy as may any 
other constituent of the urine. The results are in all cases correct 
within less than 10 mg. ammonia in the litre. 


TABLES OF RESULTS BY THE NEW FOLIN METHOD, AND BY ‘THE MODIFIED VACUUM 
DISTILLATION METHOD. 


Unless otherwise stated, “ By vacuum distillation,” mentioned in the following tables, 
means the operation as described on page 348. The amount of urine used for the deter- 
minations and the time is mentioned in each case. When the temperature is not men- 
tioned with the results by the Folin method, it was from 22° to 25° C, 


Urine No. 8. Normal. 


By the Folin method 50 c.c. urine 2 hours 
By vacuum distillation S0ices =< 15 min. 


According to Wurster 
Ba(OH), 50 c.c. 


Urine No.9. Normal. 


By the Folin method 25 c.c. urine 2 hours 


By vacuum distillation 5Oiees oS 15 min. 


Urine No. 10. Normal. 


By the Folin method 50 c.c. urine 2 hours 


By vacuum distillation Sica 15 min. 


Urine No. 4. Normal. 


By the Folin method 25 c.c. urine 50 min. 
Further air blast gave 0. 
By vacuum distillation 50 c.c. urine =15 min. 


According to Schlésing (see page 341). 


350 Philip Shaffer. 


Urine No. 8. Normal. 


Differ- 


oe Hs | ae 
per litre. | ‘Mig. 
ree A Biche ris | 
By the Folin method 25 c.c. urine 50 min. 0.741 | 
Further air blast gave 0. yoio8 2 
By vacuum distillation 50le:c, urine! |) 15) min: 0.743 | 
According to Schlosing (see page 341). 0.727 


Urine No. 11. Normal. 


By the Folin method 25 c.c. urine 1 hour 0401 
¢ PISO Las 0.408 

By vacuum distillation Oexe, 15 min. 0.410 
= Shere; My es 0.407 

< ‘ SO) ee, AG ot SISENCe 0.407 


Urine No. 12. Normal horse urine. 


By the Folin method 50 c.c. urine 1} hour 0.073 


Further air blast, 14 hours, gave 0. 


——~+- 
So 


By vacuum distillation 50 c.c. urine 15 min. 0.073 


3 " SOC es Sve 0.073 


Urine No.1. Normal. 


By the Folin method 25 c.c. urine 10 hours 0.486 
ff 25°C.C, a 1] i 0.496 
¥ 2SNC:C5 mus 3) a O.489 


Next 6 hours’ air blast gave 0 

By vacuum distillation 25 c.c. urine 15 min. 0.493 
Next 15 minutes’ distillation gave 0. 

By vacuum distillation 25 c.c. urine 30 min. 0.493 
Next 30 minutes’ distillation gave 3.4 mg. per litre. 


(See page 340 for Schlosing’s results with this urine.) 


Quantitative Determination of Ammonia tn Urine. 


Urine No. 2. Normal. 


Gm. NH. 


3 


per litre. 


By the Folin method 


vacuum distillation 


“ “ec 


Next 60 minutes’ distillation gave 6.6. mg. per litre. 


35! 


Urine No 13. Fever urine. Much albumin. 


the Folin method 5 c.c. urine 1} hours 


y vacuum distillation Seto ma 15 min. 


Next 15 minutes’ distillation gave 6 mg. per litre. 


Urine No. 14. Fever urine. 


y the Folin method 25-c.c. urine 0.646 


7 vacuum distillation S0iGee | 15 min. 0.663 


Same Urine. Later. 


y the Folin method 25 c.c. urine 0.659 
y vacuum distillation 25 Ciena 15 min. 0.673 
Next ]5 minutes’ distillation gave 3.4 mg. per litre. 
vacuum distillation 50 cc. urine 50min. 38°C. 


Next 50 minutes’ distillation at 38° C. gave 3.4 ing. 
per litre. 


Urine No. 15. Contained much blood. 


By the Folin method 5 c.c. urine 1 hour 


By vacuum distillation 50 c.c. 15 min. 


352 Philip Shaffer. 


Urine. Later. See page 341, Urine No. 7. 


ins NE || ee 


per litre. | Mg. 


By the Folin method 25 c.c. urine 2 hours 0.389 
. = SOieye 11 hours at 0° C. 0.393 
By vacuum distillation Sexe, 94 15 min. 0.389 


Next 15 minutes’ distillation gave 3.4 mg. per litre. 


Urine No. 16. Normal dog’s urine. 


By the Folin method 20 c.c. urine 2 hours 


By vacuum distillation ZA KCiCsmns 15 min. 


Urine No. 17. Normal dog’s urine. 


By the Folin method 50 c.c. urine 1} hours 


By vacuum distillation SOicc es 15 min. 


It will be observed from the tables above that there is in some 
urines a difference between the results according to the Folin method 
and those according to the proposed method by vacuum distillation. 
From a series of experiments it has been concluded that this occa- 
sional difference is due to the complete decomposition, under the con- 
ditions of the vacuum distillation, of an unstable substance which 
occurs in some urines in very small quantities. Under the condi- 
tions of the Folin method, it is not decomposed, but at a temperature 
of about 25° C. is driven off with the ammonia, and partially retained 
in the receiver, as may be shown by heating the acid solution to 
boiling, or by redistilling the ammonia from the solution. In this 
way from urines containing this substance more ammonia may be 
obtained than is shown by the original titration. 

That the substance is completely decomposed during the fifteen 
minutes of the vacuum distillation is proved by the fact that on con- 
tinuing the distillation for a second period under the same conditions 
there is either no ammonia obtained or but a trifling amount which 
could not account for the difference. 


Quantitative Determination of Ammonia in Urine. 353 


Regarding the identity of the unstable substance, I am not able to 
state. It seems to occur more frequently and in larger amounts in 
dog’s urine than in human urine, in which I have not found it to 
exceed 20 mg. ammonia in the litre. 

The following results from dog’s urine are quoted in this con- 
nection: 

When fresh, about six hours after being voided, the urine contained : 


Gm. N Hs per litre. 


By vacuum distillation — 25 c.c. — 15 min. — 50° C. 2.081 
According to the Folin method — 25 c.c. — 1 hr. — about 22° C. 2.023 
fe + og 25 c.c. — 11 hrs. — about 0° C. 2.026 


There can be no doubt of the completion of the expulsion of the 
ammonia under these conditions. After being treated with the air 
blast for eleven hours, the urine residue was distilled in a vacuum 
with methyl alcohol at 50° C. for fifteen minutes. There was ob- 
tained 37.4 mg. ammonia per litre. At the low temperature, the 
urine evidently held a part of the substance, which was then decom- 
posed during the vacuum distillation. 

This urine then stood eleven days, when the determinations were 
repeated. The unstable substance had during this time been decom- 
posed, and the results by the two methods agreed. 


Gm. NHs per litre. 


By vacuum distillation — 25 c.c. — 15 min. — 50° C. 2.091 
According to the Folin method — 25 c.c. — 2 hrs. — 22° C. 2.091 


In using the vaccum distillation method, it must therefore be con- 
sidered that it includes in its results the ammonia formed by the 
decomposition of this unknown substance, which, however, exists not 
only in such small quantities, but in so few urines as to make the 
fact of but little consequence. 


VII. Summary. 


1. The Schloésing method as described in textbooks, and as gener- 
ally used, is unreliable and its results are usually incorrect. The 
method depends upon certain conditions which have been entirely 
disregarded by previous workers. By fulfilling these conditions, 
results may be obtained which are satisfactory for clinical purposes. 


354 Philip Shaffer. 


2. The methods by vacuum distillation as described by Wurster, 
Nencki and Zaleski, and Séldner may give correct results, but, owing 
to practical difficulties, simpler methods are to be preferred. 

3. The method published by Folin in 1901 is inaccurate. 

4. The new method by Folin gives accurate and reliable results. 

5. The old but unknown method of Boussingault is quite as 
accurate as the other vacuum distillation methods and less cumber- 
some, 

6. With the use of modifications which are proposed the Bous- 
singault vacuum distillation method is quick and free from difficulty, 
and its results accurate. 

7. For scientific or other careful work the recent Folin method, or 
the modified vacuum distillation method should be used. With 
either of these methods a higher degree of accuracy than has yet 
been attained in ammonia determinations is easily possible. 

In conclusion I wish to thank Dr. Otto Folin for his kind counsel 
during the course of the work. 


thet tee 


THE INFLUENCE OF FATIGUE UPON THE SPEED OF 
VOLUNTARY CONTRACTION OF HUMAN MUSCLE. 


By THOMAS ANDREW STOREY. 


[From the Physiological Laboratory of the University of Michigan.) 


CONTENTS. 
Page 
Apparatus and method of work : 355 
The influence of fatigue upon the speed of contraction . 362 
Relation between height of contraction and speed of contraction 367 
Central and peripheral fatigue 369 


es the appearance of Mosso’s well-known paper ‘‘ Ueber die 

Gesetze der Ermidung,” a very large amount of work has been 
done by different investigators on the effect of fatigue upon the 
height of voluntary muscular contractions of short duration. The 
effect of fatigue upon the speed of voluntary contraction, however, 
does not seem to have been studied; at least an examination of the 
literature has failed to discover any papers bearing immediately upon 
this subject. It is generally admitted that voluntary muscular con- 
tractions are tetani, the motor cells of the cord sending impulses to 
the muscle, according to most of the later observers, at the rate of 
ten or twelve per second. The cause of the decrease in the height 
of the voluntary contractions as fatigue comes on is attributed by 
some writers to fatigue of the muscles themselves, while others 
believe that it is rather an expression of fatigue of the central 
nervous system. 

The experiments herein reported were undertaken in the hope that 
they might throw some light upon these points." 


APPARATUS AND METHOD OF WORK. 


The abductor indicis, the first dorsal interosseous muscle, was used 
throughout this investigation. This muscle has been found by a 


1 This work was done during the spring and summer of Igoo. 
35 


un 


356 Thomas Andrew Storey. 


number of observers, notably Fick and Lombard, to offer many 
-advantages over the flexors of the middle finger employed in the 
experiments of Mosso, and in many other ergographic studies. 
The abductor indicis lies close to 

5 e¢ the surface, is isolated, responds 
readily to making or breaking in- 

duction shocks, and is relatively 

feeble, and consequently does not 

require the use of heavy weights or 

FIGURE 1]. — Cross-section of finger-car- cumbersome apparatus. The hand 
riage. 1. Steel bearing, fitting into the man’ belfored! with@ut 19 Dade 
brass cylinder numbered 4. 2. Steel pe ae Lee oh od 
bearing in place. 3. Pulleys for attach- freedom of the movement of the 
ment of various forms of resistance and finger, and means for obtaining 


recording devices. 4. Brass cylinder accurate records of that movement 


rotating about the steel axis numbered : ; 
2. 5. Support for finger. 6. Horizon- C4 be easily devised. 


tal recording point. The ergograph (Figs. 1 to 4) 
used in this research was a new 

form devised especially for work with the abductor indicis muscle. 
It may also be used with the abductor minimi digiti muscle. The 
essential part of this machine is a device for communicating to a 


FIGURE 2.— General view of ergograph. 


recording point the angular movement of the metacarpo-phalangeal 
joint when acted upon by the abductor muscle. The recording point 
is carried at the end of a horizontal arm which supports the finger 
and rotates about a vertical axis (see Fig. 1). The distal end of the 


, Lnfluence of Fatigue on Contraction of Human Muscle. 357 


first phalanx is fastened to the supporting arm by a light brass clamp 
(see Fig. 2). The vertical axis supporting the horizontal arm is 


| cial 


Figure 3. — Working drawing of ergograph. Viewed from above. a and 4, arm-rests ; 
d, hand-rest and electrode combined; ¢, clamp resting against ulnar side of the hand; 
Jf, clamp between index and middle finger; g, clamp resting against ends of fingers ; 
h, clamp holding thumb in place; 7 and 4, clamps fixing index finger in place over 
the moving arm below. Dotted lines indicate position of finger-carriage when the 
muscle is contracted. The spring pointer on the end of the carriage is arranged for 
tracing the movements made on a horizontal drum. 


FicureE 4.— Working drawing of ergograph. Viewed from side. Lettering the same 
as in Fig. 3. 


formed of a brass cylinder with a closely fitting steel core (see Fig. 1 
The brass cylinder carries two pulleys (more may be added if needed). 


358 Thomas Andrew Storey. 


To one of these pulleys is fastened the cord to which the weight is 
attached. As in the case of an ordinary ergograph this insures con- 
stant tension of the finger throughout the movement, and, if the 
pulley be small, reduces the throw of the weights. 

The second pulley is for a thread which connects with a lever or 
other recording device. This arrangement permits the simultaneous 
records of the movement on different drums going at different speeds, 
one record being made by the point at the end of the horizontal arm, 
and the other by the recording device with which the thread from 
the second pulley is connected. These pulleys, or additional pulleys, 
may be used for other purposes. They may connect with various 
devices arranged to bring tension on the muscle ; or they may con- 
nect with recording mechanisms, such as Fick’s Arbeitsammler or 
Zuntz’s Ergometer. Of course the size of the pulleys can be chosen 
to suit the needs of the experiment. 

All parts of this ergograph were made as light as possible without 
sacrificing too much strength and stiffness.. The error from the 
momentum of moving parts was further reduced by the attachment 
of the weight cord to a small pulley. As far as could be: judged, the 
throw of the lever had little influence in the experiments under con- 
sideration. It may be noted here that in the movement of extreme 
abduction the throw of the horizontal arm was opposed by the elastic 
web of skin between the bases of the first and second fingers. This 
influence would in such an event assist in producing relaxation. It 
would not be present, however, in any but extreme movements of 
abduction. 

With slight modifications of the support for the hand and the base, 
the ergograph can be turned on its side, and the horizontal arm be 
then made to write on an upright drum, as in case of the model 
demonstrated by Professor Lombard at the meeting of the American 
Physiological Society in Chicago, December, 1901.! 

The weight method (isotonic) was used throughout in these experi- 
ments. The isometric method, introduced by Fick, and advocated by 
Franz, Hough, and Schenck, would be less satisfactory for the pur- 
poses of this investigation.? 


1 LOMBARD: This journal, 1902, vi, p. xxiv. 


* For a discussion of isotonic and isometric methods, see FRANZ: This journal, 
1900, iv, p. 348; HouGuH: This journal, 1go1, v, p. 239; TREVES: Archiv fiir 
die gesammte Physiologie, 1896, Ixxviii, p. 45; SCHENCK : Archiv fiir die gesammte 


Physiologie, 1900, Ixxxii, p. 384. 


Lifluence of Fatigue on Contraction of Human Muscle. 359 


The signal for voluntary contraction was given automatically by a 
rapidly revolving horizontal drum striking a metal peg against a 
wooden lever. The lever rested at right angles to the axis of the 
drum near its bearing. The peg was screwed at right angles into the 
axis of the drum so that on each revolution it would strike the lever 
and lift it. The sound of the impact was the signal for contraction. 
During that portion of the experiment in which the contractions were 
not to be recorded on the fast drum, an oblong brass plate was inter- 
posed between the recording point and the surface of the drum. 
This in no way interfered with the action of the muscle or with the 
continuity of the experiment. The brass plate was attached to a 
wooden rod like a metal flag fixed to its pole. The wooden rod was 
hinged to the bed of the fast drum. When the rod was in a hort- 
zontal position, it was out of service; when it was drawn up toa 
vertical position, the brass plate on its end came between the writing- 
point and the drum. The interposition of the brass plate could be 
accomplished at any time during the experiment. 

The time-record was made by the vibration of a tuning fork whose 
rate was fifty double vibrations a second. The vibrations were traced 
upon the horizontal drum by means of an electric time-marker or 
signal. In order that the time-records might not be superimposed 
at each successive revolution of the drum, the position of this electric 
time-marker was changed. automatically so that the time-record was 
inscribed as a spiral. This was accomplished by supporting the 
electric time-marker on a carriage which was driven by a long hori- 
zontal screw. The motion of the screw was derived by belts and 
pulleys from the rotation of the axis of the drum. The same carriage 
supported a signal which indicated the moment at which the primary 
current was made and broken in those experiments where the muscle 
was excited electrically. 

The recording point at the end of the horizontal finger-support 
was a horizontal, flexible phosphor-bronze writing-point (see Fig. 1). 
The writing-point moved parallel to and level with the axis of the 
horizontal drum. Its width gave it sufficient strength to be uninflu- 
enced by the movement of the drum. 

The lever-arm and writing-point magnified the shortening of the 
muscle about twenty-five fold. The amount of movement present at 
the proximal end of the first phalanx during greatest contraction, with 
no opposition, is about 4 mm.; the amount present in the region of 
the writing-point under the same conditions is about 100 mm. 


360 Thomas Andrew Storey. 


(These figures apply to the writer.!) It might be well to note here 
that this movement is not one of pure rotation. The metacarpo- 
phalangeal joint of the index-finger is a sliding joint. 


FiGuRE 5.— One-half the original size. Curves drawn by the abductor indicis at the 
beginning and the end of one hundred and three contractions against a weight of 
about four pounds, and at a rate of about forty-eight a minute. 


In working with the electrical stimulus, the make or break shock of 
the induced current was cut out automatically, by means of two levers 


UTP AAA DUO TTUNMU TEAR AATEC ATER MARAODOU OUTTA TUE TH OA CA 


| AT Wyn HALAL LOUDON TU TL EE 


FIGURE 6.— Upper curve marks time in seconds. Lower curve compares voluntarily 
and electrically stimulated contractions. Weight, unless otherwise indicated, 165 
grams. Course of experiment (read from left to right). Five electrically stimulated 
contractions.. Five voluntary contractions. Fatigue with voluntary contractions 
(weight, 1.5 kilograms; rate, two contractions a second). Series of voluntary con- 
tractions. Series of electrically stimulated contractions. 


like that described above, with the signal for voluntary contraction. 
The levers were provided with platinum points which dipped into 


' SCHENCK: Archiv fiir die gesammte Physiologie, 1900, Ixxxii, p. 3¢ finds the 
s » ) ? 
Same amount of movement at this joint. 


Influence of Fatigue on Contraction of Human Muscle. 361 


mercury cups. They were so arranged that the secondary circuit 
should be open during either the make or break of the primary. 
During an experiment, the meniscus in the mercury cup was fre- 
quently cleaned with a soft brush to prevent the accumulation of 
electrolytic products. 

In most cases records were made upon the two drums by means of 
devices described on pages 358and 370. The rate of one, the upright, 
drum was slow, giving an ordinary fatigue-tracing. (See Figs. 6 and 


DUDPOPPRPPARAPDDD | VEPIEE Pee eee TY 


FicurE 7.— Tracing of alternating electrically and voluntarily stimulated contractions 
with a 165 grams weight, before and after a series of voluntary contractions made 
against a weight of 2.7 kilograms, in which the muscle finally is unable to lift the 
weight. 


7.) The rate of the other, the horizontal, drum was rapid, giving 
myograms and time-records for the study of changes in different 
phases of each curve (see Fig. 5). 

A rapid change of weights was made possible by means of a lever 
device. Upon referring to Figs. 2, 3, and 4, it will be seen that this 
device is a lever of the third class, and that it consists of a long arm 
which supports the weight. There is a hub on the fulcrum-end of 
the lever over which runs the cord which connects the lever with the 
weight pulley on the ergograph. When ina horizontal position, the 
weight lever is put in motion by the movement of the horizontal arm 
of the ergograph. By sliding the weight along its support, the resist- 
ance is changed in amount. By lifting the weight lever to a vertical 
position, its resistance is taken out entirely. 


362 Thomas Andrew Storey. 


A platinum point on the hub of the weight lever can be made to 
dip into a mercury cup when the weight lever is in a vertical position. 
This mercury key may be used in this manner to close the primary 
circuit and thus make it possible to change from a voluntary to an 
electrical stimulation immediately upon taking out the resistance of 
the weight lever (Figs. 6 and 7). 


THE INFLUENCE OF FATIGUE UPON THE SPEED OF VOLUNTARY 
MuSCULAR CONTRACTION. 


The first set of experiments was performed in order to discover the 
influence of fatigue upon the speed of voluntary muscular contraction. 
Each experiment consisted of one hundred and three voluntary con- 
tractions against a weight of about one and a half kilograms. The 
rate of contraction was about forty-eight to the minute. At this rate 
and with this weight, one hundred and three contractions were 
enough to produce some fatigue. Experiments were made at inter- 
vals of several hours. Records of the first and last three contractions 
in each series were made on the fast drum (Fig. 5). These myo- 
grams were then measured. The results are shown in Tables I, II, 
and III. 

An examination of Table I brings out the following facts : — 

(1) The height of contraction grows less as fatigue increases. For 
instance, on the 6th of June at 9 A. M., the highest normal contraction 
was 58.7 mm., while the highest fatigued contraction was only 
22.6 mm. 

(2) The duration of the fatigued and lesser movement is sometimes 
as great as that of the normal and greater movement. This may be 
seen on comparing the duration of the first contraction with the 
duration of the one hundred and first contraction at 9 A.M. and at 
2.5 P.M. on June 6th. 

(3) The duration of the phase of rising energy, or, as it will be 
called in this paper, ‘“ the phase of shortening,” after fatigue appears, 
is ordinarily greater than when the muscle is unfatigued, even though 
the fatigued shortening is much less in extent. This fact may be 
noted in every comparison in Table I, except in that made between 
the third and the one hundred and third contractions at 4.25 P. M. on 
June 6. 

(4) The time required for the phase of relaxation after fatigue is 


Lnfluence of Fatigue on Contraction of Human Muscle. 363 


always less than when the musclé is unfatigued. This is due to the 
smaller distance through which the finger travels at that stage. 

Table II records the speed per second of total contraction, phase 
of shortening, and phase of relaxation in normal and fatigued volun- 
tary contraction. The figures were carefully compiled and are suffi- 
ciently accurate. They show that after the appearance of some 
fatigue, the speed of contraction is greatly reduced. The reduction 
varies with the fatigue produced. For example, the most rapid of 
the first three contractions recorded in Table II was made at the rate 
of 242 mm. per second, while the most rapid of the last three contrac- 
tions at the end of that series, was made at the rate of 110 mm. per 
second. 

The question arises, is this reduction in speed confined to some 
phase of the whole contraction or is it general? Table III furnishes 
evidence on this point. It appears that there is a reduction in the 
speed of the total phase of shortening and total phase of relaxation 
and also of each period of .o5 second throughout the entire con- 
traction. 

It may be noted that in the normal and fatigued curve, the shorten- 
ing is at first slow, then faster, then slower as the end of the phase is 
reached. At the beginning of relaxation, the movement is slow and 
usually, under the above conditions, gains in speed throughout the 
rest of its extent. These relations are of course largely due to 
mechanical conditions. 

Summary: — In these experiments fatigue produced: a less exten- 
sive contraction; an increase in the time required for the total con- 
traction; a decrease in speed during the phase of shortening; a 
decrease in speed during the phase of relaxation; a decrease in 
speed during each five hundredths of a second throughout the phases 
of shortening and relaxation. 


304 Thomas Andrew Storey. 


TABLE I. 


HEIGHT AND DURATION OF NORMAL AND FATIGUED VOLUNTARY CONTRACTION. 


Height of Duration in ;35 | Duration in ;3, | Duration in 74> 
contraction in | seconds of total | seconds of phase | seconds of phase 
Number of millimetres. contraction. of shortening. of relaxation. 
contraction. 


Fa- 
tigued., 


Fa- 
tigued. 


Fa- 
tigued. 


Fa- 


Normal. tigued. 


Normal. Normal. Normal. 


June 6, 1900 
9 A.M. 


1 and 101 


2 and 102 


3 and 103 


June 6, 1900 
2.05 P.M. 


1 and 101 
2 and 102 
3 and 103 


June 6, 1900 
4.25 P.M. 


1 and 101 


2 and 102 
3 and 103 
June 7, 1900 
9.10 A.M. 
1 and 101 


2 and 102 


3 and 103 


June 7, 1900 | 
2.05 P.M. 


1 and 10) 


2 and 102 


3 and 103 


Lifluence of Fatigue on Contraction of Human Muscle. 365 


TABLE II. 


SPEED IN MILLIMETRES PER SECOND OF TOTAL CONTRACTION, PHASE OF SHORT- 
ENING AND PHASE OF RELAXATION IN NORMAL AND FATIGUED VOLUNTARY 
CONTRACTION. 


Total contraction. Shortening. Relaxation. 


Number of 
curve. 


Normal. | Fatigued. | Normal. | Fatigued. | Normal. | Fatigued. 


June 6, 1900 
9 A.M. 


land 101 
2 and 102 
3 and 103 


June 6, 1900 
2.05 P. M. 


1 and 101 
2 and 102 
3 and 103 


June 6, 1900 
4.25 P.M. 


1 and 101 


2 and 102 
3 and 103 


June 7, 1900 
9.10 A.M. 


1 and 101 
2 and 102 
3 and 103 


June 7, 1900 
2.05 P. M. 


1 and 101] 
2 and 102 
3 and 103 


366 Thomas Andrew Storey. 


TABLE III. 


SPEED IN CONSECUTIVE PERIODS OF FIVE HUNDREDTHS OF A SECOND THROUGHOUT 
THE VOLUNTARY CONTRACTION OF NORMAL AND FATIGUED MUSCLE. 


Speed in millimetres per second. 


Periods of | 
0.05 sec. 


No. 101. No. 2. No. 102. No. 3. 


June 6, 1900 
9.00 A.M. 


Shortening. 


Relaxation. 


on 
of 
= 
(=) 
Oo 
= 
) 
S 
Nn 


Relaxation. 


June 6, 1900 | 
4.25 P.M. | 


Shortening. 
mw wp MN w~N LO 
wo +p 


= 


Relaxation. 


L[nfluence of Fatigue on Contraction of Human Muscle. 367 


TABLE II] — continued. 


: Speed in millimetres per second. 
Periods of 
0.05 sec. 


1 No: 101. pp GNer2. No. 102. 


June 7, 1900 
9.10 A. M. 


Shortening. 


= 
i} 
=s 
a 
z 
S 
a 
o 
me 


Shortening. 


Relaxation. 


RELATION BETWEEN HEIGHT AND SPEED OF CONTRACTION. 


In the unfatigued condition the speed of shortening is greater in 
the higher contractions than in the lower, but there are many excep- 
tions to the rule. In the case of the fatigued contractions this rela- 
tion between height of contraction and speed of shortening is much 
closer, as appears in Table IV. 


368 Thomas Andrew Storey. 


TABLE, LV 


THE SPEED OF SHORTENING COMPARED WITH THE HEIGHT OF VOLUNTARY CON- 
TRACTION IN THE NORMAL AND FATIGUED CONDITION. 


Normal. Fatigued. 


Height of con- | Speed of short- || Height of con- | Speed of short- 
tractions ening in mm. tractions ening in mm. 
in mm. per sec. in mm. per sec. 


16 
yA 


The same statements may be made concerning the relation between 


the height of contractions and the speed in the phase of relaxation 
(Table V). 


Lnfluence of Fatigue on Contraction of Human Muscle. 369 
TABLE V. 


THE SPEED OF RELAXATION COMPARED WITH THE HEIGHT OF VOLUNTARY Con- 
TRACTION IN THE NORMAL AND FATIGUED CONDITION. 


Normal muscle. Fatigued muscle. 


: S ax- : Spee relax- 
Height of con- Speed oF relax Height of con- peed of relax 
ation In mm. = 


aes y ta Pav ation in mm. 
traction 1n mm. traction in mm. pase 
per sec. per sec. 


Why the height and the speed of contraction should be closely related 
in the fatigued and should differ widely in the fresh muscle is not clear. 


Tue HEIGHT AND SPEED OF’ UNWEIGHTED VOLUNTARY CONTRAC— 
TIONS MADE BEFORE AND AFTER A SERIES OF WEIGHTED 
VOLUNTARY CONTRACTIONS PRODUCING FATIGUE. 


In these experiments the muscles were made to contract volun- 
tarily six or eight times, with no weight to overcome, except the 
resistance of the levers and recording apparatus, which was about 
185 grams. Each contraction was willed to be one of maximal 


370 Thomas Andrew Storey. 


velocity and extent. Then the muscle was contracted voluntarily 
against a weight sufficient to produce exhaustion (Zz. e. inability to 
work longer with that weight and at that rate). The amount of 
weight was varied in different experiments. As soon as the muscle 
was unable to lift the weight, the weight was removed and six or 
eight voluntary contractions were made under the same conditions as 
in the first unweighted voluntary contractions. The whole series of 
contractions was recorded on a slow moving drum -by means of a 
+-shaped device. The horizontal arm was used for a writing-point. 
The upper vertical arm was suspended by a light rubber band, and 
the lower vertical arm was connected by means of a thread with the 
pulley described on page 358. By twisting the rubber band, the 
writing-point was made to rest against the recording surface. 

The voluntary unweighted contractions were recorded also on a 
fast moving belt of tracing paper which was about twelve feet long. 
The slow moving drum gave a picture of the whole experiment 
(Figs. 6 and 7); the fast moving belt gave a picture of the velocity of 
the movement at the desired time. 

A record of a tuning-fork which vibrated one hundred times a 
second was made on the fast moving tracings; on the slow drum the 
time-record was in seconds. 

An examination of Table VI, in which the results of the above 
experiments are tabulated, reveals the fact that the unweighted con- 
tractions after the weighted series producing fatigue, are in many 
cases as high as the unweighted contractions produced before fatigue ; 
and that the speed is often as great in one case as the other (see Figs. 
6 and 7), 


Tue HEIGHT AND SPEED OF UNWEIGHTED CONTRACTIONS PRO- 
DUCED BY THE MAKE SHOCK OF THE INDUCED CURRENT 
BEFORE AND AFTER A SERIES OF VOLUNTARY CONTRACTIONS 
AGAINST A WEIGHT, THE VOLUNTARY CONTRACTIONS BEING 
CoNTINUED UNTIL THE MUSCLE WAS UNABLE TO LIFT THE 
WEIGHT BECAUSE OF FATIGUE. 


In these experiments the electric current was made and broken 
automatically (page 360). Its strength was tested before and after 
each experiment and every effort was made to maintain the constancy 
of the conditions under which the experiment was performed. The 
current was generated by a series of seven Daniel cells. The induc- 


Influence of Fatigue on Contraction of Human Muscle. 371 


TABLE VI. 


THE HEIGHT AND SPEED OF UNWEIGHTED VOLUNTARY CONTRACTIONS MADE 
BEFORE AND AFTER A SERIES OF WEIGHTED VOLUNTARY CONTRACTIONS 
PRODUCING FATIGUE. 


Time consumed in Velocity in millimetres 
Height of xia seconds. per second. 
No. of contraction 
contrac- in mm. 
tions. Shortening. Relaxation. | Shortening. | Relaxation. 


Before.| After. |Before.| After. Before. After. Before.| After. |Before.| After. 


0.14 | 
0.15 
0.15 
0.15 


XS: (00) =f Oy). On = 10" (tS) 


— 
oO 


July 20, 1900 
11.30 A.M. 


| aly, : 0.14 


0.17 0.19 
0.21 0.21 
| 0.18 | 0.21 
| 0.18 0.17 
Reet ye 0.19 
OOo |) Tae 016 


— 
otnnNtwnelt 


1 Time record injured. 


272 Thomas Andrew Storey. 


tion coil was one of 13,000 windings. A record of the whole experi- 
ment was made on a slow moving drum, and examples taken from 
time to time during the experiment on a long belt of tracing paper 
carried on double drums and moving rapidly. This belt was about 
twelve feet long. In this way two records were made of each experi- 
ment, one showing all its details; the other showing the speed of 
contraction at different times during the experiment, as in the preced- 
ing experiments. 

An automatic device registered on the long sheet the time when 
the muscle was stimulated electrically, thus affording a rough basis for 
estimating the latent period. 

The electrodes were arranged as follows: The anode was a flat 
brass disc covered with chamois skin moistened with a weak solution 
of common salt in distilled water. This electrode rested against the 
palm of the hand. The kathode was about the size of a large thimble 
and was placed directly upon the abductor muscle. It was held in 
place by means of a rubber band. Its position was about the same 
in all experiments, and especial care was taken to prevent any changes 
during each single experiment. 

It will be observed, on looking over the tracings reproduced in this 
paper, that the make-induced shock was used in most cases where 
electricity was employed as a stimulus. This was done because the 
strength of make shock used here appeared to be more efficient than 
that of the break, not only when used to excite the human abductor 
indicis muscle, but also when used to excite the excised gastrocnemius 
of the frog. This fact was noted, but the attempt to find an explana- 
tion has been reserved for a later time. 

The rhythm of contraction was made constant in the like parts of 
the same experiment. The variation in rhythm in the different 
experiments may be detected by referring to the time-records in the 
tracings given. 

Table VII (page 374) demonstrates that the height and speed of the 
unweighted contractions, produced electrically, are reduced during the 
period of fatiguing voluntary contractions against the weight. For 
example, it may be noted in the record of the first contraction before 
fatigue, and in the record of the first contraction after fatigue, that 
there was a fall in the amplitude of movement from 52.5 mm. to 
37-9 mm.; that the time spent in executing the phase of contraction 
increased from 0.14 to 0.15 second; and that in the execution of the 
phase of relaxation there was an increase in time consumed from 0.15 


Influence of Fatigue on Contraction of Human Muscle. 373 


second to 0.18 second. It maybe seen further that the speed of 
contraction recorded in millimetres per second, decreased in the phase 
of shortening from 388.8 mm. before fatigue to 212.6 mm. after fatigue. 
The decrease in speed in the phase of relaxation was from 354.7 mm. 
per second to 210.7 mm. per second.! 

Then, the height of contraction is reduced, and the phases of short- 
ening and relaxation are accomplished at a slower speed. 


CENTRAL AND PERIPHERAL FATIGUE 


A comparison of the conclusions drawn from Tables VI and VII 
brings out the following facts bearing upon the seat of fatigue in 
voluntary muscular contraction. 

Voluntary stimulation with no weight, before and after voluntary 
fatigue with a weight, gives no good evidence of fatigue (Table VI). 
The neuro-muscular machine may be subjected to a set of conditions 
which seem to exhaust it so that it loses its ability to act in response 
to the will. But there is apparently enough power left to accomplish 
under more advantageous conditions a normal amplitude and speed of 
shortening. Then it would appear that either the central or periphe- 
ral mechanisms or both, are fatigued for a given weight, but are still 
normally active for a lighter weight. 


1 It may be noted in passing that staircase contractions appeared in every series 
of electrically excited contractions produced in this research (Figs. 6 and 7). 

2 A reference to the following papers cited will show the diversity of opinion 
existing concerning the seat of fatigue in voluntary muscular contraction. FERE: 
Journal de l’anatomie et de la physiologie, 1901, xxxviil, p. 14. FRANZ: This 
journal, 1900, iv, p. 348. HARLEY: Journal of physiology, 1894, xvi, p. 97. 
HouGu: This journal, 1go!, v, p. 239. JOTEYKO: Travaux de I’Institut Solvay, 
1900, iii, p. 47. KRAEPELIN: Neue Heidelbergische Jahrbiicher, 1896, vi, 2, 
p- 222. LomBarD: Journal of physiology, 1893, xiv, p. 97; 1892, xiii, p. 6; 
American journal of psychology, 18go, iii, p. 24. MAGGIORA: Archives italiennes 
de biologie, 1898, xxix, p. 267. Mosso: Die Ermiidung, Leipzig, 1892; Archiv 
fiir Physiologie, 1890, p. 129; Archives italiennes de biologie, 1890, xiii, p. 123. 
MULLER, G. C.: Zeitschrift fiir Psychologie und Physiologie der Sinnesorgane, 
1893, iv, p- 122. MULLER, R.: Philosophische Studien, 1901, xvii, p. 1. 
OSERETZKOWSKY and KRAEPELIN: Psychologische Arbeiten, 1898, iii, p. 587. 
SCHENCK: Archiv fiir die gesammte Physiologie, 1890, ]xxxii, p. 393. TREVES: 
Archives italiennes de biologie, 1898, xxx, p. 1. WALLER: Brain, 1891, liv, 
p. 174. WoopwortH: The New York University Bulletin of the Medical 
Sciences, 1901, i, No. 3, p. 133. 


Thomas Andrew Storey. 


+ 
IC™ 
on 


f61E 
£602 
£S61 
Sst 
T'S0l 


S'tlt 
L'L0b 
0'S0F 
FHSe 


9°99¢ 
CLV 
O'0FZ 
Opole $'06€ 


ICIS iSitststa 


“ON DIVLF 
IV 


“ONSIeF 
a10Jaq 


‘anSyLy 


‘puosas sad ‘uu ul 
uolexe[at Jo poads 


‘onstyey 
a10jJoq 


I0}FV 


‘ONSYRI 
ONY 


“ANST}LF 
a10joq 


106 ‘Oz Ainf 


9 


1061 ‘Oz 41m 


‘ONSLF 
IdyV 


‘ONSI}VF 
a10FO 


‘onsyey 
TFV 


‘ansi}ey 
a10jaq 


‘puosas sod ‘uu ut 
Surusj10oys Jo psads 


*spuooas 
UL WOTJEXPOY 


*spuosas 
OOT ; : 
Tr Ul SuruazI0YS 


“tau UL 
UOTJIVIJUOD JO JYSIAT] 


“ANOILVA OL ONIGVAT SNOILOVULNOD AUVINATOA GILHSIT A 


4O SHINAS V YAIAV GNV AXOAAH ACVW NOILVIONWILS TVOINLOATY NOUN SNOILOVYLNOD GALHSIAMNY, AO GAAdS ANV LHSIGH 


TTAY ATV 


Liflucnce of Fatigue on Contraction of Human Muscle. 375 


On the other hand, it has been shown in Table VII that electrical 
stimulation with no weight, before and after voluntary fatigue with 
a weight, gives unmistakable evidence of fatigue. 

If these experiments are not at fault, one is justified in stating that, 
under these conditions, peripheral fatigue is apparent and that central 
fatigue, though possibly present, is not apparent (Figs. 6 and 7). 


SUMMARY OF RESULTs. 


This paper has reported : — 

(1) The invention of a new form of ergograph employing a simple 
muscle easily excited to contraction by the single-induced shock. 

(2) Evidence that fatigue produces a less extensive contraction ; 
an increase in the time required for the total contraction; a decrease 
in speed during the phase of shortening; a decrease in speed during 
the phase of relaxation; and a decrease in speed during each five 
one-hundredth seconds throughout the phases of shortening and 
relaxation. 

(3) In the unfatigued condition the speed of shortening is greater 
in the higher contractions than in the lower, but there are many 
exceptions to this rule. This relation is much closer in the fatigued 
contractions. The same statement may be made concerning the 
relation between the height of contraction and the speed of relaxation. 

(4) Unweighted (voluntary) contractions after a weighted series 
producing fatigue are in many cases as high as the unweighted 
(voluntary) contractions produced before fatigue, and the speed is 
often as great in the one case as in the other. 

(5) The height and speed of unweighted contractions produced 
electrically are reduced during a period of fatiguing voluntary con- 
traction against a weight. 

Under the experimental conditions outlined in this paper, peripheral 
fatigue was apparent, and central fatigue was not apparent. 


In conclusion, the writer wishes to express his obligation to Dr. 
Warren P. Lombard for his generous expenditure of time and advice 
upon this research. 


Pea YoOLOLOGICAL STUDY OF NUCLEIC ACID. 


By LAFAYETTE B. MENDEL, FRANK P. UNDERHILL, ANpD 
BENJAMIN. WHITE. 
[From the Sheffield Laboratory of Physiological Chemistry, Yale University.] 


CONTENTS. 


Page 

EDS Pree, Seat Y, acre ec gteh OF Lk We Pogue Pat i) ez elo) ey WARE 
LETS STAD, 2 Ue Sarg tee Met chs 0a rig SLA man ery | ok Oe, a PTY Uriel ge os 
PBI tyLOvPHe MUCICIC ACIGS: F. = 3) friicius, Jotes crdust om ss). eee eo nae FOS 

EB ArAMONCOfMUCIEICAGCIG. win eeirakce tod fs a, tas ude 2 § ve ees an On 
Pe eA GHODOLMUCleiG: ACI...) cus) 3 8 a ol ek Sw es eas eee wy OOO 
BEMetae x CUI CtS ay be) eee oman Cee ouddys tea? ts to's Nl & am oe te OOO 
PDICCEH Pex DCUMC MCS miei? Gl ii oth torte RA cyto auch suck. fo oe las Bays a EOSO 

PTE CIS ON DIOGU=PrESSULE ecu (ac) oy. set ylcs SAR GeE S) cot SE hen whew 3s oy ee SO 
BEEECESiOn Dl ood=coapulation ay.) % Pe eit a fof Os sf ee dees! | ODL 
BALERS One VIN PUOW sect wipe ic. tS) ie lon cathe waren ae ont tee dees Nin a! OB 
RCL Ven Ae COMET es el smc IRL. Qe iy LS s (a) Gel eo paeaes poo 
ati nice CeACIasin the: DOAY, 24 Til lees sek du el eg os, Wl, oe ees) YSOD 
RE EDMEU ISS Ui] CCHIOL Beech asa tye™ ge aay. at | tc oe one Tee a, ) ost euesae ae eer OOD, 
HapeEMCOnedieInjECtiON: 11s col as8 c/o ele) a) is Se a east eee) eae LOO 

PMP EATeC OUST CCUIOI: wari ta, Seale) tsp eros! oc!) he oa tonne Mes Bums aOR 
PRGEICELPCTICIEOIS ON ARNMALS© 2.0 cps fs, cw ude |e! M. tee ake ee eh ee gee 
PPGOLAIBICC CUO ery Lie So coe ya. est we Moke as) LS gt, ee en 

SP CPBEIMC TIES OMPANAD gt Ware Rpinet es foe | on Ms aoe fo ei aek ow 4.8 ian al fo) pee ae ese TOD 
SUPRA 2g. RAAGL sei SR ine Te re aR rain ER tr coe 30) 


INTRODUCTORY. 


HE successive changes of opinion regarding the origin and 

elimination of uric acid in the animal body form an interesting 
chapter in the history of physiological research. The abandonment 
of the earlier theory which connected the formation of uric acid 
directly with the decomposition products of the proteids was due to 
the failure of experimental proofs. In its place has arisen a series 
of hypotheses which have stimulated investigators to undertake new 
lines of research and have given direction to the modern teaching 
regarding the metabolism of the so-called “ nucleins.” 

So long as ordinary meat diets were used in the experimental 
studies, it was not difficult to show an apparent relationship between 
nitrogenous metabolism and uric acid production. This fact led to 

377 


378 L. B. Mendel, F. P. Underhill, and B. White. 


the attempt to establish a ‘‘normal ratio” between urea and uric 
acid output in man. But it was soon found that wide variations in 
this ratio might occur in healthy individuals. Thus with certain diets 
the uric acid output remained noticeably low despite the normally 
high execretion of urea. This was particularly evident with dietaries 
in which milk and vegetable foods predominated. On the other 
hand the results of the feeding experiments of Weintraud!and others 
with thymus again directed attention to the peculiar significance of 
the nucleins in the production of uric acid. 

Early in the course of his study of the nucleins of cells Kossel ? 
directed attention to the possible genetic relationship between these 
compounds and the uric acid output. Subsequent investigation has 
tended to verify this assumption. Stadthagen? and Gumlich* both 
failed to obtain any experimental evidence of uric acid production 
from isolated nuclein derivatives. The former fed nuclein obtained 
from yeast to a dog ; and Gumlich likewise fed 22 grams of nucleic 
acid prepared from thymus (and containing about 10 per cent of 
phosphorus) with equally negative results. These experiments have, 
however, been criticised in respect to the methods employed? Hor- 
baczewski ® was the first to demonstrate that the ingestion of nuclein 
(obtained from the spleen) gave rise to an increased uric acid elimina- 
tion in man and in the rabbit; but he did not attribute this increase 
directly to the ingested nuclein. According to Horbaczewski the 
formation of uric acid is dependent on the decomposition of leucocytes, 
the latter thus furnishing the real antecedents of the uric acid. 
Leucocytosis and cellular decomposition thus become essential con- 
ditions for uric acid production ; and according to the author, ingested 
nuclein is efficient mainly because of the digestive leucocytosis to 
which it gives rise. 

Despite the fact that experimental aidente in apparent support of 
the leucocytosis hypothesis has not been lacking,’ it can no longer be 
regarded as the only tenable one. Demonstration of increased uric 

' WEINTRAUD: Berliner klinische Wochenschrift, 1895, p. 405; Archiv fiir 
Physiologie, 1895, p. 382. 

* KOssEL: Zeitschrift fiir physiologische Chemie, 1882, vii, p. 19. 


* STADTHAGEN: Archiv fiir pathologische Anatomie, 1887, cix, p. 390. 
* GUMLICH: Zeitschrift fiir physiologische Chemie, 1893, xviii, p. 508. 


5 Cf. SCHREIBER: Die Harnsidure, Stuttgart, 1899, p. So. 
® HORBACZEWSKI: Monatshefte fiir Chemie, 1891, xii, p. 221. 
" Cf for example, KUEHNAU: Zeitschrift fiir klinische Medicin, 1895, xxviii, 


P- 534- 


A Physiological Study of Nucleic Acid. 379 


acid output unaccompanied by leucocyte changes has repeatedly been 
afforded in recent years; ! and reversely, marked leucocytosis may 
occur quite independent of any change in uric acid execretion. The 
parallelism between the two factors is by no means constant. Two 
striking illustrations in support of this statement may suffice. 
Weintraud? found only a very slight increase in the number of 
leucocytes in his thymus-feeding experiments, although the excretion 
of uric acid was unusually large. On the other hand Milroy and 
Malcolm * (and likewise Henderson and Edwards?) have studied 
cases of lymphatic leukemia in which an enormous leucocytosis was 
unaccompanied by any corresponding increase in uric acid output. 
An essential relation between the two is not apparent; and _ further- 
more it is not unlikely that at times increase in the number of 
leucocytes may not be accompanied by any simultaneous destruction 
of these elements. 

It will be seen that an increased production of uric acid from dis- 
integrated leucocytes (as it occurs in some forms of leukzemia and 
after the use of certain drugs) is in no way precluded.® In this case 
we may look to the liberated nuclein compounds of these cells or to 
their decomposition products for the antecedents of the uric acid or 
other excreted purin compounds. We shall attempt to demonstrate 
further that similar materials introduced into the organism are sub- 
ject to analogous metabolic changes. The purin bodies which arise 
from the nuclein may accordingly be regarded as intermediary pro- 
ducts of metabolism. One part of these alloxuric compounds is 
perhaps excreted as such, while another part is first oxidized further 
and then eliminated. A rather extensive literature has already 
arisen in support of this idea, as the outcome of numerous experi- 
ments of varied character. Most of these have been concordant in 
indicating that the ingestion of nuclein-containing tissues in man 


1 Cf. RicuTER: Zeitschrift fiir klinische Medicin, 1895, xxvii, p. 290; Hop- 
Kins and Hope: Journal of physiology, 1898, xxiii, p. 285; MILroy and Mat- 
coLm: J/ézd., 1898, xxiii, p. 217; SCHREIBER: Die Harnsaure, 1899; BURIAN 
and Scour: Archiv fiir die gesammte Physiologie, 1900, Ixxx, p. 258. 

2 WEINTRAUD: Berliner klinische Wochenschrift, 1895, p. 405. 

8 MiLtRoy and MALCOLM: Journal of physiology, 1898, xxiii, p. 217; 1899, 
<xv, ph Ios. 

4 HENDERSON and Epwarps: This journal, 1902, vi, p. xxii. 

5 Cf. Loewi: Archiv fiir experimentelle Pathologie und Pharmakologie, 1900, 
xliv, p. 2, where the literature is reviewed. 

6 Cf. BurIAN and Scuur: Archiv fiir die gesammte Physiologie, 1900, Ixxx, 


p- 241. 


380 «6oL. B. Mendel, F. P. Underhill, and B. White. 


and animals is followed by an increased output of uric acid and other 
purin compounds varying in quantity with the relative richness of 
the tissue in the nuclein precursor.!. In the course of these earlier 
investigations it was discovered independently by Minkowski ? and 
by Cohn ® that allantoin may be excreted in the dog as an end-product 
in the metabolic processes which give rise to uric acid elimination 
after thymus-feeding. Subsequently Salkowski* obtained allantoin 
after pancreas-feeding ; and Mendel and Brown® demonstrated that 
allantoin excretion can be brought about in the cat also, after inges- 
tion of pancreas, thymus, or lymphatic glands. In man, allantoin has 
not been detected under similar conditions.® It has, however, been 
reported to occur in the urine of pregnant women and of newly 
born infants; and in view of the unsatisfactory methods available 
for its separation and identification it is not improbable that traces 
present have repeatedly escaped detection. The significance of the 
presence of allantoin in the urine of animals fed on a diet rich in nuclein 
constituents is more apparent when we recall the older observations of 
Salkowski‘ that the feeding of uric acid to dogs is followed by an 
excretion of allantoin. Since the latter is directly obtainable in the 
laboratory from uric acid by oxidation, a genetic relationship between 
the two compounds in the organism is at once suggested. Podushka® 


1 Cf. WEINTRAUD: Berliner klinische Wochenschrift, 1895, No. 19, p. 405; 
KUEHNAU: Zeitschrift fiir klinische Medicin, 1895, xxviii, p. 561; ROSENFELD 
and ORGLER: Centralblatt fiir innere Medicin, 1896, p. 42; Hess and SCHMOLL : 
Archiv fiir experimentelle Pathologie und Pharmakologie, 1896, xxxvii, p. 243; 
UMBER: Zeitschrift fiir klinische Medicin, 1896, xxix, p. 174; MAYER: Deutsche 
medicinische Wochenschrift, 1896, p. 186; LUETHJE: Zeitschrift fur klinische 
Medicin, 1897, xxxi, p. 112; WeEIss: Zeitschrift fiir physiologische Chemie, 1899, 
xxvil, p. 216; HopKINS and Hope: Journal of physiology, 1898, xxiii, p. 271; 
JEROME: /did., 1899, xxv, p. 98; TAYLOR: American journal of the medical 
sciences, 1899, cvili, p. 50; MENDEL and Brown: This journal, 1900, iii, p. 261; 
MENDEL and JACKSON: This journal, 1900, iv, p. 163; and others. 

* MINKOWSKI: Centralblatt fiir innere Medicin, 1898, No. 19; Archiv ftir 
experimentelle Pathologie und Pharmakologie, 1898, xli, p. 376. 

® COHN: Zeitschrift fiir physiologische Chemie, 1898, xxv, p. 507- 

* SALKOWSKI: Centralblatt fiir die medicinische Wissenschaften, 1898, p. 929. 

® MENDEL and Brown: This journal, 1900, iii, p. 261. 

® Coun: Loe. cit., p. 509; MINKOWSKI: Archiv fiir experimentelle Pathologie 
und Pharmakologie, 1898, xli, p. 398; Lorw1: /déd., 1900, xliv, p. 22. 

7 SALKOWSKI: Berichte der deutschen chemischen Gesellschaft, 1876, ix, 
p- 719; Zeitschrift fir physiologische Chemie, 1902, xxxv, p. 495. 

* PopuscHKA: Archiv fiir experimentelle Pathologie und Pharmakologie, 
1900, xliv,’ p. 65. 


A Physiological Study of Nuclere Acid. 381 


o 


failed to obtain any allantoin after feeding uric acid; but Salkowski’s 
observations have been verified by Minkowski! and, in this laboratory, 
by Swain; while Mendel and Brown ®* have demonstrated that a 
similar oxidation may occur in the cat. Finally, it must be added 
that Minkowski obtained allantoin to the extent of 77 per cent after 
feeding hypoxanthin to adog.t Experimental evidence thus indicates 
that purin bases (and nucleins) may give rise to allantoin as an inter- 
mediate product in their oxidation. Ordinarily, where the conditions 
are favorable, these compounds are, as Swain has suggested, more 
completely oxidized in the system, and their nitrogen presum- 
ably reappears in large part as urea.” This might be represented as 
the ordinary fate of uric acid, for example, when introduced into 
the Auman body in such doses as are permissible. Furthermore, 
allantoin itself may be oxidized almost completely in the body.® In 
the dog, also, comparable quantities of uric acid may apparently be 
metabolized beyond the stage where allantoin appears as an end- 
product. But when larger quantities (per kilo of body weight) 
are fed, the system is evidently unable to bring about so. complete a 
decomposition of the purin radical; and under these conditions 
allantoin escapes as an end-product of the change. The difference 
between man and other animals may thus be due, in part at least, to 
quantitative variations in the extent of metabolism in the involved 
organs, such as the liver, in addition to any specific peculiarities of 
different animals.’ 

The investigations of recent years have thus indicated clearly that 
the purin (or alloxuric) compounds of the urine owe their origin 
almost entirely to antecedent purin groups introduced into the body 
as food or otherwise. A small quantity of purin compounds, largely 
in the form of uric acid, is excreted even on a “ purin-free”’ diet. 
This quantity, which Burian and Schur have designated as endogenous 
urinary purin constituents apparently varies somewhat with different 


1 MINKOWSKI: Loc. cit., xli, p. 398 (foot-note 3). 

2 SwaIN: This journal, 1gor, vi, p. 38. 

3 MENDEL and Brown: ‘his journal, 1900, iii, p. 267.! 

4 MINKOWSKI: Loe. cit., p. 404. 

® Cf. BuRIAN and ScuuR: Archiv fiir die gesammte Physiologie, 1901, Ixxxvii, 
p- 239; SALKOWSKI: Zeitschrift fiir physiologische Chemie, 1902, xxxv, p. 495. 

6 Cf. MINKOWSKI: Loc. cit., p. 399; PopusCcHKA: Loc. cit., p. 64. 

‘ Cf. BuRIAN and ScuHur: Archiv fiir die gesammte Physiologie, rgor, 
Ixxxvil, pp. 261, 335; KAUFMANN and Moar: Deutsches Archiv fiir klinische 
Medicin, 1902, Ixxiv, p. 141. 


382 L.B. Mendel, F. P. Underhill, and B. White. 


individuals, although it tends to remain remarkably constant for the 
same person, even where extensive changes in the nitrogenous ele- 
ments of the diet are introduced. In general it varies in man 
between 0.1 and 0.2 gram of nitrogen per day. This represents the 
end-products of the metabolism of purin complexes of the body cells, 
and can be ascertained by estimating the purin compounds (chiefly 
uric acid) of the urine on a “ purin-free” diet of milk, eggs, cheese, 
potatoes, rice, etc. The additional output of purin derivatives which 
follows a diet containing the antecedents already referred to earlier, 
may be termed the exogenous portion. 

Problems. — \Vhat are the immediate antecedents of the erogenous 
purin compounds? What changes do the so-called nucleins undergo 
in metabolism? Do the nuclein compounds from various sources 
exhibit similar transformations in the organism; that is, do they 
possess the same physiological value? These and related questions 
can at present be answered in part only. It is the purpose of this 
paper to afford further experimental contributions to the modern 
theory of purin metabolism. Most investigators have employed 
glands (e.g. thymus) rich in “nucleoproteids” as the nuclein-con- 
taining substance. Experiments with isolated “ nucleoproteids,”’ 
nucleates,” or nucleic acids are few in number, and the 
results are by no means concordant. Still fewer experiments have 
been tried with the purin derivatives: hypoxanthin, xanthin, guanin, 
and adenin, obtainable from the above.! The failure of Stadthagen? 
to obtain allantoin after guanin-feeding, and the almost similar result 
of Minkowski? with adenin, contrasted strongly with the marked 
allantoin output which follows ingestion of pancreas or thymus. In 
view of the fact that the nucleoproteid of the pancreas yields guanin 
(2-amino-6-oxypurin) almost exclusively as its purin constituent, and 


> 66 


‘“nucleins, 


that the thymus affords a preponderance of adenin (6-amino-purin ) 
on decomposition, it might be assumed that the amino-purin deriva- 
tives are not readily transformed in metabolism, or that the transfor- 
mation is dependent on a peculiar organic group or complex in which 
these derivatives exist in the tissue cells. Krueger and Schmid #4 


' The valuable contribution of KRUEGER and ScuMmib: Zeitschrift fiir physi- 
ologische Chemie, 1902, xxxiv, p. 549, on the behavior of the free purin bases, 
appeared after our experiments were undertaken. 

* STADTHAGEN: Archiv fiir pathologische Anatomie, 1887, cix, p. 418. 


* MINKOWSKI: Loc. cit., pp. 401, 406. 


* KRUEGER and Scumip;: Zeitschrift fiir physiologische Chemie, 1902, xxxiv, 


p. 549 


A Physiological Study of Nuclete Acid. 383 


have, however, quite recently been able to demonstrate the direct 
transformation of the four common purin bases: xanthin, hypoxan- 
thin, adenin, and guanin, into uric acid in man. 

Chemistry of the nucleic acids.— The preceding remarks with 
reference to the pancreas and thymus will serve to emphasize the 
differences which are now recognized to exist in the chemical struc- 
ture of the nuclein bodies. Whether the various related compounds 
separated from the tissues occur as actual constituents of the latter, 
or are merely the product of the manipulations employed, cannot be 
definitely asserted. Common to all of these products, however, is an 
acid constituent, the zacletc acid, which unites with the proteids to 
form proteid compounds, the xwac/eins or nucleoprotetds, the latter 
designation usually being reserved for those compounds which contain 
a smaller proportion of the nucleic acid radical. The nucleic acids 
may also be obtained combined with the protamins and free from 
ordinary proteids. The careful studies of Osborne! have led him to 
the conclusion “that the true nucleic acids are strong polybasic acids, 
containing the purin, pyrimidin, and carbohydrate groups, and yield 
on hydrolysis orthophosphoric acid; that there is at least one other 
acid which contains the purin and carbohydrate group and also yields 
orthophosphoric acid, but is a substance of a different order, since it 
contains glycerin and lacks the pyrimidin group; that there are at 
least two true nucleic acids, — one containing thymin, the other uracy] ; 
that the ultimate composition of these acids is not yet settled, though 
the more carefully purified preparations have a similar composition.” 

Up to the present time the thymus, spleen, pancreas, Arbacia-, 
salmon-, codfish-, and herring-milt have served as sources for the 
more carefully investigated nucleic acids. The nucleic acid of the 
wheat embryo (tritico-nucleic acid) is the only one thus far obtained 
from the higher orders of plants, although the yeast nucleic acid has 
been known for some time. We have used the wheat product 
largely in the present research for several reasons. First, it seemed 
desirable to compare the physiological behavior of the acid of 
vegetable origin with that of the other nucleic acids. Again, the 
wheat acid can readily be obtained in sufficient quantity and purity ; 
and finally its chemical composition has probably been ascertained 
more nearly than that of any other known nucleic acid. A pro- 


1 OSBORNE and Harris: Connecticut agricultural experiment station report 
for 1901, p. 387; cf. also Zeitschrift fur physiologische Chemie, 1902, xxxvi, 
p- 85. ‘ 


384. L. B. Mendel, F. P. Underhill, and B. White. 


o 


visional chemical structure has been assigned to it by Osborne,’ as 
follows : 


yOu 
C;H,O; — p— C;H,0; 
| SOH 
O 
HO. | 
i P —C,H,N.O, 
xX 
OH 
HO—p% 
| \C,H3N,0, 
CHaNg a 
at ge 
C;H,N; OH 


since it yields one molecule of adenin (C;H;N,) and of guanin 
(C,H;N,O), two molecules of uracyl (CyH,N,O,), and three mole- 
cules of a pentose (C,H,,O;). The wheat nucleic acid thus differs 
from the investigated animal acids in having the pyrimidin group repre- 
sented by uracyl in place of thymin (methyl uracyl). The pentose 
derived by Neuberg? from the guanylic acid of the pancreas is xylose. 
The thymus nucleic acid presumably yields a hexose. The wheat 
nucleic acid differs from Bang’s pancreas guanylic ® acid in containing 
no glycerin radical. The purest preparations of the nucleic acid of 
the wheat embryo obtained by Osborne are represented by the formula 
C,,H¢1Ni¢P,O3,, with which we have compared our preparations as a 
standard. 

Preparation of nucleic acid. — In preparing nucleic acid com- 
pounds from the wheat embryo, the suggestions of Levene?* and of 
Osborne® were applied. The material used was the “yellow germ 
meal” of commerce, care being taken to obtain a preparation as free 
as possible from endosperm. The meal consists of small flakes con- 
taining the embryo of the wheat kernel flattened into thin scales by 
the milling processes used. The yield of nucleic acid is variable ; 
and Osborne, who obtained 1.25 per cent, has pointed out that it is 


1 Cf. OsBORNE and HARRIS: Zeitschrift fiir physiologische Chemie, 1902, 
XXXVI, Pp. 120. 


* NEUBERG: Berichte der deutschen chemischen Gesellschaft, 1902, xxxv, 
p. 1467. 

* BANG: Zeitschrift fur physiologische Chemie, 1901, xxxi, p. 416. 

' LEVENE: Journal of the American chemical society, 1900, xxii, p. 329. 

® OSBORNE and CAMPBELL: Connecticut agricultural experiment station report 


for 1899, p- 305; Journal of the American chemical society, 1900, &xii, p. 379. 


A Physiological Study of Nuclete Acid. 385 


necessary to use very fresh meal, Since the yield decreases after a few 
weeks. The nucleic acid is present in the embryo in combination 
with proteids in the form of nucleates (nucleoproteids) from which 
it must be separated. Our purest preparations (as judged by their 
freedom from proteid) were obtained as follows: 


The wheat germ meal was twice extracted with seven times its weight of water, 
the supernatant liquids being siphoned off and strained through cloth. 
The united extracts were then saturated with sodium chloride, hydro- 
chloric acid being added until the precipitate separated out well. The 
bulky nucleate thus obtained contained a variable but large proportion of 
proteid as indicated by its low phosphorus content.’ It was digested 
for nearly a week with o.2 per cent hydrochloric acid containing an abun- 
dance of very active scale pepsin, with repeated renewal of the digestion 
medium. ‘The undissolved residue was repeatedly washed by decanta- 
tion and finally dissolved in water by careful addition of sodium hydroxide. 
The solution was alkaline to phenolphthalein. Dilute hydrochloric acid 
was then carefully added until a proteid precipitate began to separate, 
and this was filtered off. After the usual treatment with water, alcohol, 
ether, and drying at roo’, it was found to contain 1.15 per cent of phos- 
phorus.” The filtrate was treated with stronger hydrochloric acid which 
deposited a granular precipitate. ‘The later was treated with water, alcohol, 
and ether; and dried at 100°, it formed preparation H. The analysis 
was as follows: 


Preparation Osborne’s purest 
preparation 
Wey. i507 Perecemt 2 2. "e. 15.00 Delicene 
Ee ta CA ew, SP ears ret POET) 


A part of the nucleic acid precipitated by the hydrochloric acid was 
redissolved in sodium hydroxide, and the solution, after filtration, poured 
into a large volume of alcohol (95 per cent). ‘The fine cream-white pre- 
cipitate thus deposited was washed, dehydrated and dried in the usual 
way. An analysis of this sodzum nucleate, preparation K, gave 


Pieri. eet aee wel ee ee ai ee A peCe me 


1 Cf. OSBORNE and CAMPBELL : Loc. cit. 

2 The determinations of phosphorus in these compounds were all made by the 
usual gravimetric process after fusing the substances with pure sodium hydrate 
and potassium nitrate in a nickel crucible. Nitrogen estimations were made by 
the KJELDAHL-GUNNING method. 


386 L.B. Mendel, F. P. Underhill, and B. White. 


Another preparation, M, consisting of 31 grams obtained from about 20 kilos 
of meal, gave the following analysis : 


Pe Wor os) ue! Gept ae tare wore! | th, Jean, teu ree an aie ae Cola 
| Te a ea TPP ae ie cs WAL EY oe yy 


and was accordingly not quite as pure as the preceding products. 

Other preparations (nucleates) still containing some proteid in combination 
with the nucleic acid were also made. ‘The preliminary treatment was 
the same in each case, the proteid extracted from the meal being sub- 
jected to vigorous pepsin-acid digestion to eliminate a portion of the pro- 
teid. One residue which was found to contain 3.7 per cent of phosphorus 
after digestion for several days, was subjected to repeated digestion with 
pepsin-acid for three days at 38°. ‘This final product, nucleate C, now 
contained 


Pio Mes sl ile lle PE OR ee Be ee Gu em@enin 


and, assuming a phosphorus content of g per cent in the nucleic acid, 
was composed of the latter to the extent of about one-half. Anotber 
digestion residue, nucleate D, gave an analysis of 


Pes tn ets tap (En Ree eee 2.9 per cent, 


indicating a nucleic acid content of about one-third. Finally, one prep- 
aration was separated in part by Levene’s method. The digestion resi- 
due was dissolved in sodium hydroxide and then almost neutralized with 
acetic acid. Sufficient picric acid solution was then added to make the 
fluid slightly acid. The proteid-picrate precipitate, containing only a 
trace of phosphorus, was removed by filtration. To the filtrate strong 
hydrochloric acid was added to precipitate the nucleic acid, and then 
alcohol to facilitate its separation. The product, F, was only about two- 
thirds nucleic acid, since the analysis indicated 


Pe els) cate Ate ye? Ta bagi OA a conan 


In addition to these preparations we have used a commercial 
‘“ yeast nucleinic acid” which contained some admixture of proteid.! 


PHYSIOLOGICAL ACTION OF NucLeic ACID. 


Earlier experiments. — The physiological action of the nucleic acids 
and their derivatives after direct introduction into the circulation has 


' This was kindly furnished to us by PARKE, Davis and Co., of Detroit. It 


was reported to contain about 6 per cent of phosphorus. 


A Physiological Study of Nucleic Acid. 387 


received little attention up to the present time. Wooldridge! dis- 
covered some years ago that his so-called “ tissue-fibrinogens,” pre- 
pared from extracts of various organs and injected into the veins of 
a dog, produced thrombosis of the portal vein and its affluents; and 
he later ascertained that a degree of “immunity” toward a second 
injection could be brought about by a previous one.2 Wright? con- 
tinued these investigations. Halliburton‘ and his co-workers showed 
that the various tissue extracts which produce intra-vascular clotting 
contain nucleoalbumins, as indicated by their noticeable content 
of organic phosphorus and the formation of an insoluble phosphorus- 
containing residue by artificial gastric digestion. In rabbits a 
solution of the separated nucleoproteids usually produced death in 
a short time by respiratory failure. The thrombosis was ordinarily 
limited to the venous system, the portal vein sometimes being 
involved. Nucleoproteids prepared from the thymus, kidney, testis, 
liver, lymphatic glands, brain, and red marrow — containing from 0.4 
to 1.4 per cent of phosphorus — all showed this property. It was 
discovered that albino rabbits and the Arctic hare in its albino stage 
are immune to these effects.® Lilienfeld ® found that his “ nucleohis- 
ton,” prepared from leucocytes, likewise gave rise to intravascular 
clotting, and he concluded that the tendency toward production of 
thrombosis is attributable to the nuclein component of the compound 
injected. But it must be remembered in this connection that Martin? 
similarly obtained thrombosis with snake-venom which he found to 
be free from nucleoproteid; and Halliburton and Pickering * were 
able to induce intravascular coagulation in various animals by injec- 
tion of Grimaux’s proteid-like colloide amidobenzoique and colloide 
aspartique, each of which is free from phosphorus and in no way 
directly related to the nucleoproteids. Five to ten c.c. of 1.5 per cent 
solutions are fatal, while small doses of both nucleoproteids and the 
synthesized colloids may retard coagulation. Gioffredi® has found 
that a nuclein prepared from tubercle bacilli brings about death 


1 WooLpDRIDGE: Archiv fiir Physiologie, 1886, p. 397. 

2 WooLDRIDGE: J/d7d., 1888, p. 526. 

WriGurt: Journal of physiology, 1891, xii, p. 184. 

HALLIBURTON and BropieE: Journal of physiology, 1894, xvii, p. 135. 

5 PICKERING: Journal of physiology, 1896, xx, p. 310. 

6 LILIENFELD : Zeitschrift fiir physiologische Chemie, 1894, xx, p. 141. 

7 MARTIN: Journal of physiology, 1894, xv, p. 380. 

8 HALLIBURTON and PICKERING: Journal of physiology, 1895, xviii, p. 285. 
® GIOFFREDI: Jahresbericht fiir Thierchemie, 1goo, xxx, p. 1028. 


388 L. B. Mendel, F. P. Underhill, and B. White. 


speedily in rabbits when it is injected intravenously in doses of 
0.02-0.08 grams. In dogs also thrombosis is brought about by this 
preparation. 

The most important recent contribution to this subject is that of 
Bang,! who has studied the physiological action of Hammarsten’s 
pancreas nucleoproteid and the guanylic acid obtained from it. The 
latter has been extensively investigated from the chemical standpoint 
by Bang, who obtained as characteristic decomposition products : 
guanin, a pentose, glycerin, and phosphoric acid. It will thus be 
seen to differ essentially from the wheat nucleic acid used by us. 
The phenomena observed by Bang, after intravenous injection of 
pancreas guanylic acid into dogs, were characteristic, resembling in 
many ways the effects produced by protamin injections? and by 
larger doses of proteoses.*? The quantities injected varied from 0.02 
to 0.06 grams per kilo of body weight. In every case the injection 
immediately provoked a state of excitation in the animal, followed by 
transitory narcosis. The clotting of blood withdrawn from the vessels 
was greatly delayed, a dose of 0.04 grams per kilo prolonging the 
clotting time from the normal of six minutes to two hours. With 
0.018 gram per kilo the delay was only from a clotting time of ten 
minutes (before the injection) to nineteen minutes. The effects on 
respiration were also pronounced. Blood-pressure quickly fell after 
the injection, the pulse-waves of the records becoming smaller, while 
the heart-beat gradually became more forcible with a return of normal 
pressure. The fate of the injected guanylic acid was not ascertained. 
It was noted, however, that the urine always became distinctly alka- 
line after the injection, and albuminuria was observed. The results 
of the injection of the B-nucleoproteid of the pancreas were similar in 
nearly every respect. A dose of 0.05 grams per kilo was sufficient to 
abolish the coagulability of the blood in dogs completely, or almost 
so. In rabbits a noticeable retardation in blood-clotting was obtained 
with 0.14 gram of the pancreas nucleoproteid per kilo. The effect 
was far less marked than with dogs; and it may be recalled that 
intravenous injections of proteoses are practically ineffective in the 
case of the rabbit. In the dog, Bang noted a fall of arterial pressure 
after injection of the nucleoproteid. The urine did not show an 


' BANG: Zeitschrift fiir physiologische Chemie, 1g00, xxxi, p. 410; 1901, 
XXXII, p. 20I. 

* C/. THOMPSON: Zeitschrift fiir physiologische Chemie, 1899, xxix, p. 1. 

* Cf. CHITTENDEN, MENDEL, and HENDERSON: This journal, 1899, ii, p. 142. 


A Physiological Study of Nucleic Acid. 389 


alkaline reaction in this case, but contained dextrose (not pentose) 
in three of the five trials. 

Present experiments. — Our experiments on the physiological action 
of the wheat nucleic acid introduced directly into the circulation were 
carried out on dogs, cats, and a rabbit. Careful observations were 
made with reference to the effect on the coagulability of the blood, 
on blood-pressure, on lymph-flow, on “immunity,” and the fate of 
the injected acid. The dogs were anesthetized with A.C. E. mix- 
ture after receiving a hypodermic injection of 1 cgm. of morphin sul- 
phate and 1 mgm. atropin sulphate per kilo of body weight. The 


TABLE SHOWING EFFECTS UPON ARTERIAL PRESSURE. 


Pressure in milligrams of Hg. 


50min. after. | 


Experiment. 
in kilos. 
injected. 

Preparation 

employed. 
40 min. after. 


Grams per kilo 
tion in seconds. 


Weight of dog 


Just after 
injection. 


Duration of injec- 
injection. 


1 hour after. 


30min. after. 


Just before 
20 min. after. | 


5 min. after. 


10 min. after. 


_ 
Ww 
oO 


Pale. §2'|:0. 13 | H(8.43% P) 
| 13.0 | 0. H 
| 12.7 |0. H 

17.0 | 0. hee O4 


8.0 | 0.1: H | 140 | 
11.0 | 0.055 K (8.47% P) | 110 


| 


6.5 | 0.065 90 | K 


7.0 |0.056| 58 | F (604% P) | 


0.100 C (4.56% P) 


1 Slow injection. 
2 In this experiment a rabbit was used. 


substances were dissolved as sodium nucleates in a small volume 
(15-30 c.c.) and injected rapidly, 7. ¢., within a minute, from a burette 
into the jugular vein. Arterial pressure was recorded with a mercu- 
rial manometer in the carotid artery, and the blood samples were 
collected from a clean cannula introduced into a femoral vein, the 
coagulation time being noted as in previous experiments in the 


L. B. Mendel, F P. Underhill, and B. White. 


599 


‘opt tad “wi cot jo wonsalul “H uonriedatg “sop ZT Jo Sod “IA quowiiadx7] —"¢ ANAS 


NOLLITINI Lh Seal ir * aaron mila 


a go cas 


eee NS eT cae 


‘opey zed ‘wi cg‘Q Jo uonsaluy “H uoneiedaig ‘solr £ZT JO Sod “TITA quowiadxy —"Z AMADA 


— ll me “MIW 0 "MIA IT “IW £2 “wiv ef NIN oS 
NOMLIUNE Pannnount ee ee 


IMAM anna AWW 


‘opiy aod ‘wi $z0'9 Jo uonoaluy “W uonesredatg “sory s'ZT Jo Soq ‘XJ Juewuedx” —*[ aAYNSIY 


F 7 Talilalnkedalmlmlanb ashi) esite ine md eanon TT TITTtT 0 a ie 


as Ntw s¢ 


NOILITONE ASUS (INOLdAd) NOLLIMNI ONODTS 


> 


AAU MAMA) a a 


aii | NM Vall 


Ay Myr yh Ay ar a 1, VV iy 


A Physiological Study of Nucleic Acid. 391 


laboratory.!. Lymph was collected from the thoracic duct and ana- 
lyzed in the usual manner. Urine was withdrawn through a catheter 
when desired. 

Effects on blood-pressure. — The effects observed after the intraven- 
ous injection of solutions of nucleic acid on arterial pressure are 
summarized in the table on page 380. 

Three typical manometer tracings are reproduced in Figs. 1, 2, and 3 
from Experiments IX, VIII, and VI, which show the differences in the 
effects produced by varying doses. The curves were recorded from 
right to left. The line of zero pressure is marked below, and the 
time is recorded in seconds. 

From these protocols it will be observed that the nucleates ob- 
tained from the wheat embryo, when injected rapidly in doses larger 
than 0.04 gram per kilo of body weight, produce a fall in arterial 
pressure comparable with that obtained by Bang with the quite 
different guanylic acid of the pancreas, and somewhat resembling the 
effects observed after injection of larger doses of albumoses. So 
small a dose as 0.024 gram per kilo was practically without influ- 
ence; and with doses only slightly larger, the return of pressure to 
its previous height was rapid. The significance of the rate of injec- 
tion is indicated by the results of Experiment III in which the 
nucleic acid was introduced more slowly into the circulation. In the 
case of similar phenomena observed after injections of albumoses, it 
has already been pointed out that the intensity of the vaso-motor 
effects is probably not so much a function directly of the absolute 
quantity injected as of the quantity in the circulation at any time, or, 
in other words, of the concentration of the substance in the blood.’ 
A brief period of excitation was frequently noticeable, as in Bang’s 
experiments, immediately after the injections were made. 

Effects on blood-coagulation. — The action of the nucleic acid injec- 
tions on extravascular blood-coagulation are summarized in the 
appended table. It will be seen that doses of 0.05 gram per kilo 
tend, in the dog, to diminish the coagulability of the blood. In 
Experiment III the injection was, as already observed, somewhat 
slower than usual. The dog may have been naturally immune. To 
what extent the diversion of the lymph from the blood stream has 
modified the tendency of the blood to remain fluid after injections of 


1 CHITTENDEN, MENDEL, and HENDERSON: This journal, 1899, ii, p. 144. 
2 CHITTENDEN, MENDEL, and HENDERSON: This journal, 1899, ii, p. 150. 


*(¢) aunt SoqY = ‘uonsalur Mo[S 

*(sanoy Oz) YSU 1aA0 Sutpurys 19}Je paqjo[9 you pey sojduies ay} yey SoJOUsp (—) ysep Y ¢ 
‘SUOTPBAIISGO 9AIY} 10 OM] WOLF pze[N[wd ou] 9BvAAAY OY] AS¥D YOvE UT SI SIL z 

"c6¢ pure [6¢ ‘dd ‘syivuiar 9ag_ *syuautiadxa yduikq_ { 


All ©) 7 O 6 


(HG, (0) Z 0 | uonsalur s10jaq D ws O10 | Wades 


ss 0 


+ 


+e = FI 
OT Q | uonsalut a1ojag 


A “ws 9500 


0 
I 


Nw 


“dl ONT 
‘yoaful pz 


OT O | ; S aes 
ST uonsalut a10jaq H “ws $900 


mr} Ne | He 


N 


(0) 
ch 0 
Of + 


> 
uonoalur o10jacq 


M “Ws Ss0'0 


N 
|}oOor |OO}o0O;]coo0;}o° 


Cc 


Tal 
uoljoalut a10jaq 


‘yoalur pz 


I 
S 


nO 
wae 


H “ws sT 0 


[+o +m 
a 


+ 
uorjoalur a10jsag 


H ‘ws colo 


wy 


S 


Phe ud CQ’ 
uonoalur a1ojag H £0'0 


oe | 


j tan + 


£ nie : 
uonoalut a10Jaq H Wis tt0'O 


‘d on 
‘yoolut pz 


wy 


}CON | +r 


‘d 911M 
yoalut pz 
| “UL “" 


0 As 
QO  ,uoTIDAlUI s10Jaq 
“" 


H ‘Ws +70 0 


-mcn 


Heo 
Poo oS 


wn | 
st 
Jado loo HO|;|NoO!|cooO 


“QUIT]-3UI]JO[D JY SAATS OUT] IOMOT AY], 
‘uotoalur oy} 19378 pasdeya yor ‘soynurut ut ‘awit oy} Saas yuouttIadxea yove ur aur, saddn oy, 


‘SHINT.L-ONILLOTD 


‘opty rad ‘yudWI 
‘asoc] -10dx] 


cy ‘doo0Tqd AHL AO NOILVWTONSVOY AHL NO SNOLLVANASTQD AO AUVWWOS 


A Physiological Study of Nuclete Acid. 393 


nucleic acid cannot be answered. 


It is quite possible, for example, 


that the absence of more marked effects on blood-coagulation in 
Experiment IV or V is attributable to the failure of an anti-clotting 


Experiment. 


TABLE SHOWING EFFECTS UPON THE LYMPH. 


II 
Dog of 7 kilos 


Inject. of 0.056 gm. 


F per kilo 


| Before injection 

| After injection: 
lst portion 
2d portion 
3d portion 
4th portion 


V 
Dog of 11 kilos 


Inject. of 0.055 gm. | 


K per kilo 


Before injection 
After injection : 
lst portion 
2d portion 
3d portion 
4th portion 


Lymph-flow in 
periods of 
10 minutes. 

c.c. 


lymph. 


Per cent. 


Total solids 
Ash in 


in lymph. 
Per cent. 


1-3-1-5—1-5—1-4 


Remarks. 


Clots. 


Does not clot. 
Does not clot. 
Does not clot. 
Does not clot. 


“I 


ie 
Oo 


INI © 
S10) =) 
SOR 


VIII 


Dog of 12.7 kilos | 
Inject. of 0.05 gm. | 


H per kilo 


Before injection | 
After injection : 
Ist portion 
2d portion 
3d portion 
4th portion 


IV 
Dog of 13 kilos 


Inject. of 0.044 gm. | 


H per kilo 


Before injection | 
After injection : 
lst portion 
2d portion 
| After injection 
of ‘“ peptone ” 
(Witte) 


| 


13.0-18 0 


12.8412.3-(12.5-12 5) 
(11.5-11.5)-(9.0-9.0) 


(7.0-7.0) 


2.84.0 


4.0-3.0 
3.5-(3.0-3.0)—(4.6+4.6) 


10.2-7.8-6.0 


OY 
wn 


io) 


“ID IH 


Om Ore 
“ISI + Go 


Clots. 


Does not clot. 


| Does not clot 


Clots. 


Clots. 


| Clots. 


Does not clot. 
Does not clot. 
Does not clot. 
Does not clot. 


Clots. 


Clots imperfectly. 
Clots. 


Does not clot. 


IX 
Dog of 12.5 kilos 
Inject. of 0.024 gm 
H per kilo 


Before injection | 
. | After injection 
| 


(20 min. periods) 


2.5-1.0 
3.0-4.0-5.5-2.5 


substance formed in the liver to reach the blood through the lymph. 


Other observations of this character have been made.! 


Effects on lymph-flow.— The characteristic effects of proteoses on 
the coagulability, composition, and rate of flow of lymph collected 


1 Cf. CHITTENDEN, MENDEL, and HENDERSON: Loc. cit., p. 162. 


394 L. B. Mendel, F. P. Underhill, and B. White. 


from the thoracic duct is well known. In view of the resemblances 
already noted between the action of the nucleic acids and the prote- 
oses on blood-pressure and coagulability typical lymphagogic effects 
might also be anticipated. The experiments confirm this. In five 
trials it was observed that small doses (Experiments IX, IV) pro- 
duced only slight effects on the flow and the composition of the 
lymph. With larger doses which suffice to accelerate the flow of 
lymph, the latter becomes richer in solids, as after the injection of 
other lymphagogues of this class. Protocols of the experiments are 
given in the table on page 393. 

Immunity. — It is well known that a sufficiently large injection of 
albumoses may confer a certain degree of immunity upon an animal. 
In such cases, when the blood returns to a condition under which it 
is once more coagulable as usual, a second injection fails to render it 
non-coagulable.! We have made a few observations regarding the 
influence of nucleic acid injections in this respect. In Experiments 
IX and IV in which small doses (0.024 gram and 0.044 gram per 
kilo respectively) of preparation H were injected, second injections 
of Witte peptone (0.5 gram per kilo) made after the blood-pressure 
and the clotting-time had returned to the normal, still produced the 
typical effects. The pressure quickly fell, and the blood became non- 
coagulable. In Experiment IV the ugual effect of proteose injections 
on the lymph was also manifested, as the lymph protocols show. 
However, injection of 0.15 gram per kilo of preparation H in Experi- 
ment VII conferred immunity against a subsequent injection of a still 
larger dose (0.19 gram per kilo). The pressure-curve showed only 
a transitory fall, and the clotting of blood samples subsequently taken 
was not retarded. In Experiment III in which 0.065 gram per kilo 
of preparation K was injected rather slowly without producing very 
marked effects (see previous protocols), the animal was found to be 
immune towards a subsequent injection of 0.5 gram of Witte peptone 
per kilo, at least as far as retardation of blood-clotting was concerned.” 
A single experiment (I) on a medium-sized rabbit which received an 
intravenous injection of one-third gram of wheat nuclein C (4.56 per 
cent of phosphorus) equivalent to about 0.05 gram of nucleic acid 
per kilo of body weight, may be recorded here. The animal was 
anaesthetized with one-half gram urethane followed by ether adminis- 


' Cf. CHITTENDEN, MENDEL, and HENDERSON: Loc. cét., p. 158. 
* The slight effects observed in this animal may have been due to a natural 
immunity which has sometimes been noted in dogs. 


A Physiological Study of Nuclete Acid. 395 


tration. The blood samples clotted in nine and one-half minutes 
before the injection (30 c.c. of fluid) which lasted nineteen seconds. 
Blood samples taken two, ten, twenty-two, forty-five, and fifty-five 
minutes after, clotted in twenty-two, ten, twelve, one, and nine minutes 
respectively, the pressure fell more gradually than was usually observed 
in dogs, and began to rise again in forty minutes. Immediately after 
the injection, disturbances in respiration and other evidences of 
excitation were observed, as in the case of the dogs. Bang’s similar 
observations on the relative immunity of the rabbit to injections of 
pancreas nucleoproteid have been mentioned (page 388). 


THe FATE oF NUCLEIC ACID IN THE Bopy. 


The fate of nucleic acid (or nucleins) in the body has usually been 
studied by feeding experiments on man and animals. So far as we 
recall, no extensive study has heretofore been made of the transfor- 
mations which occur when the products are introduced in some other 
way than by the alimentary canal. Consequently it is not definitely 
known whether an important influence on the subsequent utilization 
of the nucleic acid radicals is exerted by the digestive processes. 
From the experiments available, it seems probable that the nucleic 
acid radical is split up only to a very slight extent, if at all, in the 
digestive tract. The proteid compounds of the nucleic acids (nucleins) 
may be broken up by the digestive enzymes into an organic phosphorus- 
containing portion and albumose or peptone. But there is no evidence 
that the purin constituents are liberated before absorption.! We 
have introduced nucleic acid or its compounds into the organism by 
direct injection into the blood-current, into the peritoneal cavity, 
under the skin and per rectum, and have attempted to follow the 
reactions of the substance introduced by searching for its character- 
istic metabolism product, allantoin, in the urine. Some details are 
given below. 

Intravenous injection. — The allantoin was separated by crystalli- 
zation from the concentrated urine, and identified by its characteristic 
crystalline appearance and melting point (210 -220° C.). The ani- 
mals were either starved on the preceding day, or fed on a “ purin- 
free” diet of casein, cracker-meal, milk, and lard. 


1 Cf. PopoFF: Zeitschrift fiir physiologische Chemie, 1894, xviii, p. 533: 
Mivroy: /ézd., 1896-97, xxii, p. 307; LoEw1: Archiv fiir experimentelle Patho- 
logie und Pharmakologie, tgor, xlv, p. 165. 


396 «= L.. B. Mendel, F. P. Underhill, and B. White. 


1. In Experiment VII (see previous protocols), the dog of 8 kilos received 
two injections containing a total of 2.7 grams of preparation H at an 
interval of less than two hours. The urine was removed by means of a 
catheter at intervals of two-thirds, six, and seven and one-half hours after 
the last injection. From the first portion a very small quantity of allan- 
toin was obtained ; from the second, 0.076 gram (m. p. 217°); from the 
third, o.oro gram (m. p. 219°) before the animal was killed. 

From the urine obtained in Experiment VIII, in which 0.6 gram of H was 
injected into a dog of 12.7 kilos, no allantoin could be separated. 

2. A cat of 4 kilos was anesthetized with 75 mgm. of chloralose followed 
by A.C. E. mixture. One and one-half grams of sodium nucleate pre- 
pared by dissolving preparation H in 14 c.c. of very dilute sodium hydrox- 
ide were injected as aseptically as practicable, from a burette through a 
cannula into the jugular vein in thirteen and one-half minutes. Somewhat 
later a second slow injection of 2.14 grams dissolved in 23 c.c. of fluid 
followed. The wound was sewed up and the animal placed under obser- 
vation. ‘The urine (100 c.c.) collected on the following morning con- 
tained o.11r2 gram allantoin (m.p. 217°). The cat remained asleep 
during this entire day and was found dead on the second morning. No 
more allantoin could be obtained. 

4. A medium-sized cat was aneethetized with ether. Two and one-half grams 
of yeast “ nucleinic acid” (P. D. and Co.), dissolved in 38 c.c. of very dilute 
sodium hydroxide, were injected into the jugular vein in thirty minutes. 
The effect on respiration, which became deep and prolonged, was very 
noticeable. The wound was closed, and the animal sat up again in fif- 
teen minutes. The urine collected at the end of the next hour was acid, 
and contained much proteid; but no allantoin separated. From the 
urine of the following three hours an abundant yield of allantoin was 
obtained ; and small quantities were separated from the next day’s urine. 
Proteid continued to be present in the acid urine.’ 


NO 


Intraperitoneal injection. — In his study of the interrelation between 
uric acid excretion and leucocytosis Kuehnau? made intraperitoneal 
injections of thymus emulsions and of thymus nuclein in dogs. This 
procedure gave rise in each case to anincreased output of uric acid 
which the author was unable to attribute entirely to leucocyte influ- 
ences. A fall in body temperature, noted after the thymus injections, 
was attributed to ‘“ peritoneal shock.” Physiological salt solution 
thus introduced produced no change in the uric acid output. 


1 Mr. Wuire has recently observed a large output of allantoin after the injec- 
tion of urates into the systemic and portal circulation. The investigation is being 
continued. 

2 KUEHNAU: Zeitschrift fiir klinische Medicin, 1895, xxviii, p. 561. 


A Physiological Study of Nucleic Acid. 397 


Our animals were fed on a “ purin-free”” diet (milk and cracker- 
meal) just before and during the experiments. The intraperitoneal 
injections were made (without anesthesia) with a large hollow needle. 
Vomiting and diarrhoea were common symptoms observed after the 
injections were made. 


5. A dog of 1o kilos received 3 grams of yeast “ nucleinic acid” (P. D. and 
Co.), dissolved in 125 c.c. of water containing a trace of sodium hydrox- 
ide. The urine collected on this and the following day was acid and 
contained much proteid. Large quantities of allantoin were separated, 
particularly from the portions obtained within the first twenty hours. 
The dog recovered completely, and four days later received a second 
injection (4 grams) for the purpose of ascertaining the effects on the 
uric acid output. The results were negative. Three days later the same 
dog was given a third injection of 8 grams of yeast “‘ nucleinic acid.” 
The symptoms already described were marked, and the animal appeared 
stupid. Within three hours allantoin was excreted in the urine; and 
from the urine of the following two hours there was separated the largest 
yield of allantoin obtained from one animal. No marked increase in 
uric acid output was noted. 

6. A medium-sized dog received an intraperitoneal injection of 4 grams of 
yeast “nucleinic acid.” Within an hour vomiting ensued, followed by 
a rise in temperature and indifferent attitude. About two hours later 
4 grams more were injected. The symptoms were not aggravated. 
The urine collected in the succeeding five hours was very acid, con- 
tained proteid and only little allantoin. 

7. A large cat received an intraperitoneal injection of 4 grams of the yeast 
“nucleinic acid.” Within five minutes vomiting ensued and the animal 
became indifferent. Seven hours later the cat was found dead. The 
30 c.c. of urine previously collected contained a very small quantity of 
allantoin. 


Subcutaneous injection. — Experiments involving subcutaneous ad- 
ministration of nuclein preparations are fairly numerous in medical 
literature. They have been carried out mainly with reference to 
their bearing on leucocytosis, although the effect on uric acid output 
has been studied in a few instances." 


8. The dog used in Experiment 5 received subcutaneously 4 grams of yeast 
“nucleinic acid” in 50 c.c. of very dilute sodium hydroxide. Severe 


1 KUEHNAU : Zeitschrift fiir klinische Medicin, 1895, xxviii, p. 564; MILRoY 
and MALcoLm: Journal of physiology, 1899, xxv, p. 105; BURIAN and SCHuUR: 
Archiv fiir die gesammte Physiologie, 1901, Ixxxvii, p. 304. 


398 = oL. B. Mendel, F. P. Underhill, and B. White. 


diarrhoea followed, but further symptoms were not marked. ‘The urine 
collected during the remainder of the day contained a small quantity of 


allantoin. 


Burian and Schur have noted a marked rise in uric acid output in 
dogs after subcutaneous injections of sodium nucleate (from thymus). 
This was accompanied by increased nitrogen execretion and fever; 
the authors regard the uric acid eliminated in this case as exdogenous 
in origin, and attribute it to pathological processes (cell destruction ? ) 
provoked. In the absence of nitrogen estimations and leucocyte 
counts, we are unable to deny a similar origin to the allantoin obtained 
in the preceding experiments. It seems unlikely, however, that the 
amount of allantoin obtained in some of the experiments (No. 5, for 
example) is entirely attributable to cellular disintegration and endoge- 
nous purin sources. The further possibility of such a reaction is 
suggested, however, by the experiments of Borissow,’ who found 
allantoin in the urine of the dog after poisoning with hydrazine sul- 
phate. In the following feeding experiments, the absence of marked 
toxic symptoms makes it more probable that the allantoin excreted was 
at least in part exogenous in origin and derived from the nuclein 
compounds fed. 

Feeding experiments. -— Numerous feeding experiments with gland- 
ular materials rich in nuclein compounds are recorded in the litera- 
ture. It will suffice here to refer to those observations which were 
made with the isolated substances from which alone more definite 
conclusions regarding the action of the nucleic acid components can 
be drawn. In addition to the experiments of Stadthagen and Gum- 
lich already mentioned, nucleins or nucleates have been fed by 
Horbaczewski,2 Richter,®? Mayer,* Jerome,®? Minkowski,® Milroy and 
Malcolm,’ and Loewi.’ Minkowski was the first to describe allantoin 

1 Borissow : Zeitschrift fiir physiologische Chemie, 1894, xix, p 499. 

2 HorepaczEwskt: Monatshefte fiir Chemie, 1889, x, p. 624; 1891, xii, p. 221 
(spleen nuclein ; man and rabbit). 

® RICHTER: Zeitschrift fiir klinische Medicin, 1895. xxvii, p. 311 (sodium 
nucleate from thymus; man). 

4 Mayer: Deutsche medicinische Wochenschrift, 1896, xxii, p. 186 (spleen 
nuclein; man). 

6 JEROME: Journal of physiology, 1899, xxv, p. 98 (yeast nuclein; man). 

6 MiINKowskKI: Archiv fiir experimentelle Pathologie und Pharmakologie, 
1898, xli, p. 403 (salmon nucleic acid; dog and man). 

Mitroy and Matcoum: Journal of physiology, 1898, xxiii, p. 217 (nucleic 
acid ; man). 

§ Loewr: Archiv fiir experimentelle Pathologie und Pharmakologie, 1901, xlv, 
p. 157 (various nucleic acids and nuclein products; man). 


_ A Physiological Study of Nucletc Acid. 399 


as a metabolism product of an isolated nucleic acid in the dog. The 
above observers have noted a pronounced rise in uric acid output. The 
apparent relation between uric acid (noted in man) and allantoin 
(in the dog) has already been suggested. A few extracts from our 
protocols follow. The negative results which were occasionally 
obtained, are not recorded here, since they do not affect the observa- 
tions made, and are not surprising in view of the difficulty in separat- 
ing small quantities of allantoin, and in the absence of data regarding 
the extent to which the substances fed were absorbed. As before, 
the animals were always fed on a “ purin-free”’ diet of milk, cracker- 
meal, and lard for some time previous to the feeding experiment with 
the nucleic acid compounds. 


9. Dog of 6.8 kilos. Ona “ purin-free ” diet, such as that described above, 
no allantoin appeared in the urine. During two days a total of 17 grams 
of preparation D (containing 2.9 per cent of phosphorus and equivalent 
to about 5.6 grams of nucleic acid) were added to the diet. The urine 
of these days yielded 0.21 gram of allantoin (m.p. 217°). Two days 
later, when the urine was again free from allantoin, 30 grams of dessi- 
cated salivary glands! of the ox were added to the diet. The urine of 
the following day furnished 0.023 grams of allantoin (m. p. 219°). 

to. Dog of 3.3 kilos. The diet consisted of casein, cracker-meal, and lard. 
On one day, 5 grams of preparation H were added to the food. The 
urine for this day yielded 0.445 gram of allantoin (m. p. 219°). None 
was obtained on the following days, even when 15 grams of Armour’s des- 
sicated salivary gland were added to the diet. 

11. A cat of medium size was fed on cracker-meal and milk. In the course 
of two days, 10 grams of preparation D (equivalent to about 3 grams of 
nucleic acid) were added. A small yield of allantoin (m. p. 217°) was 
obtained on the day following only. The subsequent addition of 0.36 
grams of preparation F produced no detectable trace of allantoin. In 
another cat negative results were obtained after feeding 6 grams of prep- 
aration D. : 

12. A kitten about four weeks old was fed on milk. ‘The daily addition of 
I gram of yeast “nucleinic acid” to the milk was followed by an excre- 
tion of allantoin. ‘The urine previously collected was free from the 
latter. 

13. The attempt to induce an excretion of allantoin by feeding emulsions of 
thymus glands to rabbits was unsuccessful. 


1 Specially prepared for us by ARMouR and Co., Chicago. 


400 L.B. Mendel, F. P. Underhill, and B. White. 


Rectal feeding. — Mochizucki! has shown that the administration 
of thymus gland substance per rectum is also followed by a typical 
increase in uric acid excretion in man. In two cases out of three 
experiments on dogs, we have obtained allantoin excretion after a 
similar procedure. The third animal eliminated uric acid in appar- 
ently increased amount. In each instance a “purin-free” diet was 
simultaneously fed. Thymus glands were macerated with water and 
strained through cloth. About 70 c.c. of the resulting emulsion 
were injected after addition of 2 grams of common salt and 5-8 drops 
of laudanum. Sometimes the gut was emptied on the previous day 
by an enema of soap and water. A single experiment on man will 
be referred to later. 

Experiments on man. — Milroy and Malcolm? have studied the 
effect of nucleic acid on nitrogenous metabolism and leucocytosis in 
man, They found only a very small rise in uric acid output. The 
P,O, elimination and the number of leucocytes were both noticeably 
increased, although only one-half and one gram of nucleic acid were 
taken respectively on the two days selected. They were unable to 
give larger doses of nucleic acid because of certain somewhat disagree- 
able symptoms (severe muscular tremors) which arose after the 
larger quantity had been given. Loewi? has attributed the latter to 
some foreign toxicity in the nucleic acid fed. In his own experiments, 
larger quantities (30 grams) were taken without apparent ill effects, 
although he noted unfavorable symptoms after taking a commercial 
preparation of sodium nucleate prepared from yeast. Neumann * has 
reported the absence of disagreeable symptoms after administration 
of the thymus nucleic acid. Loewi concluded from his experiments 
that uric acid is the only specific nitrogenous end-product of nuclein 
metabolism in man. 

The present feeding experiments were carried out on two of us. 
The daily diet selected from foods which are practically ‘“ purin-free,” 
was maintained unchanged throughout the entire periods. Prepara- 
tion M, containing 7.35 per cent of phosphorus and 16.3 per cent of 
nitrogen was added to the diet on two successive days. It was dis- 


1 MocuHiIzuck!1: Archiv fiir Verdauungskrankheiten, 19o1, vii, p. 221. 


; MILRoy and MALCOLM: Journal of physiology, 1898, xxiii, p. 227. 

* Loewl: Archiv fiir experimentelle Pathologie und Pharmakologie, 1901, xlv, 
p. 157. 

* NEUMANN: Verhandlungen der physiologischen Gesellschaft, Berlin, 1898. 


No. I1, 12, 13 (according to LoEwt). 


A Physiological Study of Nuclete Acid. 401 


solved in a minimum quantity of dilute sodium hydroxide and added 
to the milk taken at noon. The total daily food consisted of : 


Bi W: Rigekre is 
Grams. Grams. 
Bape ea edi st Je wag maha SSO 445 
Dresee toe oe hes «sb 275 275 
BREE Wile aren teen: 6 one BOO 1850 
Pcie ats da gc 75 96 
QELS ig: Fee ae Molen We I50 115 
estimated to contain 
jeeeted See hae LE 138 
Vidhan oA es oc. en we LAD 191 
beacnchediate Werk «a a eZ0 235 
ReeC@UMted 16> VICI as tet Ae es . 29¥0Cal. tagronCal. 


The food was taken in three portions, about one gram of salt being 
added at each meal. The urine and feces were collected for each 
twenty-four hours and analyzed. The faces were covered with 
alcohol containing a few drops of sulphuric acid, and were dried on a 
water-bath and pulverized. Nitrogen was estimated by the Kjeldahl- 
Gunning method ; P,O, in the urine by titration with uranium solu- 
tion, in the faeces, gravimetrically by Neumann’s method;! uric acid 
by the Ludwig-Salkowski process; ethereal SO, by Baumann’s 
method. On one day nucleic acid (dissolved as usual and mixed 
with a part of the daily milk portion and a few drops of laudanum) 
was introduced per rectum through a long flexible catheter. The 
intestine had previously been subjected to an enema of soap-water. 
(See B. W., 8th day.) The data are summarized in the table on 
page 402. 

Leucocyte counts made on the two days before and during the 
nucleic acid feeding period showed only slight changes, if any. Thus 
the counts on B.W. during the first three days gave the following 
numbers: 5780—5460—5780—5150—7812; during the nucleic acid 
days: 5200—6400—5000—5210—5 780—7320—6580. The counts on 
F. P. U. were as follows; on the first two days: 9360—8120—10620; 
on the nucleic acid days: 10620—11240—11400—12280. 

Both subjects complained of slight muscular soreness on one of 
the nucleic acid days. The writer was at the time inclined to over- 


1 NEUMANN: Archiv fiir Physiologie, 1897, p. 552; cf. also ZADIK: Archiv 
fiir die gesammte Physiologie, 1899, Ixxvii, p. 2. 


“snut}I] 0} plow 


A19A SBA OUTIN OY], 


620 | LCT se ICI FOLT'O 16¢°€ SLe0 LYSE STS'91 9¢0'T LOOT 6 
FLOT crL'0 0¢ 69 FOST'O EP9'E ICE O10’ f6F 21 | ¥20'T [sol 8 
aS SIS'T cba O€ Of 6SZT0 Lye ere'0 soc’ 699° LT 9¢0'T [gol L 
~S FLT 9LO'T 6¢ 16 | séet‘o OIZE | Spe. | 69€°¢ C60 OL 9CO' | FrOL 9 
S Cle + 69EC S$ 161 | SLbT‘0 619'€ OO) SIGE CtO'ST 0c0'T O¢sT S$ 
if 0062 | +6S'T LE Zee se OOSLON a GSSiS 89s°0 $99°¢ LSE LT 1co'T OOF v 
RQ cer I L960 €¢ 94 | +¥ELT‘0 LSPS 6£ee0 Tae err OL +0" T cOOl £ 
S Ie6'T FOL" 1 TE S6 ccSl0 SSSKE 99€0 | S0ZEe 096'°ST £70'T OFTT c 
. £60 | E90 ST Oo Ssorlo S88'S 9EL0 LIS @ Gua LT0'T SOST I 
~ ‘SO[Iy 1°79 ‘3Yystam Apog ‘plo sawak ¢Z “MA “gq 
3 Lei : £ Mt 
< | a a Se a ae ———— — = = 
Y | 
7 Sé8'0 £60 (Ke 69 6lelo | CEL’ 06+'0 F88'¢ 996 LT 9C0'T O9TT 6 
dD LS6'T 660 C es S6T | S6t'0 lpecOuic: esr'0 £6L°¢ €hs'Sl | L420'T Occl 8 
: SLTZ STS"? 8S PLT LS9T‘0 | eSs"e gss0 cr" 99¢61 | cc0'l Scel £ 
Q S00'C S67 OS SST cEELO | -L90°¢ 1 +980 LOS'€ C69 91 9C0'T OST 9 
ke eryTt | Oss T 9¢ Scl | {2900 | 6L6€ | 1Z€S°0 LIV $1602 LE0'T OO€T s 
98E'S OSS 6b OST 6cLT0 | SOF | I8E0 C66 6LL ol Sd0'T O09 1 
= Sté6'T Lesley 802 | 9F9TO | 966¢ €£E'0 OSL ¢ e102 £c0'1 SO¢T is 
8 +969 L9S9 | eel E16 |) L6ST0 CLT E clr0 Love 6LE LT 1c0'l oll é 
S S8o'r Or | 88 WG |) (0ldke|h0) S8S"t 6s 0 9E9'C SIMON ad) | S86 I 
eS = fe 
: ‘sug | sug f-sury | csuy | wy “sur “wy “SUILT) ‘sun | aft) 
el ong, | Uasoaytu |3y81Iem|1y3Iem | “Fos WoS *pioe ae ‘uaforiu | a3 ‘ds if: 
é O'd [eq0 J, Aq | 3stoy | Teorey | peioy, II) ‘O'd [210], ION 
N a ae 
‘SHOW ‘ANIN 
‘SOTTYT OL ystem Apog ‘po savak $7 “Qa ‘ay 


402 


‘NVI NO SLNAWINAIXY WSITIOAVLAIY 


“SUIpUL}S UO DULIN dy} Woy payeavdas s[eysks9 plow uy _ 


1 § winq9901 
| }aad Fy ploe o1aponu “ws 9 


poy JW ploe s1aponu “ws g 


| paj JX ploe oraponu ‘wid g 


p2j JW plow oraponu "ws + 
p2y JW plore o1ayonu “ws + 


A Physiological Study of Nuclete Acid. 403 


look any direct bearing of these remarks; their possible significance 
was again emphasized after reading the experience of the subjects of 
Milroy and Malcolm and of Loewi. For this reason also no larger 
quantities of nucleic acid were fed to man. 

The results obtained with F. P. U. are somewhat irregular, and 
indicate a slight “lag” in the excretions on certain days. In each 
subject a noticeable increase in uric acid output followed the in- 
gestion of the nucleic acid. In the absence of any other purin 
constituent in the diet and (in B.W. at least) of any marked 
leucocytosis (and probable leucolysis), we must attribute the increase 
in uric acid, as Loewi does, directly to the nucleic acid ingested. It 
is, in the language of Burian and Schur, exogenous in nature. The 
low average daily output of endogenous uric acid in our two subjects 
on a “purin-free ” diet falls within the limits ascertained by Burian 
and Schur! in different individuals. 


SUMMARY. 


The more important observations recorded in this paper indicate 
that the vegetable nucleic acid obtained from the wheat embryo 
resembles, in its physiological effects, the guanylic acid of the pan- 
creas. Introduced in sufficient doses into the circulation, it may 
produce a fall in arterial pressure; a change in the coagulability of 
the blood; an increase in the flow of lymph and a change in its 
composition; and perhaps, also, a degree of immunity toward subse- 
quent injections. 

The ingestion of nucleic acid is followed in man by an increased 
output of uric acid, and in the dog by the excretion of allantoin. 
These products correspond in either case to only a portion of the 
purin radicals introduced. In animals, allantoin excretion was also 
observed after the introduction of vegetable nucleic acids into the 
body per rectum, intravenously, intraperitoneally, and subcutaneously. 
Some features of intermediary purin-metabolism are discussed. 


1 BuRIAN and Scuur: Archiv fiir die gesammte Physiologie, 1900, Ixxx, 
p- 302. 


THE ACTION OF ACIDS AND ACID SALTS ON, SECar— 
CORPUSCLES AND OTHERSCEELS: 


By 5S. PESKIND2 


From the Physiological Laboratory, Western Reserve University. 
Ly. S J 3B; 


Ween investigating a large number of laking agents, in con- 
junction with Dr. G. N. Stewart, I made the observation that 
small quantities of ferric chloride or hydrochloric acid cause aggluti- 
nation and precipitation of blood-corpuscles. 

To determine what substances would give this reaction, the follow- 
ing list of chemicals was tested. The purest reagents were used, 
Merck’s and Schuchart’s for the most part. Defibrinated dog’s blood 
was employed in all the experiments, although the reaction was found 
to take place also with human blood, cow’s blood, and chicken’s 
blood. 

The following acids caused good agglutination and precipitation : 


Sulphuric Acetic Chlorine water 
Hydrochloric Oxalic Bromine water 
Nitric Citric 
Phosphoric ‘Tartaric 
Iodic Lactic 
Chromic ‘Tannic 
Molybdic Pyrogallic 
Phosphomolybdic Benzoic 
Phosphotungstic Hippuric 
Sulphanilic 


The corpuscles precipitated by the above acids laked very rapidly. 
No extraneous precipitate (z. ¢., no precipitate of serum constituents) 
was thrown down with the corpuscles. 

The following acids did not produce agglutination and precipitation : 


Arsenious Uric 
Boric Eosic 
Hydrogen sulphide Acid fuchsin 


Carbon dioxide 
Carbolic 
Osmic 
' H. M. Hanna Fellow for 1go2. 
404 


Action of Acids and Acid Salts on the Blood. 405 


Very large as well as very small, amounts of a solution of arsenious 
acid failed to give the reaction with whole blood or washed corpuscles. 
After adding the arsenious acid a drop of 0.2 per cent hydrochloric 
acid was added, which immediately gave a good reaction. Washed 
corpuscles treated with excess of hydrogen sulphide turned a greenish 
color, owing to the formation of sulfo-methazmoglobin.! 

Pyrogallic acid changed the color of washed corpuscles to olive 
green when just enough was added to cause precipitation. 

Chromic acid, if used in slight excess, hardens the corpuscles at the 
same time that it precipitates them, so that they show no laking 
after many hours. Under the microscope, they look angular, have 
sharp margins, and are clearly hardened. 

The following salts were tested. All the salts mentioned in the 
table cause agglutination and precipitation, ercept those marked with 
an asterisk. 

“ A” denotes acid reaction, ‘““N” neutral reaction. 


Reaction of 
solutions to 
litmus paper. 


DOMME ee SUIDLALE ° oy ke a" Ske ga cc 


*Potassium acid tartrate . 

*Potassium and sodium tartrate 

*Potassium and antimony tartrate . 
Potassium acid sulphate . 


*Calcium glycerinphosphate 


*Ferrous sulphate . ; 
Ammonio-ferrous sulphate . 
Hypophosphite of iron (ferrous) . 
Ferric nitrate : 
Ammonio-ferric sulphate lites avy He 
Eealyze@iron: i. -.. sey kis, -s 2s? . Bledehed: himus 
Ferric chloride . 

*Pyrophosphate of iron 


>PrPrrZz PP PPap 


*Niuckel sulphate ©. +s 4: says 
*Nickel and ammonium sulphat 


Zinc sulphate 
Zinc chloride 


rr azagZzaar 


1 Hoppe-SEYLER: Handbuch der chemischen Analyse, p. 281. 


406 S. Peskind. 


Reaction of 
solutions to 
litmus paper. 


Aluminium chlondet.s29'.. ena eae ee 
Aluminium sulphate. . . . ARNE Se As 
Alum (d’ble sulphate of potassium “ai alstulibithea) Fa 


*Cobaltschloride® | 2. 04 SN eee oe eee 
*Cobal€ nitrate ©. 20) | Ss Gee eee 
*Manganese sulphate... \. G0 0 eo ee 
* Manganese nitrate, “s.r ee ys 
*Chrome lune.” Sr.) caked ee ee ee 
*Potassium bichromate 
* Ammonium chromate N 
*Potassium chromate . N 
Copper sulphate A 
* Lead acetate N 
*Mercuric cyanide . N 
*Mercuric bichloride N 
Mercury bisulphate A 
Silver nitrate A 
Gold chloride A 
Gold and sodium chloride A 
*Platinic chloride 4 ‘ 
Stannous chloride A 
Uranium acetate A 
Ammonium molybdate A 
*Sodium tungstate . N 
*(Quinine bisulphate A 


*Methylene blue 


The reagents to be tested were dissolved in 0.9 per cent sodium 
chloride solution in such proportion that the solution contained from 
0.1 per cent to 0.§ per cent of the substances. A\ll these solutions 
are therefore slightly hyperisotonic. In working with leucocytes 
(see experiments on leucocytes), 0.9 per cent sodium chloride solu- 
tions containing about 0.04 per cent of the reagents were employed, 
as stronger solutions did not allow the reaction to take place. Some 
reagents are but slightly soluble in water, ¢.g., hippuric acid, Of 


Action of Acids and Acid Salts on the Blood. 407 


these a saturated solution in 0.9"per cent sodium chloride solution 
was used. 

In making the experiments a few drops of defibrinated dog’s blood 
were suspended in a little 0.9 per cent saline (a 10 per cent blood- 
suspension is best), and the reagents were added drop by drop until 
the blood-corpuscles came down as a flocculent precipitate. One can 
thus determine quite accurately the amount of precipitant necessary 
for the exact precipitation of all the corpuscles. If undiluted blood is 
used, it becomes solidified from the precipitate of corpuscles that 
forms, and one cannot discern the exact points where precipitation 
begins and ends. 

Excess of reagent must be avoided. — A slight excess causes rapid 
laking of the precipitated corpuscles, if the latter be kept at room 
temperature. But if the precipitated blood be kept on ice at 0°, the 
laking is retarded for many hours. If more than a slight excess of 
precipitant is added, the reaction will not take place. In the case of 
some reagents, where every excess must be avoided in order to 
demonstrate the reaction, it is advisable to reverse the procedure 
given above. Blood is added a little at a time to a weak solution of 
the substance to be tested. As long as the reagent is in excess, no 
precipitation occurs, but when enough blood has been added an 
immediate agglutination and precipitation of the corpuscles takes 
place. : 

On looking over the above table, it is observed that many inorganic 
and organic acids produce agglutination and precipitation of blood- 
corpuscles, the only exceptions among the acids tested being osmic 
acid, arsenious acid, boric acid, uric acid, hydrogen sulphide, car- 
bonic acid (carbon dioxide), carbolic acid. It is seen further that many 
salts give the reaction. The active salts have the following character- 
istics: they all have a strongly acid reaction; and they have for bases 
one of the following heavy metals, iron, zinc, aluminium, silver, mer- 
cury, gold, tin, molybdenum, exceptions being the acid sulphates of 
sodium and potassium, which must be included in the list of active 
salts. Only the acid salts of these metals cause precipitation of the 
blood-corpuscles. Cream of tartar, which hasa strongly acid reaction, 
as also chrome alum, calcium glycerine-phosphate, potassium and 
antimony tartrate, quinine bisulphate, all strongly acid salts, do not 
give the reaction. Combinations which contain the metalloids, 
boron, arsenicum, and antimony are apparently unable to agglutinate 
and precipitate corpuscles. No neutral salt, no matter what metal it 


408 S. Peskind. 


has for a base, was found to give the reaction. One might think that 
possibly there was enough free acid contained in the acid salts to 
cause precipitation of the corpuscles, but it will be shown later 
that the reaction is due to the salts themselves, and not to free acids 
which might exist as impurities. 

The corpuscles precipitated by acids lake very rapidly at room 
temperature, if even a slight excess of acid be present. The ferric 
chloride precipitates lake not so readily, but still very rapidly. Cop- 
per sulphate precipitates remain unlaked for a considerable time, even 
at room temperature. In another part of the paper will be described 
in detail a procedure for preventing the laking of precipitated cor- 
puscles. 

The flocculent precipitate of blood-corpuscles produced by any of 
the above-mentioned acids settles rapidly, leaving a clear and colorless 
supernatant liquid which shows no evidence of laking and no extra- 
neous macro- or microscopic precipitate, ¢. ¢., of globulin. On centri- 
fugalizing for two or three minutes, the corpuscles are found to be 
agglutinated into a firm gelatinous mass at the bottom of the centri- 
fuge tubes. With the microscope, no extraneous precipitate can be 
seen in or around the masses of agglutinated blood-corpuscles. If 
ferric chloride is used for precipitating, a small amount of a white 
granular precipitate is seen on the top of the cake of corpuscles and 
in the supernatant liquid. 

In the case of copper sulphate, a greenish-white granular precipitate 
is seen above the sediment of corpuscles, and, with the microscope, 
in and around the masses of corpuscles. Similar precipitates are 
found in the case of other acid salts used. They are soluble in a few 
drops of acid or alkali and consist apparently of globulin. They 
may be removed almost completely by several washings with 0.9 per 
cent saline, and subsequent centrifugalization, as they are very light, 
and are floated off by the saline and possibly in part dissolved. 

The color of the sediment of blood-corpuscles precipitated by any 
of the reagents (with two exceptions) is perfectly normal, provided the 
reagents are not added in excess. Only in the case of pyrogallic acid 
and gold chloride have I found the color of the corpuscles altered in 
the precipitation, due to a simultaneous action on the hemoglobin. 

Microscopic examination of the precipitate of blood-corpuscles pro- 
duced by acids shows the following: One sees big clumps consisting 
of agglutinated masses of red corpuscles of normal color, absolutely 
no extraneous precipitate being visible. The contour of the cor- 


tt ale tee til 


Action of Acids and Acid Salts on the Blood. 409 


puscles is perfectly filled out, and’even if they are very much cre- 
nated before, they look full and rounded after precipitation, all the 
crenation having disappeared, showing that something has entered 
the corpuscles. Corpuscles precipitated by ferric chloride and 
other acid salts tend to assume a globular form, and usually are sub- 
normal in size. 

The corpuscles are all quite uniform in size, even when in the 
original blood-suspension they may have shown marked disparity in 
regard to size. 

The corpuscles show no swelling if the reagents are not in excess. 
If the precipitated corpuscles are watched, on the slide, at room tem- 
perature, they will (if even a very slight excess of reagent has been 
added) be seen to swell, grow paler and dimmer, and then lose their 
hemoglobin; finally only agglutinated masses of stromata remain. 

If a slight trace of a caustic alkali (ammonia or sodium hydrate) 
is added to the liquid containing the precipitated corpuscles, and the 
whole shaken a little, the clumping is immediately and completely 
broken up, so that the corpuscles remain suspended singly in the 
liquid and do not settle for a long time, just as in normal blood. 
Under the microscope no agglutinated masses are to be seen, each 
corpuscle being separate from its fellows. On adding one of the pre- 
cipitants to this restored blood, the corpuscles are reprecipitated. A 
moderate amount of serum -added to precipitated blood is unable to 
restore the normal suspension of the corpuscles. 

If blood is added to an excess of reagent, agglutination and precipi- 
tation do not occur until the blood added reaches acertain amount. It 
seems that an excess of acid or acid salt surrounding the corpuscles 
is inimical to the agglutination and precipitation of the latter, 
although red corpuscles once agglutinated always remain in that 
condition in spite of any excess of reagent that is afterward added. 

If corpuscles precipitated by hydrochloric acid or ferric chloride are 
washed, well shaken with saline and centrifugalized, it becomes pro- 
gressively more difficult after each washing to remove the corpuscles 
by centrifugalization. If this process be repeated five or six times, 
and the corpuscles be shaken well each time, they will resume an 
almost normal suspension, which becomes almost quite normal if the 
corpuscles are allowed to stand in the ice-chest over night suspended 
in saline solution. 

If these washed hydrochloric acid or ferric chloride corpuscles are 
suspended in saline they can be reprecipitated by very minute quanti- 


410 S. Peskind. 


ties of the precipitants. The precipitates have the same macroscopic 
and microscopic appearances as when ordinary blood is acted upon 
by the same reagents, the only difference being that the extraneous 
granular precipitate ordinarily seen when the metallic salts are used 
is absent in this case. 

What takes place in the precipitated corpuscles when they are so 
thoroughly washed that they resume, practically, their normal suspen- 
sion and can be reprecipitated on again adding the reagents? Is the 
precipitant removed from the corpuscles by an osmotic process which 
is favored by removal of all adherent serum? This is very improb- 
able, as the combination between reagent and corpuscle is a firm one, 
as will be shown elsewhere. 

Two other explanations may be advanced. We know that alkalies 
break up the agglutination of the corpuscles. Possibly the distilled 
water used in preparing the saline contains a sufficient amount of free 
alkali (ammonia ?) to combine with the precipitating acid or salt con- 
tained in the corpuscles. This again is hardly probable, as but a 
very small amount of ammonia exists in distilled water. The most 
reasonable explanation is that the alkali present in the interior of the 
corpuscles diffuses toward the periphery, and on reaching the sur- 
face layer of the corpuscle (which for convenience we shall call the 
envelope) neutralizes the acid or acid salt contained therein. As 
soon as the normal alkaline reaction of the envelope is restored 
wholly or in large part, the agglutination will be broken up. The 
repeated washings required in order to restore the normal suspen- 
sion of the corpuscles acts, according to this view, by removing the 
acid or acid salt on the outside of the corpuscles, which would tend 
to keep the envelope acid. As an illustration, the following experi- 
ment may be quoted. 

The washed corpuscles of one cubic centimetre of dog’s blood were 
diluted with ten parts saline, and 0.2 per cent hydrochloric acid in 
saline was added drop by drop, and the tube shaken. After 0,2 
or 0.3 c.c. of the acid has been added, good precipitation occurred. 
On shaking the tube thoroughly, the agglutinated and precipitated 
corpuscles resumed their normal suspension. Another drop of the 
acid was added, and again the corpuscles were agglutinated and pre- 
cipitated. After vigorous shaking, the corpuscles again resumed their 
normal suspension. This could be repeated until 0.6 c.c. of acid had 
been added, when the corpuscles came down as a very heavy floccu- 
lent precipitate, The agglutination could no longer be broken up, 


Action of Acids and Actd Salts on the Blood. 411 


and soon after the corpuscles laked. It seems probable that in this 
case after each addition of the acid the alkali of the interior of the 
corpuscle diffused to the surface and neutralized the acid in the sur- 
face-layer of the corpuscle. After 0.6 c.c. of the acid had been 
added, the alkali was almost completely used up, and the agglutina- 
tion could not be destroyed. 

In what ways can agglutination and precipitation of corpuscles be 
prevented? I made attempts to harden the surface of the corpuscles 
(or envelope) by small quantities of hardening agents, thinking 
thereby to make them incapable of being agglutinated. In the case 
of washed corpuscles, precipitation cannot be prevented by adding 1 
or 2 per cent of formaldehyde in 0.9 per cent sodium chloride, or 0.2 
per cent osmic acid (in sodium chloride solution) immediately before 
the addition of the precipitant. Blood-corpuscles hardened in 2 per 
cent formaldehyde (in saline) for forty-eight hours are still aggluti- 
nated and precipitated by ferric chloride. The hamoglobin in these 
corpuscles is thoroughly fixed, as is shown by their dark brown color, 
and the fact that a large excess of laking agents, such as hydrochloric 
acid, fails to lake them. 

Why does cold retard the laking of corpuscles precipitated by acids 
and acid salts? Loewy and Zuntz have noticed that in the titration 
of blood-corpuscles with acids in order to determine their alkalinity, 
lower figures were obtained if the titration were carried out at 0° C. 
So it seems most likely that the diffusion of the alkalies in the interior 
of the corpuscles towards the surface is greatly retarded by cold. 
This would serve to explain the effect of cold in retarding laking. 


Incidentally an experiment is perhaps worth mentioning which seems to show 
that after the action of precipitating agents the hamoglobin of the corpus- 
cles is in a peculiarly labile condition, so that a very slight change is suffi- 
cient to cause its liberation. After a slight excess of ammonio-ferrous 
sulphate the corpuscles settle rapidly, leaving a perfectly colorless super- 
natant liquid. Microscopically, agglutinated masses of normal looking 
cells are to be seen. On slightly pressing the cover slip, the hemoglobin 
was seen to pass out of the corpuscles and away from the point of pres- 
sure, leaving almost hemoglobin-free corpuscles. 


If blood be allowed to run direct from an artery into saline solution 
containing an amount of the ferric chloride exactly calculated to pre- 
cipitate all the corpuscles, the serum will not clot even after several 
days. 


A412 S. Peskind. 


The following experiments were made. Four glass graduates were 
employed. 


Graduate I contained 3.9 c.c. of a solution of 1.2 per cent ferric chloride in 
saline + sufficient saline to make 4o c.c. Into this mixture 1o c.c. of 
blood was allowed to flow directly from a cannula in the carotid of a dog, 
and stirred thoroughly. ‘The blood-corpuscles were precipitated imme- 
diately and completely. After half an hour no clotting had taken place. 
The mixture was centrifugalized and the serum decanted. The blood 
cake was absolutely unlaked after twenty-four hours, standing at room 
temperature. The serum, which was fevfectly neutral in reaction, did not 
clot after two days. 

Control. Graduate II contained four parts saline. One part blood was 
allowed to flow in, and mixed. After three minutes the blood was com- 
pletely clotted. 

Graduate III contained 2 c.c. of ferric chloride + saline up to 15 c.c. Added 
5 c.c. blood ; precipitated ; no clotting after half an hour. Centrifugal- 
ized; serum unclotted after two days, 

Control. Graduate 1V contained three parts saline + one part blood. Clotted 
within three minutes. 


Whether the absence of clotting was due to the precipitation of all 
the corpuscles which thus prevented any liberation of fibrin ferment 
(which is the most probable cause), or whether it was due to the 
neutralization of the alkalinity of the serum, must be determined by 
subsequent experiment. 

What constituent of the blood ts concerned in this reaction — the 
serum or the corpuscles ? That the serum constituents take no part, 
is shown by the fact that thoroughly-washed corpuscles, absolutely 
free from serum,’ are agglutinated and precipitated by the same 
reagents which act on entire blood. 

In this connection it may be of interest to give the results of some 
quantitative experiments showing the amounts of reagents required 
to produce the reaction. 


Thoroughly washed dog’s corpuscles were suspended in o.g9 per cent sodium 
chloride solution, so that 1 c.c. contained 2,150,000 corpuscles (by 


' The corpuscles are washed until the last wash water (0.9 per cent saline) gives 


no precipitate with ferro-cyanide of potassium and acetic acid, showing an entire 
absence of serum proteids and of any free haemoglobin which might be present if 
some of the corpuscles had laked. This last wash water also should give no pre- 
cipitate with copper sulphate or ferric chloride. The object of these tests is to 
show that there is nothing in the liquid surrounding the corpuscles which will 
give a precipitate with the reagents. 


a i i, ee iE a 
; ‘ 


Action of Acids and Acid Salts on the Blood. 413 


Gower’s hemocytometer). To definite quantities (1 c.c. and 1} c.c.) of 
this suspension, placed in very small test tubes, were added weak solutions 
of copper sulphate, ferric chloride, and hydrochloric acid in 0.9 per cent 
sodium chloride solution. ‘These solutions were added cautiously from a 
capillary pipette until the corpuscles came down as a flocculent precipitate. 
The end-reaction is quite exactly determined, as the close agreement of 
the duplicate experiments will show. 


TABLE I. 
1 c.c. of blood-suspension required, 
Of solution of copper sulphate, 0.4 percent . . . . 0.25 C.c. 
Duplicate experiment, copper sulphate . . . . . . 0.28 c.c. 
Average.) GE. oy oh ey LOG Ete. 


Therefore 1 c.c. blood = 0.00106 gram of copper sulphate. 


1 c.c. blood required, 
Of hydrochloric acid solution, 3; percent . . . . . o28c.c. 
13 c.c. blood required, 
Hydrochloric acid, 0.42 c.c. = 0.28 c.c. for 1 c.c. blood 
AVIA te ie siete =n) ee eet 
Therefore 1 c.c. blood = 0.000112 gram of hydrochloric acid. 


1 c.c. blood required, 


imeetse Chiocide, 0.24 percent . .) 9s « ~ ~ OLR Ge. 
13 c.c. blood required, 
Of ferric chloride, 0.24 pefcent . . . ey ah ee ee” 
or 0.147 c.c. for 1 c.c. of ised: 
Average . . . bah asin SE Aa oe 


Therefore 1 c.c. blood = o. nue gram of Peat chloride. 


Therefore 1 c.c. of a suspension of dog’s corpuscles in saline, contain- 


ing 2,150,000 corpuscles to I C.c., 
0.00106 g. of copper sulphate. 


Required for precipitation } 0.ooo112 g. of hydrochloric acid. 
0.00036 g. of ferric chloride. 


The blood in the above experiments was not diluted. 
In the following experiments a 5 to 10 per cent suspension of washed 
corpuscles or whole blood was employed. 


TABLE II. 


1 c.c. of whole blood took . . 3.7 c.c. of tartaric acid (0.3 per cent). 
The washed corpuscles of 1 c.c. 
of the same blood took . . 2.5 c.c. of tartaric acid. 


AI4 S. Peskind. 


TABLE ILI (contenued). 


1 c.c. of whole blood took . 1.2 c.c. of 0.02 per ct. hydrochloric acid. 
CoM ord a nN " 4 Maer ES i 
The corpuscles of 1 c.c. blood 

took (dilutedito’s5°c.c)) Sic. 7ete-e. 


The corpuscles of t c.c. blood 


ve ee ee 


took (dilutedito5°c.c;) =. O.g5 exc. K be a 
1 c.c. whole blood took . . 2.9 c.c. of 0.02 per cent ferric chloride. 
Corpuscles of 1 c.c., diluted to 

LOrG:C, tOOk: Ga diss 2 of yee ere neces ct i 
Corpuscles of 1 c.c., diluted to 

EO 1:5 "tOOKe age. hl <s) e-8 NOrs.e-e: Pf s i 


It is seen from these figures how very minute is the quantity of 
reagent that enters into the reaction. These results demonstrate 
another fact. If the reaction produced by acid salts were due to any 
free acid present as impurity, then one would expect that a compara- 
tively large amount of the salts would be required to cause the pre- 
cipitation. Such is not the case. It takes only three times as much 
of ferric chloride as it does of hydrochloric acid to cause precipitation 
(Table I). Now it is absurd to consider that there is thirty per cent 
free hydrochloric acid in Merck’s ferric chloride, which would have to 
be the case were this reaction due to free hydrochloric acid. So with- 
out question the entire metallic salt enters into the reaction. 

Having established the fact that the agglutination and precipitation 
is due to an effect on the blood-corpuscles themselves, the question 
now arises as to what part of the corpuscles is concerned in this 
reaction. Is it the stroma or the hemoglobin? A very simple 
experiment suffices to decide this point. Whole blood or washed cor- 
puscles are laked by any hzemolytic agent, ¢. ¢., ether, water, or sapo- 
toxin, best by a small quantity of a weak solution of sapotoxin (0.2 
per cent in saline). Just enough sapotoxin should be added so that 
laking takes place slowly, say in three to five minutes; this leaves 
the precipitated ghosts well preserved. After stronger solutions of 
sapotoxin or water, the stromata appear considerably disintegrated. 
To the laked and transparent blood solution a trace of acid or any 
other precipitant is added. Immediately a heavy precipitate, consist- 
ing of agglutinated masses of ghosts, is thrown down. 

Acids do not precipitate anything but the stromata. The metallic 
precipitants, however, may throw down some of the haemoglobin at 


Action of Acids and Acid Salts on the Blood. 415 


the same time, if the solution of the blood is concentrated. The sub- 
stances which do not precipitate intact blood-corpuscles, do not preci- 
pitate stromata, an exception being carbon dioxide, which precipitates 
ghosts, but not corpuscles. 

The reaction affects the leucocytes as well as the red corpuscles. This 
may be seen in ordinary blood, and still better by working with a pure 
suspension of leucocytes. About 16 c.c. of leukzmic blood, kindly 
given to me by Dr. W. T. Howard, Jr., was received in a sterile test 
tube a few hours after the autopsy. The blood was found separated 
into two layers. The upper three-fourths consisted of a very yellow 
milky-looking serum which contained a large number of leucocytes 
and absolutely no red corpuscles. The lower one-fourth consisted of 
red corpuscles and some leucocytes. I diluted 1 cc. of the milky 
serum with 30 c.c. of 0.9 per cent sodium chloride solution, centrifu- 
galized and washed the sediment of leucocytes once with saline. 
The sediment, consisting of serum-free leucocytes, was suspended 
in 0.9 per cent saline, and portions were placed in very smal] test 
tubes and tested with the ordinary precipitating agents. 

I found that if a mere trace of a very weak solution of ferric 
chloride or copper sulphate (about 0.04 per cent) was added, very 
striking agglutination and precipitation of the leucocytes took place. 
Under the microscope, clumps of agglutinated and perfectly normal- 
looking leucocytes were seen. The precipitate of leucocytes settled 
very rapidly in the test tube, leaving a clear, supernatant liquid. Ifa 
slight excess of reagent was added, the reaction did not take place, or, 
if it had already taken place, the clumps were broken up and the 
normal suspension was resumed. 

With hydrochloric acid I could not at first obtain the reaction, but 
on repeating the experiment, using one drop of a 0.006 per cent 
hydrochloric acid, a typical reaction took place, the leucocytes becom- 
ing immediately agglutinated and settling rapidly to the bottom of the 
tube. 

These experiments will show that it is a very easy matter to miss 
the reaction if the proper quantities of reagents are not employed. 
Excess of precipitating agents produces a considerable degree of 
leucolysis. 

The investigation of the serum of this leukzemic blood gave rise to 
an observation which seems to show that there is an antagonism 
between agglutinins and hemolysins when acting in the same serum. 

The following experiment was performed: 


A416 S. Peskind. 


To one part of a ro per cent suspension of washed dog’s corpuscles was 
added an equal part of the leukaemic serum. A very feeble attempt at 
agglutination occurred, but this was so slight that on preparing a micro- 
scopic specimen the agglutination was completely broken up by the mere 
act of placing the cover slip on the drop of blood. The blood was not at 
all precipitated in the test tube. Ze blood-corpuscles laked rapidly, in 
five minutes, first exhibiting considerable swelling, just as occurs when 
corpuscles precipitated by acids andvacid salts are allowed to remain at 
room temperature. 

Thinking that the presence of the hamolysin in the leukaemic serum inter- 
fered with the action of the agglutinin, I made the following experiment : 

To the same washed dog’s corpuscles was added an equal portion of the serum 
from another dog. No agglutination or precipitation of the corpuscles 
occurred. ‘To this mixture (equal parts } c.c.) were added two drops of 
the leukeemic serum. There was immediately a good agglutination and 
precipitation of all the blood-corpuscles, which settled completely in five 
minutes. After twenty-four hours, absolutely no laking of the precipitated 
corpuscles had taken place. Apparently the second dog’s serum con- 
tained an antihzemoiysin which protected the corpuscles of the first dog 
from the action of the hamolysin contained in the leukemic serum. 
As soon as the latter hemolysin was neutralized, the agglutinin of the 
leukaemic serum had free play and produced a marked agglutination of 
the dog’s corpuscles. 

To two drops of leukaemic blood in 1 c.c. of saline was added one drop of 
ferric chloride. A fairly good agglutination of the corpuscles occurred. 
On the slide this agglutination was completely broken up by the pressure 
of the cover slip. ‘The corpuscles appeared enormously swollen. In the 
test tube at the same time the corpuscles were beginning to resume 
normal suspension, the agglutination becoming broken up. One more 
drop of ferric chloride was added. A heavy flocculent precipitate of 
corpuscles came down, the .corpuscles having resumed approximately 
normal size. The precipitate settled in a minute, leaving a_ perfectly 
clear and colorless supernatant liquid, showing no trace of laking. The 
lesson I would draw from this experiment is that the condition of the 
blood-corpuscles when they are swelled is antagonistic to their agglutina- 
tion. ‘This is seen also in the first experiment, where the hamolysin of 
the leukeemic serum also produced swelling of the dog’s corpuscles, and 
the agglutination of the corpuscles was almost entirely prevented. As 
soon as the effect of the hemolysin on the corpuscles was removed, then 
the agglutination took place readily. 

To one part of a ro per cent suspension of my own blood in o.g per cent 
saline was added one part of the leukamic serum. <A good agglutination 
and precipitation of the corpuscles occurred. After twenty minutes, one- 


Action of Acids and Acid Salts on the Blood. 417 


half part of leukaemic serum was,added. Laking started soon after and 
in an hour was almost complete. 


Red corpuscles and leucocytes are not the only cells affected by the 
reaction. It seems to bea very general one, being given by all the 
cells which I have hitherto tested — namely, spermatozoa, ciliated 
epithelium, yeast cells, a fungus, and a bacillus. 


Human spermatic fluid, possessing a strongly alkaline reaction, was sus- 
pended in saline. ‘To this a little ferric chloride solution was added. A 
precipitate came down consisting of masses of fine granules. On the 
further addition of ferric chloride, a flocculent precipitate appeared con- 
sisting of agglutinated masses of spermatozoa surrounded by the granular 
precipitate already mentioned. If the spermatic fluid was neutralized 
with hydrochloric acid before the ferric chloride was added, then practi- 
cally no granular precipitate was thrown down, but the spermatozoa alone 
were immediately agglutinated and came down as a flocculent precipitate. 
Under the microscope the spermatozoa were seen in clumps, a tangle of 
heads and tails, the motility of the spermatozoa being completely sup- 
pressed. Around these agglutinated masses of spermatozoa and in 
their midst were seen masses of the granular material that is present 
normally in spermatic fluid (detritus of spermatozoa, spermatic cor- 
puscles, etc.). 

Human spermatic fluid, diluted with saline, was treated to fractional cen- 
trifugalization, so that the first sediment contained principally the granular 
material, spermatic corpuscles, etc., while the supernatant fluid contained 
practically nothing but spermatozoa. This supernatant fluid was centri- 
fugalized, and the sediment, consisting almost entirely of spermatozoa, was 
washed with saline, and again centrifugalized. The sediment of sperma- 
tozoa thus obtained was suspended in o.9 per cent saline, and treated with 
various precipitants. 

Ferric chloride caused a marked agglutination and precipitation of the 
spermatozoa. I could not obtain the reaction with either hydrochloric 
acid or copper sulphate ; possibly 1 worked with too large amounts of 
these reagents. 

If uranium acetate is added to spermatic fluid suspended in saline, a heavy 
precipitate is thrown down. Under the microscope, it is seen that but 
very few of the spermatozoa have been caught and carried down by the 
precipitate. They retain their motility and lie free in the liquid. If the 
bulky precipitate produced by the uranium acetate is allowed to settle, 
the supernatant liquid containing most of the spermatozoa decanted, and 
to this fluid a drop of ferric chloride is added, all the spermatozoa 
are immediately aggiutinated and precipitated, showing conclusively that 


A18 S. Peskind. 


this is a specific reaction and is not caused by any extraneous precipitate 
carrying down the spermatozoa. 

Ciliated epithelium obtained by scraping the laryngeal mucosa of a rabbit 
was suspended in saline, and any blood present laked by a little sapotoxin. 
Ferric chloride gave a good agglutination and precipitation of the epithe- 
lial cells. I could not obtain the reaction with hydrochloric acid or 
copper sulphate. 

A piece of ordinary cake of dried yeast was rubbed up with o.g per cent 
sodium chloride solution, so that under the microscope the yeast cells 
were seen to lie separately, absolutely no clumps of cells being visible. 
The cells were immediately agglutinated and precipitated in flocculent 
form by a little ferric chloride. This precipitate consisted of agglutinated 
masses of well-preserved, normal-looking yeast cells, no extraneous pre- 
cipitate being visible. Centrifugalized and washed yeast cells (7. e., 
washed with saline) gave the same reaction. 

A small quantity of a mould was shaken up with saline. Under the micro- 
scope, mycelia and spores were seen, but no clumps to speak of. To 
this suspension a trace of ferric chloride was added. A flocculent pre- 
cipitate was thrown down, consisting of agglutinated masses of mycelia 
and spores. 

A strong solution of peptone was inoculated with a motile bacillus. After 
several days the solution was centrifugalized and a heavy sediment con- 
sisting entirely of bacilli was obtained. It was almost a pure culture of a 
short, very motile bacillus. This sediment was washed and _recen- 
trifugalized until free from the peptone solution. The bacilli were then 
shaken up in saline until all clumps of bacilli were broken up. Ferric 
chloride as well as hydrochloric acid caused the bacilli to lose their 
motility and immediately to become agglutinated and precipitated. 


UsEsS AND APPLICATIONS OF THE REACTION. 


This reaction may be put to the following uses: 

1. To obtain large quantities of blood-corpuscles. 

For this purpose the following procedure is recommended. 

All the operations are best done with the temperature of the solu- 
tions kept as near as possible to 0° C. 


Fresh defibrinated dog’s blood is diluted with ten parts of o.g per cent saline 
and the beaker with this mixture is cooled to o° C. Then, with constant 
stirring, a solution of o.2 per cent hydrochloric acid in saline, or of 
1.2 per cent ferric chloride in saline, is allowed to drop into the mixture 
from a burette. For every 5 c.c. of dog’s blood, 5 to 10 c.c. of the 
hydrochloric acid or 1.7 to 1.8 c.c. of the ferric chloride solution are 


Ee 


je Geen 


Action of Acids and Acid Salts on the Blood. 419 


required. ‘he corpuscles are rapidly precipitated and settle well, leaving 
a perfectly clear supernatant liquid, when hydrochloric acid is used as 
precipitant, while a slight amount of a granular precipitate is produced 
in the supernatant liquid if ferric chloride is employed. ‘This granular 
precipitate consists of a globulin. 

The supernatant liquid is then decanted, and the corpuscles are transferred to 
centrifuge tubes and quickly centrifugalized (for about two or three 
minutes). The supernatant liquid is then decanted or otherwise removed, 
ice-cold saline is added to the corpuscles in the tubes, the latter are well 
shaken up, and then once more centrifugalized. ‘This washing and 
centrifugalization is repeated three or four times. After being thor- 
oughly washed, the corpuscles may be suspended in saline, and may then 
remain at room temperature for many hours without the slightest laking. 

The granular precipitate produced by ferric chloride is removed in large part 
by the washings, which floats the precipitate off. Its presence may be 
avoided by using a mixture of ferric chloride and acids in precipitating, 
or by using a less proportion of ferric chloride than is given above. 

If quantitative experiments are intended, then it is best to place the measured 
blood from the first in the centrifuge tubes, diluting with ten volumes of 
saline, and adding the proportionate amounts of reagent. 

Ferric chloride is to be preferred as a precipitant, since hydrochloric acid has 
a great tendency to cause laking. By this process a large amount of 
blood-corpuscles, thoroughly free from serum, may be obtained in less 
than half an hour. 


2. To interrupt the action of all those haemolytic agents which are 
not affected by the precipitants. For instance, it is desired to know 
if sapotoxin is fixed by blood-corpuscles after acting on them for a 
certain length of time and at a definite temperature. After the de- 
sired length of time has elapsed the blood is cooled to 0°, which, I 
have found, arrests the action of the sapotoxin. The blood is then 
diluted a little with ice-cold saline solution, the corpuscles precipi- 
tated by hydrochloric acid or ferric chloride, and rapidly centrifugal- 
ized. The corpuscles are thus very quickly removed from the action 
of the sapotoxin, especially if the whole experiment be performed in 
centrifuge tubes. The supernatant liquid will then show if any 
laking has taken place, and the sediment of corpuscles, which may 
be washed with cold saline solution if desired, is allowed to stand 
suspended in saline. If the corpuscles lake, sapotoxin has been 
fixed. 

3. To determine at any given moment what percentage of cor- 
puscles has been laked by a hemolytic agent. At the desired moment 


420 S. Peskind. 


the corpuscles are precipitated and the amount of free haemoglobin in 
the supernatant liquid (which is obtained by centrifugalization) is 
determined colorimetrically. 

4. This reaction might perhaps be used for determining the alka- 
linity of the blood-serum. It is evident that there is a close connec- 
tion between agglutination and precipitation cf the corpuscles and 
the alkalinity of the blood. The serum remaining after the cor- 
puscles are exactly precipitated, is perfectly neutral in reaction. It 
is possible that the point where the corpuscles just begin to aggluti- 
nate is an indication that the alkalinity of the serum is neutralized. 


EXPLANATION OF THE REACTION. 


It has been shown above that the stromata are precipitated by the 
same reagents which precipitate and agglutinate intact blood-cor- 
puscles. Therefore the substance which combines with the precipi- 
tating agent must be a constituent of the stromata. According to 
Halliburton,! “cell globulin 8” is the principal proteid demonstrable 
in the stromata of red corpuscles, and this proteid seems to be 
common to leucocytes, liver cells, and apparently all typical cells. 
Later researches? have shown that this cytoglobulin is really a 
nucleoproteid. The nucleoproteids of blood-corpuscles, like the 
globulin of the serum, act like weak acids,and must be considered as 
existing in combination with alkalies.* The other constituents of 
the stromata were found by Woolridge to be lecithin and cholesterin. 

That the cholesterin and lecithin are not concerned in this reac- 
tion is shown by the fact that in ether-laked blood, where the choles- 
terin and lecithin are dissolved out from the stromata, the precipitation 
still occurs on addition of the reagents. We must, therefore, look to 
the proteid of the stroma for an explanation of the reaction. How 
does this substance react toward the acids and acid salts which 
produce agglutination and precipitation of blood-corpuscles? Nucleo- 
proteids combine with alkalies to form neutral or slightly acid solu- 
tions. From these solutions the nucleoproteid is precipitated by small 
quantities of most acids and acid salts. 

It seems probable therefore that the agglutination and precipitation 
of the red blood-corpuscles by acids and acid salts is due to a combi- 
nation of the latter with the nucleoproteid of the stromata. 


' HALLIBURTON and FRIEND: Journal of physiology, x, p. 537. 
* HALLIBURTON : Journal of physiology, 1895, xviii, p. 306. 
* Cf. HAMMARSTEN: Physiologische Chemie, p. 80. 


—— 


a 


a 


i i ie 


Action of Actds and Actd Salts on the Blood. 421 


That acid salts form an actual combination with nucleoproteid is 
seen from the fact that the smallest quantities of these salts form a 
precipitate in a neutral solution of the former. 

Why certain acids and acid salts will not precipitate intact cor- 
puscles and stromata, I cannot at present say. The reason of the 
inactivity of some of these substances probably lies in the fact that 
they are unable to precipitate nucleoproteids. For instance, I found 
that potassium acid tartrate, potassium and antimony tartrate, and 
calcium glycerino phosphate were unable to form a precipitate in a 
solution of sodium caseinate (casein being a typical nucleoalbumin). 
To explain the inactivity of other substances a clue may perhaps be 
found in the observation that at least some of the acids and acid salts 
which do not agglutinate and precipitate corpuscles and stromata are 
found in certain strengths to produce crenation of the corpuscles. 
This fact is well shown by arsenious and boracic acids. Those acids 
and acid salts which do agglutinate and precipitate corpuscles were 
never seen by the writer to cause crenation, but always produced a 
rounding and filling out of the contours of the corpuscles, even if 
they were previously crenated. This indicates that the active rea- 
gents penetrate the corpuscles readily, while the inactive ones pene- 
trate imperfectly. 

The question arises: how far do the precipitating agents penetrate 
into the corpuscles when they are added in just sufficient amounts to 
cause agglutination and precipitation? I believe that the reagents 
do not penetrate more than a slight distance, but remain close to the 
surface and combine with the surface layer or envelope of the cor- 
puscles. The envelope becomes sticky, thus accounting for the 
agglutination. The corpuscles increase in weight, owing to the en- 
trance of the reagent, and are therefore precipitated. 

My reasons for supposing that the reagents combine with the sur- 
face layer of the corpuscles, and do not penetrate for more than a 
slight depth, are the following: 

1. The quantity of reagent needed to produce agglutination and 
precipitation is very slight. If complete penetration of the corpuscles 
occurred, it is evident that all the alkali contained in the corpuscles 
would have to be neutralized. This would require much more reagent 
than is really necessary for precipitation. 

2. The fact that repeated washing of precipitated corpuscles re- 
stores in large part their normal suspension; also the fact that 
excess of reagent outside the corpuscles prevents agglutination and 


422 S. Peskind. 


precipitation. In the first case we must assume that the normal 
alkalinity of the surface layer or envelope has been to a great degree 
restored. In the latter case, the excessive acidity of the surface layer 
seems to prevent precipitation. Both these instances tend to show 
that the reaction of the surface layer of the corpuscles (2. ¢., the 
reaction toward litmus) is an important factor in their agglutination 
and precipitation. ; 

3. The hemoglobin contained in the corpuscles is not altered in 
the least, if only the exact amount of reagent necessary to agglutinate 
and precipitate the corpuscles is employed. Yet all the precipitating 
agents, acids and acid salts, in dilute solution convert oxyhamoglobin, 
first, into methemoglobin, and then into acid hzematin. 

If moderate excess of reagent be added, then the corpuscles will 
still be precipitated, but no methemoglobin band appears for five 
or ten minutes, depending on the amount of reagent employed. 
The same amount of reagent added to the same amount of sapotoxin 
or water-laked blood instantly converts part of the oxyhamoglobin 
into methemoglobin and this conversion proceeds very rapidly till 
complete. 

The following experiments may be quoted: 


1. 1c.c. of blood, diluted to 10 cc with saline, required 0.4 c.c. of ferric 
chloride solution (12 grams of ferric chloride to 1 litre of 0.9 per cent 
saline) in order that the corpuscles should be agglutinated and precipi- 
tated. The spectrum of the precipitated blood was that of oxyhzmo- 
globin, and it remained so for several hours. 

2. To 1 c.c. of the sapotoxin laked blood, diluted to 10 c.c. with saline, was 
added 0.4 c.c. of ferric chloride. The ghosts were precipitated; no 
methzmoglobin was formed. 

To 1 c.c. of blood, diluted to ro c.c. with saline, added was 2 c.c. of ferric 
chloride, a considerable excess over the required amount (0.4 c.c.) of 
reagent. The corpuscles were precipitated. I watched with the spectro- 
scope for the methemoglobin band. ‘The latter did not appear until eight 
minutes had elapsed. The corpuscles at the same time turned dark and 
started to lake. 


ww 


4. To 1 c.c. of sapotoxin laked blood, diluted to 10 c.c. with saline, was 
added 2 c.c. of ferric chloride. The methemoglobin band appeared 
instantly, and quickly became very strong. The color of the solution 
turned dark. 

5. To 5 c.c. of ferric chloride, diluted to to c.c. with saline, was added }§ c.c. 
of blood (therefore a large excess of reagent). ‘The corpuscles were pre- 
cipitated. ‘The methemoglobin band did not appear until five minutes 


Se 


ee se ee 


wai \ haiti 


Action of Acids and Acid Salts on the Blood. 423 


after the blood was mixed with the reagent. The blood then began to 
turn dark and to lake. 

6. To 1 c.c. blood, laked with water and diluted with water to 10 c.c., was 
added 0.4 c.c. of ferric chloride. The stromata were all agglutinated and 
precipitated, while the color of the solution remained bright red and showed 
only oxyhzmoglobin bands and no methemoglobin. ‘Then 1.6 c.c. more 
of ferric chloride was added. Immediately a methzemoglobin band ap- 
peared and the blood turned dark brown. ‘This seems to show that the 
reagents combine first with the stromata, and only after more than enough 
has been added to precipitate the stromata is the hamoglobin attacked. 
In other words, the stromata protect the hamoglobin from the action of a 
certain amount of reagent. 


From the above experiments it is evident that in the intact cor- 
puscles the protection of the oxyhzmoglobin is much more complete 
than in laked blood. This protection against the action of reagents, 
if it is of a chemical nature, might be afforded by: (1) the alkali of 
the corpuscles; (2) the stroma substance; (3) the firm combination 
of the hemoglobin with the stroma. 

That neither the alkali nor the stroma substance shields the hamo- 
globin from the action of the reagents is shown by the fact that after 
laking the presence of stromata and of liberated alkalies protects 
oxyhzmoglobin but slightly, and not at all if excess of the reagent be 
present (compare Experiments 1-6). Moreover there are many facts 
which suggest that the haemoglobin compound within the corpuscles 
is in all probability a weak one. So that it seems most likely that 
there is something in the structure of the corpuscles, some mechanical 
barrier, which shields the hemoglobin from the immediate action of 
the reagents, in other words, an envelope. 

In using the term envelope I merely suggest that at the surface 
of the corpuscle there exists a hemoglobin-free layer of stroma, but 
not necessarily a differentiated membrane. In order to explain all 
the known facts, this envelope must be supposed to contain nucleo- 
proteid, chlosterin, and lecithin. 

That the hemoglobin of mammalian corpuscles does not extend to 
the very periphery, but is separated from the latter by a thin hemo- 
globin-free layer, seems probable from the results obtained by means 
of a reaction in which, by the addition of hydroxylamine hydro- 
chlorate to blood, bubbles of gas are caused to form just within the 
periphery of the corpuscles. These bubbles are lined externally by 
a delicate membrane (a, Fig. 1) and internally by haemoglobin (4). 


424 S. Peskind. 


The membrane is hardly stained by any of those dyes which stain 
the haemoglobin in the rest of the corpuscles very deeply. This 
indicates that it does not contain haemoglobin. It appears colorless 
in the unstained corpuscle. This might be because the color of a 
very thin layer of haemoglobin might escape detection. 


€: © ; S&S ~Q 
Fig. 2 Fig. 3 Fig. 4 
© @ @ 
fi : ; $4 nS & ie) » 
Fig Fig. 6 Fig. 7 


Figures showing bubble formation in blood-corpuscles treated with 
hydroxylamine hydrochlorate. 


Only by observing the corpuscles during the formative stage of the 
bubbles is it possible to make out that they are not formed within 
the hemoglobin at all, but that the latter merely forms their inner 
boundary, 

About ten minutes after the reagent and blood are mixed, the first 
sign of the intracorpuscular bubble formation becomes evident ; 
it consists in a flattening or depression of the haemoglobin on the 
periphery of a corpuscle, the outer lining of the bubble being too 
close to the hamoglobin at this stage to be visible (Fig. 2). 

The next stage is represented by Fig. 3, and shows some separa- 
tion between the surface layer of the corpuscle (a) and the hamo- 
globin. The completed bubble appears in Fig. 4, the outer lining 
(a) often projecting considerably above the surface. 


— 


a Y 


Action of Acids and Acid Salts on the Blood. 425 


The most striking part of the reaction lies in the depression of the 
hemoglobin from the surface of the corpuscles in the first stage. It 
demonstrates that the accumulation of gas takes place between an 
enveloping layer and the hemoglobin, causing the repulsion of the 
latter from surface of the corpuscle toward the centre and at the 
same time separating the envelope from the hemoglobin underneath. 
Although gas is formed immediately on mixing the reagents, yet 
about ten minutes elapse before bubbles are formed inside of the 
corpuscles. This could be explained on the hypothesis that the 
envelope must be hardened to a certain extent before it will be 
sufficiently unyielding to retain the bubbles within the corpuscles. 

Occasionally a burst bubble is noted, the lining of which is seen 
as in Fig. 6. In order that others may be able to verify these 
statements, I shall give a detailed account of the procedure by which 
the above results were obtained. 


The reagent is a saturated solution of hydroxylamine hydrochlorate in 0.9 per 
cent saline solution. I usually prepare this solution just before use, as 
the temperature of a freshly prepared solution is much lower than the 
temperature of the room, and this may have some bearing on the obtain- 
ing of a successful result. 

The blood used was defibrinated dog’s blood, although human blood 
also showed good intracorpuscular bubble formation. 

To one part of blood in a test tube is added an exactly equal part of 
the reagent, and the tube is quickly shaken to insure intimate mixture. 
The color of the blood turns dark brown almost immediately. Bubbles 
of gas quickly begin to rise to the surface, and soon the mixture becomes 
so full of bubbles as to be almost solid, owing to their presence. Under 
the microscope, the corpuscles appear somewhat swelled, but the outlines 
are for the most part smooth. Only after ten minutes does intracor- 
puscular bubble formation become evident. After a few hours this is 
complete. 

If examined after twenty-four hours, the hydroxylamine corpuscles are 
found to be thoroughly hardened, so that neither sapotoxin, water, acids, 
nor alkalies will cause the slightest laking, although the last three cause 
great swelling of the corpuscles. The bubbles are especially well brought 
out by adding a drop or two of ammonia (1o per cent) to some corpus- 
cles on a slide. A trace of methylene blue or gentian violet or methyl] 
green, added after the ammonia, brings out in striking contrast the practi- 
cally unstained bubble lining and the deeply stained corpuscles. 

In a successful specimen most of the corpuscles should contain one or 
more bubbles. Many contain a whole peripheral ring of bubbles (Fig. 5). 


426 


S. Peskind. 


Efforts were unsuccessfully made to produce a differential stain of the 
membrane-like film covering the bubbles. 

The exact proportion and method of mixing given above must be used. 
If less of the reagent is employed, then the corpuscles will appear laked, 
the hemoglobin being precipitated in the form of granules on top of and 
around them. Hydroxylamine hydrochlorate precipitates haemoglobin 
completely from solution. If more than the proper proportion of re- 
agent is used, then the corpuscles show absolutely no intracorpuscular 
bubble formation, although gas comes off from the liquid. ‘The age of 
the blood is a very important factor in the success of the reaction. With 
perfectly fresh defibrinated dog’s blood a comparatively slight amount of 
intracorpuscular bubble formation took place. The same blood, when 
thirty hours old, was treated with the reagent, and in the majority of cor- 
puscles one or more bubbles were formed. ‘This blood, when sixty hours 
old (having been kept in the ice-chest in a beaker covered with filter 
paper), was still of a bright red color (therefore not reduced), and gave 
rise to a considerable number of bubble corpuscles when hydroxylamine 
hydrochlorate was added. 

The blood was then placed on ice in a stoppered test tube. After 
twelve hours it was still bright red. The blood was then treated with 
hydroxylamine hydrochlorate, and almost no bubbles were formed. The 
corpuscles, after the addition of the reagent, looked riddled, as though a 
large number of bubbles had escaped from the corpuscles and had left 
holes in them or their envelopes, as in Fig. 7. 

Evidently some great change had occurred in the corpuscles during 
the last twelve hours. Before this time many bubbles were retained 
within the corpuscles after treatment with the reagent. Twelve hours 
later the corpuscles were unable to retain any gas bubbles whatever. This 
is easily explainable, on the hypothesis that the envelopes of the corpus- 
cles had deteriorated and had become weaker during those few hours, so 
that a too rapid penetration of the corpuscles by the reagent took place, 
giving rise to a sudden and great evolution of gas bubbles which the weak- 
ened envelopes were unable to retain ; as the envelope hardened simul- 
taneously with the escape of the bubbles, the rents in the former remained 
permanently open, thus giving rise to the perforated appearance. 


Blood six days old, kept stoppered on ice, and whose hemoglobin appeared 


to be thoroughly reduced, yielded no intracorpuscular bubbles, although 
a large amount of gas was given off, the corpuscles having the riddled 
appearance mentioned above. ‘Through this blood oxygen was passed. 
On the addition of hydroxylamine hydrochlorate less gas was given off, 
but no intracorpuscular bubble formation took place. Illuminating gas 
was then passed through this blood ; also no intraglobular bubble forma- 
tion on addition of the reagent. 


Action of Acids and Acid Salts on the Blood. 427 


Ordinary fresh blood, mixed with equal parts of hydroxylamine hydrochlorate 
solution, was placed in a thick-walled test tube which was connected with 
a suction pump. Bubbles formed in the corpuscles just as usual, the 
suction having failed to remove them from the corpuscles. 

I was unsuccessful in modifying the usual result by carrying out the reaction 
under pressure. Fresh blood, mixed with equal parts of reagent, was ex- 
posed to considerable air pressure, which was released after a few minutes ; 
no unusual effect was produced in the corpuscles. 

If laked blood is treated with a little hydroxylamine hydrochlorate, the stromata, 
together with the hemoglobin, are precipitated. 

In attempting to determine the source of the gas, the following experiment 
was mace : 

Dog’s serum mixed with an equal part of the hydroxylamine hydro- 
chlorate solution gave off practically no gas. On adding to this mixture 
a little blood from the same dog, a large amount of gas was evolved 
and the test tube became warm. ‘This shows that the gas is formed 
entirely by the interaction of the corpuscles and the reagent. 

What does this gas consist of and how is it generated ? 

Hydroxylamine hydrochlorate is strongly acid in reaction, therefore it 
seems probable that some carbon dioxide is driven off from the corpuscles 
by the reagent. Moreover hydroxylamine hydrochlorate is easily decom- 
posed, giving off nitrogen gas. Also if a saturated solution of hydroxyl- 
amine hydrochlorate is treated with a little caustic soda, nitrogen is 
evolved. Does the reaction take place between the oxygen of the 
oxyhemoglobin and the hydroxylamine? Or between the alkali of the 
corpuscles and the reagent? Or is the reaction due to a catalytic action 
on the hydroxylamine produced by some substance of the stroma? 
Experiments are being made to determine these points. 

An analysis of the gas given off in the above reaction was made as follows : 

In a test tube was poured 8-ro c.c. of mercury, and upon this 15.5 c.c. 
of blood. Then 4 c.c. of a saturated solution of the reagent was added, 
this being just enough to fill the tube perfectly. The few bubbles on top 
of the liquid were quickly removed with filter paper, the test tube closed 
with the finger, inverted, and placed over mercury. Gas began to form 
immediately. After twelve hours the mixture of blood and reagent was 
almost solid, and contained bubbles throughout its whole extent. The 
tube was then placed in a large dish containing water, then an aluminium 
probe, around the end of which was wrapped a small piece of cotton 
which had been wetted to remove air, was passed inside of the test tube 
and worked up and down, the tube being kept quite vertical throughout 
this procedure’ The bubbles were thus released and rose to the top, 
while the other contents of the tube were washed out by the surrounding 
water. 


428 S. Peskind. 


After all the blood and reagent were removed from the tube, and the 
gas had all collected at the top, the test tube was again placed over 
mercury. By suitable means, most of the water in the tube was removed 
and mercury drawn up in its place. Caustic soda solution was then 
introduced by means of a pipette with a recurved tip. Some of the gas 
was absorbed = carbon dioxide. ‘Then a strong pyrogallic acid solution 
was allowed to enter, and the tube was agitated. More gas was absorbed 
= oxygen. The greater part of the gas remained unabsorbed after six 
hours = nitrogen. I assume this residue to be nitrogen, because it is the 
only inert gas which could be given off in this reaction. 

Roughly the following measurements were obtained : 

15.5 c.c. of blood, mixed with 4 c.c. of a saturated solution of hydroxyl- 
amine hydrochlorate, gave rise to 5 c.c. of gas which consisted of 

3-6 c.c. nitrogen, 
0.8 C.c. oxygen, 
0.6 c.c. carbon dioxide. 

It is evident that only a small part of the oxygen in the blood was 
given off. What became of the remainder? If we assume that it com- 
bined with the reagent after the following reaction : 


4NH,OH _ ©; = 2Ne, oL 6H,O, 


then we would expect to find two volumes of nitrogen given off for every 
volume of oxygen fixed. This is approximately what was found to be the 
case in the rough quantitative experiment detailed above. However, 
more experiments, exactly performed, will be required to determine what 
actually takes place during the interaction of the reagent and blood- 
corpuscles. 


The behavior of corpuscles of various ages towards hydroxylamine 
hydrochlorate, as well as their behavior toward various strengths of 
the latter reagent, speaks strongly for the presence of an envelope.! 
Evidence has already been found that the surface layer of the 
corpuscles consists in part of nucleoproteid. It seems also to con- 
tain cholesterin, and probably lecithin, for reasons which will be given, 
elsewhere, in a conjoint paper by Dr. G. N. Stewart and myself. 


CONCLUSIONS. 


As most of the conclusions of this paper were summarized ina 
preliminary note in a previous issue of this Journal (Volume VIII, 


' H. DEETJEN (in VircHow’s Archiv fiir experimentelle Pathologie, 1901, clxv, 
p. 232) gives a staining reaction whereby, according to him, the envelopes of red 
corpuscles are easily made apparent; I have not seen his statement either cor- 
roborated or refuted. 


__*; = 


SS 


Action of Acids and Acid Salts on the Blood. 429 


page 99), I shall content myself*by giving here only those which 
were not stated therein. 

1. The precipitating agents do not penetrate far into the corpuscle 
during the reaction, but remain close to the surface and combine with 
the surface layer. 

2. Theoretical considerations, as well as chemical and histological 
facts brought out in this paper, render the existence of an envelope in 
mammalian blood-corpuscles highly probable, if not absolutely certain. 
To explain all the facts, such an envelope must be assumed to con- 
tain nucleoproteid, cholesterin, and lecithin. 

Since leucocytes and probably all typical cells possess the above- 
mentioned constituents, it seems likely that these substances serve 
the same purpose in the structure of all these cells, — namely, that of 
forming an envelope. 


In conclusion I wish to express my thanks to Dr. G. N. Stewart for 
the kind interest he has taken in my work, and for the many valuable 
suggestions that I have received from him. 


HOW LONG DOES (ARBACIA) SPERM LIVE IN 
SEA-WATER? 


By MARTIN H. FISCHER. 
[From the Physiological Laboratory of the University of California. 


N the course of our experiments at Wood’s Holl, two questions had 
often arisen: first, How long after removing the sperm from the 
testicles of a marine animal do the spermatozoa retain their power of 
fertilization if kept in sea-water ; and secondly,-if fertilization occurs 
when the spermatozoa are dying, will the fertilized eggs develop into 
adult animals or only pass through the first few stages of segmenta- 
tion? It was with a view to answering these questions that, at the 
suggestion of Dr. Loeb, I undertook the following experiments upon 
sea-urchins (Arbacia). 

It has been a generally accepted fact that the sperm of marine 
animals lives but a few hours after being shed into sea-water. 
Gemmill,! for example, found that the spermatozoa of sea-urchins live 
comparatively few hours after being shed into sea-water. When 
small amounts of sperm are added to large amounts of sea-water, 
they live from three to five hours; if concentrated solutions of the 
sperm are made in sea-water (one of sperm to ten of sea-water) they 
at times live as long as seventy-two hours. That he obtained no 
better results is probably due to his failure to observe any aseptic 
precautions in his experiments. That the spermatozoa do not live 
as long in a Jarge amount of sea-water as in a small amount, Gemmill 
attributes (1) to the “stimulating” effects of the sca-water upon the 
movements of the spermatozoa, and (2) to.a dilution of the sper- 
matic fluid which is supposed to possess nutritive properties. The 
role of bacterial development, the surface action of the glass, the 
increased viscosity of the sea-water when a Jarge amount of sperm is 
added, the lack of oxygen, etc., are entirely neglected. 

All who have worked with the sperm of sea animals know that 
when it is kept in a dish of sea-water under ordinary conditions, it is 


' GEMMILL: Journal of anatomy and physiology, 1900, xxxiv, p. 163. 
430 


sei bie alll 


Flow long does (Arbacia) Sperm live in Sea-water? 43 


not long before the whole is converted into an ill-smelling, putrid 
mass, having no powers of fertilization. A large number of sperma- 
tozoa die soon after their removal from the animal body, and these, 
together with the blood and débris which may accidentally have been 
introduced into the dish containing the sperm, form an excellent 
culture ground for bacteria. The poisonous products formed by the 
bacteria soon end the life of the spermatozoa that may at first have 
survived, 

In all the experiments described in this paper the following gen- 
eral methods were followed. The male sea-urchins after having been 
kept in running sea-water for some time were scrubbed for several 
minutes in running fresh water. They were then laid upon their 
backs and allowed to drain, while the experimenter carefully washed 
his hands in fresh water and sterilized by heat all the instruments 
that were to come in contact with the animal. The ventral surface 
of the sea-urchin was then removed with sterile scissors, and the 
abdominal cavity exposed. The testicles were removed with sterile 
forceps, washed in boiled sea-water contained in sterilized finger 
bowls to remove any blood or débris that might adhere to them, and 
transferred to a second sterilized finger bowl containing sterile sea- 
water. The sperm issues from the testicles in a dense, syrupy cloud. 
This I removed with sterile pipettes to sterile flasks containing sterile 
sea-water and stoppered with sterile cotton plugs. 

Only sea-water sterilized by bringing it to the boiling point and then 
allowing it to cool off slowly in cotton-stoppered flasks was used in the 
experiments. To recharge it with oxygen, it was thoroughly shaken 
in the air. All the glassware used in the experiments was sterilized 
in an Arnold steam sterilizer for two or more hours. The flasks 
used for keeping the sperm after removal from the body were stop- 
pered with cotton. The steel instruments were sterilized by dry heat. 

The fertilizing power of the sperm at the beginning and during the 
course of the experiments was tested by adding the sperm to freshly 
obtained sea-urchin eggs. The eggs were obtained with the same 
careful precautions with regard to asepsis and to the possibility of in- 
fection with sperm which were adopted in the method for obtaining 
the spermatozoa in a sterile condition. A control experiment in 
each case served as a check upon the result obtained. The fertilizing 
power of the sperm was tested every few hours, until segmentation 
no longer followed the addition of the sperm to the eggs. Fresh 
eggs, capable of being fertilized —as shown by adding fresh sperm 


432 Martin H. Fischer. 


to them —were used in every instance. When a negative result 
was obtained, the test was repeated two or three times to corroborate 
the result. In each case the last successful fertilization was taken to 
indicate the end of the experiment. 

The following is a synopsis of the experiments performed : 


Experiment I, July 6, 1902. — 9.35 ?. M. Removed sperm from the testicles 
of two sea-urchins, and kept it in two 250 c.c. Florentine flasks, half full 
of sea-water. Last successful fertilization July 9, 1902, 5.00 P. M. 

Time: 67 hrs., 25 min. 

Experiment If, July 7, 1902. — 9.30 A.M. Removed sperm from the testicles 
of one sea-urchin, and kept it in one 500 c.c. Florentine flask, half full of 
sea-water. Last successful fertilization July 9, 1902, 5.00 P. M. 

Time: 55 hrs., 30 min. 

Experiment [11, July 7, 1902. — 3.10 P.M. Sperm from one sea-urchin was 
kept in one 250 c.c. Erlenmeyer flask containing 100 c.c. of sea-water. 
Last successful fertilization July 10, 1902, 5.00 P. M. 

Time: 73 hrs., 50 min. 

Experiment IV, July 7, 1902.— 5.45 P.M. Sperm from one sea-urchin was 
kept in two 250 c.c. Erlenmeyer flasks, each containing 100 c.c. of sea- 
water. last successful fertilization July 12, 1902, 7.00 P. M. 

Time: ten. hrs... 15min, 

Experiment V, July 12, 1902.— 5.00 P.M. Removed sperm from testicles 
of three sea-urchins, and distributed it in varying amounts into ten 50 ¢.c. 
Erlenmeyer flasks, each containing 20 c.c. of sea-water. Last successful 
fertilization July 15, 1902, 12.00 noon. 

Time: 67 hrs. 

Experiment VI, July 17, 1902.— 10.00 A.M. Removed sperm from the 
testicles of three sea-urchins, and distributed it into ten 50 c.c. Erlenmeyer 
flasks, each containing 25 c.c. sea-water. Last successful fertilization 
July 18, 1902, 10.15 A. M. 

Time: 24 hrs., 15 min. 

Experiment VII, July 18, 1902. — 9.30 A.M. Removed sperm from one male 
sea-urchin, and distributed it into ten 50 c.c. Erlenmeyer flasks, contain- 
ing from 15-30 c.c. sea-water. I was compelled to leave Wood’s Holl, 
before the completion of this experiment. I still obtained blastulz from 
a dish of eggs fertilized July 23, 1902, 11.00 P. M. 

Time: at least 133 hrs., 30 min. 

Experiment VITT, July 19, 1902,— 11.30 A.M. Removed the sperm from 
one sea-urchin, and kept it in ten Erlenmeyer flasks. Last successful 
fertilization July 21, 1902, 3.00 P. M. 

Time: 27 hrs., 30 min. 


eee 


EE — eC 


ees t—™ 


Flow long does (Arbacia) Sperm live in Sea-water? 433 


Experiment IX, July 20, 1902. — 10%@0 A.M. Removed the sperm from one 
sea-urchin, and distributed it into five Erlenmeyer flasks. Last success- 
ful fertilization July 21, 1902, 10.00 A.M. 

Time: 24 hrs. 

Experiment X, July 20, 1902. — 11.20 P.M. Removed sperm from one sea- 
urchin, and distributed it into five Erlenmeyer flasks. Last successful 
fertilization July 23, 1902, 11.00 A. M. 

Time: 59 hrs., 40 min. 

Experiment XT, July 21, 1902. — 10.15 A.M. Distributed sperm from one 
male into five Erlenmeyer flasks. Last successful fertilization July 23. 
1902, 11.00 A. M. 

Time: 48 hrs., 45 min. 

Experiment XT, july 21, 1902.— 2.45 P.M. Distributed sperm from one 
male into five Erlenmeyer flasks. I was compelled to leave Wood’s Holl 
before the completion of this experiment. I still obtained blastula from 
a dish of eggs fertilized July 23, 1902, 11.00 P. M. 

Time: at least 56 hrs.,-15 min. 


Sufficient sperm was put into each of the flasks to give them a 
milky color, and the flask was always shaken to distribute the sper- 
matozoa evenly through the sea-water. Within certain limits, the 
amount of sperm did not, however, affect the general result. I 
thought that a large amount of sperm in a small amount of sea- 
water might easily suffer.from a lack of oxygen, but even when 
enough sperm was added to the sea-water to make it distinctly 
white in color, the result was quite as good as when these conditions 
were reversed. When a very small amount of sperm is used, it is 
soon precipitated on the sides and bottom of the glass vessel. We 
seem to deal here with a surface action of the glass. 

The spermatozoa sometimes die very suddenly, and at other times 
more slowly. When the latter occurs, equal quantities of sperm — 
as measured with the pipette— from the same flask will fertilize a 
less and less number of eggs, as time goes on. 

The above experiments show that the spermatozoa of the sea- 
urchin (Arbacia) live from forty-eight to at least one hundred thirty- 
three and a half hours after being shed into sea-water. What 
determines these individual differences is reserved for further study. 

In answer to the second question raised above, it must be said 
that whenever fertilization occurs the fertilized egg developes to the 
adult stage. In all the experiments described above the test dishes 
were kept for several days, and never was a segmented egg found 


A3A Martin FI. Fischer. 


that did not develop into a pluteus. Several times the developing 
eggs were traced through every stage of their segmentation, the 
sperm being tested at intervals of one hour, but never was a different 
result obtained. In only one instance (Exp. VIII) did I find a 
number of irregularly segmented and dead eggs in one of the test 
dishes beside the swimming blastule. In the former the pigment 
was collected in the centre of the eggs, and I feel confident that 
some foreign matter (probably iron from the steam sterilizer) got 
into the dish and poisoned the eggs. 

The spermatozoa seem to retain their power of fertilization as long 
as they retain their power of ciliary motion. In every case in which 
I was able to fertilize eggs, actively moving spermatozoa were found; 
and I did not find motile spermatozoa in sperm that could no longer 
fertilize, even though the heads were still intact. These facts would 
seem to indicate that the essential function of the spermatozoon in 
fertilization is to carry something into the egg that leads to the 
successive changes in it which are termed development. 


ae a) i ie a 


VARIATIONS IN THE AMPLITUDE OF THE CONTRAC- 
TIONS OF HUMAN VOLUNTARY MUSCLE IN 
RESPONSE TO GRADED VARIATIONS 
IN THE STRENGTH OF THE 
INDUCED SHOCK. 


By THOMAS ANDREW STOREY. 


[From the Physiological Laboratory of Stanford University.] 


| dealing with the electrical excitation of human voluntary muscle, 

one of the first points to be decided is where to place the 
secondary coil in its relation to the primary coil in order to gain the 
most efficient strength of the induced shock. One finds, on looking 
over the literature bearing upon the influence of electrical excitation 
upon the contraction of voluntary muscle, that this problem has usually 
been met by simply using the strongest current that could be endured. 
The secondary coil has been placed as far over the primary as the 
pain produced by the induced current would permit. 

The writer noticed in an earlier set of experiments that he got 
better contractions from human voluntary muscles with a strength of 
the single induced shock that was not painful. That experience led 
to the investigation upon which this paper is based. 


APPARATUS. 


The myograms obtained in these experiments were made with an 
ergograph devised by the writer in the Physiological Laboratory of 
the University of Michigan in 1900. The ergograph and its manipu- 
lation have been described elsewhere.’ 

An electric pendulum was used for the control of the primary. 
current. It was so arranged as to make it possible to cut out either 
the make or the break shock. The keys were platinum points that 
were dipped into cups of mercury by each vibration of the pendulum. 
A stream of water was passed over each mercury meniscus in order 
to wash away the electrolytic products that would otherwise accumu- 
late there. 

1 StorEY: This journal, 1903, viii, p. 355. 
435 


436 Thomas Andrew Storey. 


In those experiments where tetanizing currents of short duration 
were employed, the duration of flow of the current through the 
primary coil was regulated by the pendulum which closed that cur- 
rent during a certain part of~each vibration. The interruptor on the 
primary coil was kept vibrating by an independent current. 

The inductorium used was an upright form with 10,340 turns on 
the secondary coil. 

The left abductor indicis muscle was used throughout. 

The make shock was never sufficiently abrupt to cause a visible 
contraction, so that all the contractions in these experiments may be 
regarded as having been excited by the break shock. In no case was 
the shock strong enough to cause pain. 


EXPERIMENTS. 


Experiment 7. —'Vhis experiment was performed upon Mr. L. on February 17, 
1902, at5 P.M. ‘The resistance to be overcome in contraction was that 
of a light rubber band. Four Edison-Lalande cells furnished the primary 
current. Each stimulation was caused by a single induced shock, and the 


FIGURE l. 


muscle was excited once each second for five seconds, after which there 
followed a rest for five seconds. The position of the secondary coil was 
shifted one centimetre nearer the primary coil during each rest period. 
The part of the curve labelled ‘‘ A’ was made with the kathode over the 
muscle, and the part “B” was made with the anode over the muscle." 


It will be noted on looking over Fig. 1 that, with the kathode over 
the muscle, the induced shock was able to excite a visible contraction 
when the secondary coil was placed at a distance of 11 cm. from the 
primary coil. The height of the contractions is greater at 10 cm., 
and continues to increase till the secondary is placed at 8 and 7 cm. 
when a maximal height seems to have been reached. From this 
position to that at 4 and 3 cm. there is a decrease in height at each 
position of the secondary. Then as the secondary is drawn on over 
the primary coil each centimetre of change produces an increase 


' All figures in this paper are to be read from left to right. 


——EEaEoo 


Contraction of Human Voluntary Muscle. 437 


in the height of the contractioms excited by the induced shocks. 
This is the second period of maximum contractions. 

With the anode over the muscle the induced shock is not able to 
excite a contraction until the secondary. coil is placed at 10 cm. dis- 
tance. The successive changes in the position of the secondary coil 
produce the same relative changes in the height of the contractions 
that were noted with the kathode over the muscle. The second 
period of increasing heights of contractions arrives sooner in this 
case than with the kathode over the muscle. 


FIGURE 2. 


Experiment [7. — The writer was the subject in this experiment. It was done 
on the 25th of March, 1902, at3 P.M. The resistance to be overcome 
was a weight of 100 grams. Three cells were placed on the primary cir- 
cuit. Each stimulation was from a tetanizing current of one-tenth of a 
second in duration, and was formed of six interruptions. Each vertical 
record was made by six superimposed myograms made at intervals of one 
second. A rest of ten seconds followed each series of six contractions. 
The secondary coil was shifted one centimetre during each period of rest. 
Part ‘* A” of the figure below was done with the kathode over the 
muscle, and “ B” with the anode over the muscle. 


It is apparent that the same relations are present in Fig. 2 that were 
noted in Fig. 1. As the secondary coil is drawn down over the pri- 
mary coil and the strength of the induced shock is thereby increased 
there is first an increase in the height of contraction as each centimetre 
of change is made; then, after reaching a maximal height, the contrac- 
tions decrease in height as each centimetre of change is made; after 
which comes another, the final, period of increasing heights of the 
contractions, producing a second maximum. As the secondary coil is 
shifted away from the primary, the same relations appear again. 


438 Thomas Andrew Storey. 


Experiment [1/f.— This experiment was performed upon Mr. A. B. S. on 
March 25th, at 5 P.M. ‘The only difference between the conditions in 
this experiment and those of the preceding experiment is that four Edison- 
Lalande cells were used on the primary current instead of three. 


A 
SE a A ea PS eT 
Il cm. fo) fe) Il om. 
Bb 
Il CM. oO 10) 12 cm. 
FIGURE 3. 


Fig. 3 shows the same relations that have been pointed out in Figs. 

1 and 2, and the same description may be made for it that has been 

made for them. 

Experiment 1V was made upon the writer, 
and was done on the 1st of March, 1902, 
at 4 P.M. There were three Edison- 
Lalande cells in the primary circuit. 
Each stimulation was formed of sixty 
interruptions of the primary current, and 
occupied a period of one second’s dura- 
tion. ‘The kathode was over the muscle 

during the whole experiment. A rest of ten seconds followed each tetanic 

excitation. During this period of rest the secondary coil was shifted, as 
in the preceding experiments. ‘The resistance to be overcome by the 
muscle was formed by a strong spiral spring. 


Il om. 13) 10 cM. 


FIGURE 4. 


The relations of the various parts of this figure to each other and 
to the variations in the strength of the exciting current are the same 
as were noted in the preceding figures. 


i pe A ed Die 


Contraction of Human Voluntary Muscle. 439 


CONCLUSIONS. 


The writer has performed a number of experiments like the above 
and has found the same characteristic relationships in all of them. 
This relationship may be stated generally as follows: At some 
position of the secondary coil, the excitation by the induced shock 
will cause only a faint quiver of the muscle. Increases in the strength 
of the exciting shock causes corresponding increases in the heights of 
the muscular contractions. Soon, however, additional increases in the 
strength of the induced shock fail to cause proportional increases in 
the height of the contractions. Then further increases in the strength 
of the induced shock produce contractions of decreasing heights. 
Finally still greater increases in the strength of the induced shock 
cause a second period of contractions of increasing heights. There 
is, then, with graded increases in the strength of the induced shock: 
first, a period of increasing height of contractions; second, a period of 
maximal contractions; third, a period of decreasing height of con- 
tractions; fourth, a period of increasing height of contractions; and 
finally, a second period of maximal contractions. 

In these experiments the same results were obtained with either 
the anode or the kathode applied directly to the muscle. However, 
the physiological electrotonic conditions present must have been 
complex, so that one is not justified in inferring from this last stated 
fact that electrotonic influences have nothing to do with the effects 
of graded changes in the strength of the exciting current. 


LITERATURE. 


So far as the writer is aware this phenomenon has not been 
observed before with human voluntary muscle. It has been noted 
by several investigators who have worked with the muscle-nerve 
preparation from the frog! These writers have described their 
observations under the titles of ‘‘ Die Luicke,” ‘‘ Die Intervalle,” and 
‘“The Gap,” these terms referring to the interval of lower contrac- 


1 The literature bearing upon the Gap (Die Liicke) has been considered ex- 
haustively by BIEDERMANN, Electrophysiologie, Jena, 1895, ii, pp. 624, 648, and 
by HERMANN, Handbuch der Physiologie, 1879, ii, p. 107. In later years the 
subject has been briefly considered by Locke, Journal of physiology, 1900, xxvi, 
p. xxix, and by ROSENTHAL, Allgemeine Physiologie der Muskeln und Nerven, 


Leipzig, 1899, p. 303. 


440 Thomas Andrew Storey. 


tions occurring between the periods of the first and second maximal 
contractions. The phenomenon has appeared in response to changes 
in the strength or to changes in the duration of the galvanic current 
and to changes in the strength of the induced current such as have 
been described in this paper. In these investigations the electrodes 
have been placed upon the nerve trunk and the effects of ascending 
and descending currents have been compared. One set of investi- 
gators has found that the phenomenon occurs with ascending 
currents and not with descending currents. This has given rise to 
the supposition that with the ascending current and a slight or 
short stimulation, the anodic block is not present and consequently 
the impulse started by the kathodic excitation reaches the muscle. 
With graded increases in the strength or the duration of the current 
there is for a time a corresponding increase in the heights of the 
contractions evoked. Soon, however, the anodic block begins to 
interfere with the passage of the impulse started at the kathode. This 
block may at a certain stage be sufficient to bar out the influence of 
the kathodic excitation. The muscle then ceases to contract. Later, 
with stronger or with longer stimulations, the anodic rebound or dis- 
appearance of anelectrotonus acts asa stimulus and the muscle is 
again excited to contraction. The first period of maximal contrac- 
tions is due to kathodic excitation ; the interval of lower contractions 
is due to the interference of the anodic block ; and the second period 
of maximal contractions is due to excitation from the anodic 
rebound. 

Other investigators, however, have found that this phenomenon does 
appear with descending currents. This would indicate the presence 
of some influence other than anelectrotonus. 


NT 


THE LAKING OF DRIED RED BLOOD-CORPUSCLES. 


By CHARLES CLAUDE GUTHRIE. 


[From the Physiological Laboratory, Western Reserve University.) 


AVING accidentally observed that chicken’s and human red 

blood-corpuscles, allowed to dry in the air at room temperature, 

are readily laked by 0.9 per cent NaCl solution, I mentioned the fact 

to Dr. G. N. Stewart, who told me that he had published an observa- 
tion to the same effect in regard to Necturus corpuscles.! 

He recommended me to make a more systematic investigation of 
the phenomenon; among other reasons, on account of its practical 
importance, in connection with the usual procedure for the examina- 
tion of blood-stains in medico-legal cases. 

I take this opportunity to express my thanks to him for his 
invaluable assistance. 

The action of a large number of reagents on the dried red corpus- 
cles, of man, the dog, cow, rabbit, chicken, and frog was studied. 

Briefly the procedure was this: Thin layers of blood were spread 
on glass slides, by drawing a film across the surface of one with 
the edge of another. These were then allowed to dry, either at 
room temperature or in an oven heated to various temperatures, after 
which a cover slip was applied and the slide examined with the 
microscope. A drop of the reagent was then placed on the slide at 
the margin of the cover slip and allowed to spread into the prepara- 
tion. The reaction was observed from beginning to end, through 
the microscope. 

Dried stains on cloth were also examined by scraping the cloth 
above a slide, thus causing the corpuscles to fall upon it, laying on 
a cover slip, and then applying the reagent in the manner above de- 
scribed. In some experiments the films of blood were dried in test 
tubes. 

The following is a summary of the results obtained : — 

Rapid laking is produced 

1. By addition of the animal’s own, or the serum of another animal 


1 STEWART: This journal, 1go2, viii, p. 117. 
441 


442 Charles Claude Guthrie. 


2. By addition of aqueous solutions of non-electrolytes and elec- 
trolytes (salts, acids, and bases)! except such as readily convert the 
hemoglobin into insoluble substances; ¢. g., strong solutions of 
sodium hydroxide. 

3. In general, rapidity of laking is in direct ratio to the strength 
of the solution, to a point approaching saturation, above which the 
ratio is reversed. 3 

4. Solutions of glycerine lake in proportion to the percentage of 
water they contain. 

5. Ethyl alcohol containing much more than 5 per cent of water 
produces an amount of laking, before the corpuscles are fixed, pro- 
portionate to-the percentage of water present. If it contains 50 per 
cent of water, practically complete laking occurs; if less than 5 per 
cent, the corpuscles are fixed without laking. 

6. Chloroform, ether, benzole, and xylol produce no laking, and if 
removed after a period of contact of from one to five minutes, the 
corpuscles lake as before, on addition of the solutions mentioned 
above. 

7. The laking action is not dependent on, or indeed, influenced, to 
any appreciable extent, by the dried serum constituents of the blood, 
since there is no difference when corpuscles, washed free from serum 
constituents by 0.9 per cent NaCl solution before drying, are 
employed. 

8. The laking action is not essentially affected by varying the 
temperature at which it is carried out, within the limits of 25° to 
Sees 

g. Thin films of blood or of washed corpuscles dried at room 
temperature and then heated in an oven for twenty minutes at 
temperatures up to 100° C. are laked by the reagents just as if they 
had not been heated. 

It is known that oxyhzemoglobin dried completely below o° C. can 
be heated to 100° C. without decomposition.” 

Dr. Stewart heated to 100° C. for two hours a specimen of dog’s 


' The non-electrolytes used were distilled water, ether, chloroform, chloral 


hydrate, benzole, ethyl alcohol, glycerine, phenol, formaldehyde, aniline, boracic 
acid, cane sugar, and glucose. 

The electrolytes were sodium chloride, sodium carbonate, sodium hydroxide, 
sodium sulphate, potassium dichromate, potassium ferrocyanide, ammonium chlo- 
ride, ammonium hydroxide, ammonium sulphate, mercuric chloride, hydrochloric 
acid, sulphuric acid, nitric acid, methylene blue, and eosin. 


* HopPpE-SEYLER’s Handbuch, 6th ed., p. 277. 


The Laking of Dried Red Blood-Corpuscles. 443 


oxyhzemoglobin which had beengkept dry in a closed desiccator, and 
which had been largely changed into methemoglobin, as shown by 
the spectroscope. It was still soluble to some extent in distilled 
water, although much less so than before heating, and the spectrum 
of the filtered solution was the same as before heating. It is obvious 
that the constituents of the colored corpuscles are not easily fixed by 
dry heat. As is well known, Ehrlich recommends that films be 
heated at 120° C. for two hours. Corpuscles are, of course, fixed at 
a much lower temperature if water be present; ¢.g., if they be poured, 
into boiling 0.9 per cent NaCl solution. There is then no effect pro- 
duced by any of the reagents which lake the dried corpuscles. 

10. The precise amount of water which must be removed from the 
corpuscles in order to produce the alteration in the envelope, or stroma, 
on which the action depends, has not been exactly determined. But 
that some of the intracorpuscular water, in addition to extra cor- 
puscular water, must be removed, is indicated by the following 
experiments.! 


Dog’s defibrinated blood was washed free from serum constituents with 0.9 
per cent NaCl solution, and the sediment of corpuscles separated as 
thoroughly as possible from the salt solution, by long-continued centri-. 
fugalization in the cold. The sediment was then partially dried in a flat 
dish over sulphuric acid in vacuo, in a cold room. ‘The dried crust was 
broken through and sonie of the still moist sediment removed. A portion 
of it was at once weighed in a platinum dish, then dried at 110° C. till it 
ceased to lose weight. The percentage of water in the partially dried sedi- 
ment was thus arrived at. Another portion was suspended in 0.9 per cent 
NaCl solution, to see whether the corpuscles would remain unlaked, as nor- 
mal corpuscles do in this solution. It was found that in two experiments 
many of the corpuscles remained unlaked for several hours in the salt 
solution, although the proportion of water in the sediment in one of the 
experiments was only 50 per cent and in the other 56 percent. The per- 
centage of water in the colored corpuscles is usually given as about 60 
per cent. There must certainly have been some extra corpuscular water 
in the sediment, so that on the average the corpuscles must have lost 
some water. This would necessarily happen as the concentration of the 
salt solution increased during the drying. But some of the corpuscles 
were laked and crystals of haemoglobin deposited in the partially dried 
sediment, and some of the water belonging to these corpuscles might 
have been distributed among those still intact. That the reduction of 


1 Performed by Dr. Stewart. 


444 Charles Claude Guthrie. 


water in the sediment to 50 per cent was not without injurious effect, 
even on the unlaked corpuscles, was shown by the fact that they laked 
completely in salt solution, on standing over-night in the cold. 


When gradual drying is allowed to take place, as in a small drop 
of blood or of washed suspension of corpuscles, on a slide covered 
with a slip, it may be seen, as Dr. S. Peskind pointed out to me, that 
the corpuscles at the edge of the drop become markedly swelled. 
Dr. Stewart has demonstrated to me that there is also extensive 
laking here, even in the case of corpuscles which at the moment of 
laking are floating in the liquid. It may be that drying of a portion 
of the circumference of a corpuscle (the outer edge in the case of 
corpuscles at the edge of the drop), or the action on it of the concen- 
trated salt solution produced by evaporation, may render the rest of 
the circumference more freely permeable to the salts of the solution 
or to water, and thus give rise to water laking. My own observa- 
tions on the drying of uncovered films led me to suppose that nearly 
all the water must be gotten rid of, if all the corpuscles are to lake 
in the solutions investigated. 

11. The nuclei of dried nucleated red corpuscles are eres by 
solutions which lake the corpuscles. The action is more energetic 
in the solutions of electrolytes investigated than in the non-electro- 
lytes, and bears the same relation to strength of solution as the 
laking action does. 

12. Dried corpuscles kept at room temperature for as much as 
twenty-eight days show no appreciable difference from freshly dried 
corpuscles, as regards these reactions. 

13. The white corpuscles were observed to behave much like the 
nuclei of nucleated red corpuscles, except that they are more slowly 
acted upon than the nuclei. 

14. On drying blood and suspending the resulting powder in ether, 
and then cautiously adding water, agglutination of the corpuscles 
occurred, this being shown by the formation of a sticky precipitate 
which adhered to the bottom of the test tube. It was also observed 
through the microscope by means of a hollow-ground slide. Laking 
takes place after agglutination. It also occurs when corpuscles 
thoroughly washed with salt solution are dried and treated in the 
way described. It follows that this agglutination of dried corpuscles 
is not dependent on any organic serum constituent. It occurs in 
direct ratio to the percentage of water present. It is known that 


The Laking of Dried Red Blood-Corpuscles. 445 


the addition of water to normal €orpuscles causes agglutination before 
laking. Apparently, then, although the envelope has been profoundly 
altered by drying in regard to its permeability to the solutions inves- 
tigated, it still retains those constituents, or those properties, on 
alteration of which agglutination depends. 

I shall not, at present, discuss the nature of the change which dry- 
ing produces in corpuscles, and on which the reactions described, 
depend. Dr. Stewart has shown that fixing washed corpuscles by 
pouring them into boiling salt solution greatly increases their per- 
meability to ions.’ It is very probable that this is true also for 
neutral molecules, and, no doubt drying at a temperature below that 
necessary for fixation, profoundly alters the permeability of the 
corpuscles for diffusible substances, while it allows of the escape of 
hzemoglobin when enough water to dissolve it is present. 

There is one practical application of the results on which a few 
words may be permissible. 

In the accessible literature on the microscopic examination of blood- 
stains for corpuscles, it is usually recommended to add various solu- 
tions to the stains. For instance, Reese? gives a list of six reagents 
to be used in the examination of blood-stains, viz. : 


1. Water. 
2s eEoOtassium»nhydroxid’ << S., 4. =. 1 part. 
INVIATE Eres ect a ces ee al oe CR 2 parts. 
3. Miiller’s fluid: — 
Potassium dichromate .. . . 1 part. 
Sodium:sulphate’. =). >. 3 a= 2 parts. 
WACK 9.0 aes wero sn oe TODA ES2 


4. Miiller’s fluid. 
Glycerol, q. s. 
The glycerol is added until a specific 
gravity of 1028 is obtained. 


peenGhgeerol®) 20k. wes as lew ] part. 
ae aie Si Wee fs Ae” a ee as 6 parts. 

Ga sodinmrsulphateres eo eae 6 parts. 
NERC Ce ae See Ed Soe, a ene Ae § parts. 


Now all of these, except the second (Virchow’s fluid, 30 per cent 
KOH), produce rapid laking, though the solutions containing potas- 
sium dichromate (Nos. 3 and 4), leave the ghosts stained fairly well. 


1 Journal of medical research, 1902, viii, p. 273. 
2 REESE: Text-book of medical jurisprudence and toxicology, 5th ed., 1898, 
p- 139. 


446° Charles Claude Guthrie. 


Simon ! recommends the treatment of blood-stains with water for the 
- microscopical detection of corpuscles. Sometimes it is vaguely sug- 
gested that, under certain conditions, the corpuscles are liable to 
undergo changes, although it seems rarely to be suspected that the 
liquids with which they are in contact are responsible for the changes. 
For instance, “ If the stain upon the garment has been moist, or sub- 
merged in water for a considerable time, decomposition may have led 
to the dissolution of the corpuscles,’ and again, ‘‘ Solutions of the 
same specific gravity of the blood, or lighter, will lead to the eventual 
swelling up and dissolution of the corpuscles.” ? 

In the light of the results given above, such statements need 
revision. 

Directions to wash blood-stains in water, or solutions of the same 
specific gravity as blood, if followed, will in general give a suspension 
of “ghosts” or “shadows” of red corpuscles which may easily be 
overlooked, unless staining methods are subsequently resorted to, 
and then it is difficult to bring them out distinctly. 

In any event it appears more reasonable to treat the stain with one 
of the non-laking solutions. 

By scraping dried blood-stains on cloth above-a slide, thus causing 
the corpuscles to fall upon it, and then laying ona cover slip and 
applying a solution of eosin in 96 per cent alcohol, good results were 
obtained. 

Before the discovery of the biological blood test, numerous writers 
laid stress on the importance of measuring the average diameter of 
the corpuscles in blood-stains. As a matter of fact, the corpuscles 
in an ordinary dried stain on cloth are distorted to a greater or lesser 
degree, depending largely on the rapidity with which they are dried, 
and it is difficult to find any which show no evidence of crenation.. 
Measurements, therefore, would seem to be of small value, as it is 
difficult to suppose that the corpuscles will swell up and assume their 
normal dimensions if placed in solutions of the same specific gravity 
as the blood, or that the same amount of contraction occurs, in all 


cases, in drying. 


' Simon: Manual of chemistry, 1901, p. 511. 
4 REESE x: £00. c#,, pp. 130i, 12. 


ON THE NUCLEOPROTEIDS OF THE PANCREAS, 
THYMUS, AND SUPRARENAL GLAND, WITH 
ESPECIAL REFERENCE TO THEIR 
OPTICAL. “AGCAIVITY. 


By ARTHUR GAMGEE anp WALTER JONES.! 


LL observations hitherto published concerning the optical 

activity of the albuminous bodies have led physiological chem- 

ists to believe that the bodies thus designated, whether derived from 

the vegetable or animal kingdom, without a single exception, deviate 

the plane of polarization to the left, no case having been hitherto 

made known either of a dextrogyrous, a racemic, or an otherwise 
inactive albuminous substance. 

There is one group of albuminous substances which (whether we 
consider them from the standpoint of their paramount physiological 
importance, or from that of their startlingly interesting physical 
properties, or lastly from that of their chemical relationships) pos- 
sesses quite exceptional interest, and which has, unaccountably, been 
allowed to remain uninvestigated (in so far as optical activity is con- 
cerned) until one of us, very recently, initiated an inquiry which, at 
its very threshold, revealed an unexpected and, in a sense, startling 
result. The group to which we refer is that which has_ been 
designated by German writers the “ Proteide.” This group com- 
prises those highly complex albuminous substances which can, with 
greater or less ease, be split up into, or which yield as products of 
decomposition, on the one hand, albuminous bodies, and on the other, 
such bodies as coloring matters, or nucleinic acids and the nuclein 
bases, or, in certain cases, carbohydrates. The best characterized 
and most striking members of this group are: first, hamoglobin 
and its compounds; secondly, the nucleoproteids. 

In hemoglobin we have the example of a complex proteid which 
differs from all other albuminous bodies by its color, by its marvellous 
power of forming easily dissociable compounds with oxygen and with 


1 Read before the American Physiological Society, December 30, 1902, at 
Washington, D. C. 
447 


448 Arthur Gamgee and Walter Jones. 


certain other gases, by the facility with which it admits of being crys- 
tallized and recrystallized and obtained free from all foreign mineral 
impurities, and by the startling manner in which its solutions fail to 
furnish any one of the reactions characteristic of albuminous sub- 
stances in solution, so long as the reagent has not effected a fundamen- 
tal decomposition which has liberated the albuminous and colored 
residues. The researches of one of us! had, moreover, lately demon- 
strated that although hemoglobin is definitely a diamagnetic body, 
the iron-containing products of its decomposition by acids are not 
merely paramagnetic, but actually the most powerfully ‘ ferromag- 
netic” organic bodies known to science. So complete a divergence 
was thus shown to exist in every physical and chemical property be- 
tween haemoglobin and the substances which are linked together in 
it, that it appeared to be in the highest degree interesting to ascer- 
tain whether, or not, in respect of optical activity, haemoglobin 
would behave as an albuminous substance proper and prove to be 
“lgevogyrous.” 

In a research in which one of us (A. G.) was associated with 
Dr. A. Croft Hill, and of which the results have not yet been pub- 
lished, it has been discovered that haemoglobin is a dextrogyrous 
body possessing a specific rotation for monochromatic red light 
having the mean wave length of Frauenhofer’s line C of 10.8°. On 
the other hand, the interesting histon-like albuminous substance 
globin, of which the leading characters have only been known since 
the researches of Schulz, is a normally levogyrous albuminous body 
with a specific rotation —69°. 

These interesting observations naturally suggested the probability 
that the nucleoproteids might, like haemoglobin, prove to be dextro- 
gyrous, and the research of which the first results are related in this 
paper is the outcome of this idea. The hypothesis which led to its 
being undertaken, has been confirmed, as will be shown in the sequel, 
and it has thus been proved that, contrary to the traditional belief, 
some at least of the members of a group of albuminous bodies of sig- 
nal importance in the life-history of the organism are dextrogyrous 
bodies. 


' GAMGEE: On the behavior of oxy-hamoglobin, carbonic-oxide-hamoglobin, 
methemo-globin, and certain of their derivatives, in the magnetic field, with a pre- 
liminary note on the electrolysis of the haemoglobin compounds. Proceedings of 
the Royal Society, 1901, Ixviii, p. 503; The Croonian Lecture: On certain 
chemical and physical properties of haemoglobin. Proceedings of the Royal 


Soc iety, 1902, Ixx, p- 79. 


I EE oo 


Nucleoproterds of the Pancreas, Thymus, etc. 449 


THE NUCLEOPROTEID OF THE PANCREAS.! 


The preparation of nucleoproteids that can be made into solutions 
sufficiently colorless and transparent for polarimetric work is a matter 
that is attended with considerable difficulty. In an endeavor to 
improve in this respect upon the methods already known (and this 
was found absolutely necessary in the very beginning of the work), 
we found a very interesting optical relation between the nucleo- 
proteids proper and their first hydrolytic products, the nucleins. 
We therefore give the method in some detail. 

The finely divided pancreas of the pig was treated in turn with 
50 per cent alcohol, 75 per cent alcohol, and 95 per cent alcohol, and 
finally dried with absolute alcohol and ether. The material thus 
obtained was extracted with several successive portions of a 5 per 
cent solution of ammonium acetate, the united extracts were filtered, 
and the perfectly clear fluid was poured into four times its volume 
of weak alcohol. The precipitate thus formed was washed by decanta- 
tion with an excessive amount of dilute alcohol, and finally dried with 
absolute alcohol and ether. The object of this procedure was to 
remove the coloring matter of the gland which is somewhat soluble 
in dilute alcohol, more so in an alcoholic solution of ammonium 
acetate, but soluble to a very slight extent in an aqueous solution of 
ammonium acetate. The manipulation also removes a large amount 
of inorganic salts, and renders the coagulable proteids insoluble. A 
2 per cent filtered aqueous solution of this material had only a pale 
yellow color, and could easily be polarized in a 220-mm. tube. The 
result of this optical examination showed that there was present a 
dextrorotatory substance. The solution, moreover, failed to give any 
indication of a reducing substance, even by prolonged boiling with 
Fehling’s solution, and was shown to be rich in material which pro- 
duces xanthine bases on hydrolysis with sulphuric acid. 

The main portion of this material was dissolved in twenty parts of 
water, and the filtered solution treated with dilute acetic acid a drop at 
atime. When a quantity of acid had been added sufficient to bring 
the entire solution to about 1 per cent, a well-defined white flocculent 


1 We have used the term “nucleoproteid”’ in certain instances to include all 
compound proteids which yield nucleic acids on hydrolysis, even though the sub- 
stances so designated be nucleins or nucleohistones. In other cases we have used 
the term in contradistinction to nuclein, but we have taken some care that the con- 
text shall show the exact meaning in each instance. 


450 Arthur Gamgee and Walter Jones. 


precipitate was produced. This precipitate, which for the present 
we will call nucleoproteid, was separated by the centrifuge, suspended 
‘in water, and treated with an extremely dilute solution of ammonia, 
drop by drop, as the reaction of the fluid was carefully noted with 
litmus. A very small amount of alkali was required to neutralize 
the adherent acetic acid when the solution became neutral, and 
remained so until approximately twice as much ammonia had been 
used as was required to completely dissolve the nucleoproteid. 
Evidently the nucleoproteid is at least a dibasic acid whose acid 
ammonium salt is soluble in water and neutral to litmus. Purifica- 
tion was effected by alternate solution in ammonia and _precipita- 
tion with a minimal amount of acetic acid. The final solution was 
poured into five volumes of 95 per cent alcohol, washed repeatedly 
by decantation with excessive quantities of 95 per cent alcohol, 
finally dried with absolute alcohol and ether, and placed in a 
desiccator with sulphuric acid. 

The fluid from which the nucleoproteid was originally precipitated 
was treated with 20 per cent acetic acid, a drop at a time. When the 
fluid contained 2 per cent of the acid, not the slightest precipitation 
had occurred. Continued addition of acetic acid, however, soon 
causes a turbidity, and when the acidity reaches 5-6 per cent, a well 
defined flocculent precipitate occurs. This precipitate, which we will 
for the present call nuclein, was separated in the centrifuge, and, at 
a great sacrifice of material, was washed twice in the centrifuge with 
small portions of water. The nuclein was suspended in water, and 
ammonia added in small portions, a drop at a time; when the nuclein 
was completely dissolved, the fluid was still acid to litmus. This 
solution was poured into four volumes of 95 per cent alcohol, and the 
precipitated nuclein washed and dried by the method described for 
the nucleoproteid. 

The fluid from which the nuclein was precipitated by a great excess 
of acetic acid, was poured into four volumes of alcohol, and the pre- 
cipitate washed and dried with alcohol and ether. This preparation, 
which is necessarily very impure and especially rich in inorganic 
salts, will be designated as “residual material.” 

Thus, by functional precipitation with acetic acid in the presence 
of inorganic salts, we have gotten possession of three preparations. 
The nucleoproteid is almost insoluble in water, but may be dissolved 
by the addition of ammonia or caustic soda. The nuclein is soluble 
in water with the greatest ease. By the addition of a trace of copper 


~ 


Nucleoproterds of the Pancreas, Thymus, etc. 451 


sulphate to a solution of the nugleoproteid in caustic soda, a fine pink 
color is produced, but not a shade of violet makes its appearance until 
a comparatively large amount of copper solution has been used, a 
reaction which resembles closely the biuret test with proteoses. The 
nuclein by similar treatment gives only the faintest pink color, the 
violet shade being observed even when a very small amount of copper 
sulphate is used, while the “ residual material” produces a violet color 
from the beginning. 

It has lately been shown by one of us that the nucleoproteid yields 
two xanthine bases, guanine and adenine, and in a ratio which closely 
approximates four equivalents of the former to one of the latter.! 
The nuclein and residual material were also shown to yield xanthine 
bases on hydrolysis with sulphuric acid, and it may be conveniently 
stated here that all substances which we have described were found 
to yield, on complete bydrolysis, products which respond to the xan- 
thine color reaction, and which form compounds with silver nitrate 
which are insoluble in ammonia. All three of the preparations under 
discussion contain phosphorus, all are completely precipitated from 
aqueous or faintly alkaline solutions by the addition of a trace of 
hydrochloric acid, and all yield precipitates on boiling in neutral 
solution. 

We believe sufficient evidence has been cited to justify the terms 
nucleoproteid and nuclein, and we may add that, in our opinion, there 
is contained, in the preparation called “residual material,” a nuclein 
still more closely related to a nucleic acid. In all of our experiments 
we have noticed that any property which is more marked in the 
nuclein than in the nucleoproteid finds still greater expression in the 
residual material. This was found well illustrated by a study of their 
action on polarized light. All three preparations are dextrorotatory. 
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°. 


OPTICAL PROPERTIES. 


A weighed amount of the nucleoproteid was suspended in water, 
and dissolved by the addition of a trace of ammonia. The solution 
was then made up to a definite volume with water, and polarized. 


1 JonEs and WHIPPLE: This journal, 1902, vii, p. 423. 


A52 Arthur Gamgee and Walter Jones. 


Weight of material (w). .. .. . . 1.0064gm.,, 

Volume of solution (2) 9-5) 20d ae econe ce 

@bserved'anele (a) 3 9) ne eee oes 

Wength oftubes(2)) ae) = mee ee eee eee acne 
(Ones Soke 


A similar experiment with another preparation of the nucleoproteid 
gave the following data: 


Wieightiof material. (once es eee tO nOUs ome 
Volunieiof solution) (5.1 5 ee nee ecole ce 
Observed/angle, =. 6 S90) eee loro Ue 
enethiot tuber) wurst sae ae depen nme 
a7 ane x 
= — => EBS 
(> lw 


The solution was treated with an excess of acetic acid and the 
precipitate filtered off. The filtrate was found inactive. 

A solution of a preparation of the nucleoproteid which was known 
to be somewhat contaminated with nuclein was polarized in a 200-mm. 
tube and gave a reading of 1° 34’. Acetic acid was added and the 
precipitate filtered off. The filtrate, polarized in a 200-mm. tube, gave 
a reading of 22’. The solution was then treated with hydrochloric 
acid, when the filtrate from the precipitated nuclein was found 
inactive. 

We gained good evidence that the specific rotation of the nucleo- 
proteid would be found less than that of the nuclein before we had 
made quantitative experiments with the latter substance. A perfectly 
neutral solution of nucleoproteid was prepared by treating some of the 
material with water and an insufficient amount of ammonia to effect 
complete solution. The filtered fluid, polarized in a 200-mm. tube 
gave a reading of +1° 46’. The solution was heated to boiling and 
the coagulum filtered off. The filtrate, polarized in a 200-mm. tube, 
gave a reading of 1° 49’. It is well known that boiling converts nucleo- 
proteids into nucleins. As the length of tube and angle of rotation 
remained constant, a decrease in the amount of material must mean 
an increase in the specific rotation. 

The following observations were, therefore, not surprising. A solu- 
tion of the nuclein was made in water, and as the fluid was somewhat 
colored, it was necessary to use a dilute solution and a short tube. 


Weight of material .. 7 a 9) ye OOS2 em, 
Volume of solution) > 2.55) eee ore 
Observed angle 9. 4 G i9% +, 5. Peel ae 
Length of tube: +s, < (en a ele eR 
az 
(a), pale 1-64.4,° 


Nucleoproterds of the Pancreas, Thymus, etc. 453 


The solution was treated with hydrochloric acid, and the fil- 
tered fluid polarized in a 200-mm. tube. The rotation was slightly 
negative (—9’).} 

As is the case with the nucleoproteid, a solution of the nuclein 
yields a coagulum on heating, and the rotation of the solution is not 
appreciably changed. This would lead us to assume the existence 
of a nuclein whose specific rotation is greater than +64.4°. It can 
easily be proved that such a substance exists in the preparation which 
we have designated as “residual material.” A weighed amount of 
this material was dissolved in a measured volume of water. The 
solution was polarized, treated with hydrochloric acid, and the amount 
of solid matter determined in the optically inactive filtrate. The 
following data were obtained: 


Wieisht of materaljused:-= - sa = 2s ue 055207 om, 

Weight of optically inactive material . . . 0.2691 gm., 

Weight of optically active material . . . . 0.2516 gm., 

Volumeor the salution yi.) 0h <li #11 1. pee 

Mbservedtancle; 2.) fies) sole 2s OG 

Rencthomtube: 0.505 <p koa <) cole =~ 200 mm: 
@,=7, = 0u- 


On HAMMARSTEN’S PREPARATION. 


The results which have already been detailed made an optical ex- 
amination of this substance highly desirable. By slight departures ” 
from the method described by Hammarsten, which were absolutely 
necessary to remove coloring matter, and which cannot possibly have 
exercised any great influence on the chemical nature of the product, 
we were able to prepare a nuclein which must be identical with Ham- 
marsten’s preparation. From its method of preparation, we would 
imagine that the substance is a nuclein. The material which we 
obtained gave a violet biuret reaction, and is soluble in water. Its 
solution was highly colored, but possessed so great a rotatory power 


1! We noticed several times that very slightly levorotatory filtrates were 
obtained when hydrochloric acid was used for precipitating the proteid, and 
especially so when the acid fluid was allowed to stand in contact with the precip- 
itate. It is probable that the negative rotation is due to slight hydrolysis, in which 
an optically negative proteid is formed which is soluble in hydrochloric acid. 

2 We used ammonia for redissolving the nuclein, where HAMMERSTEN used an 
alkali. We also finally poured an aqueous solution of the proteid into 95 per cent 
alcohol, and washed by decantation instead of washing on a filter. 


454 Arthur Gamgee and Walter Jones. 


that fairly satisfactory polarimetric observations could be made in 
solutions of extreme dilution. The proteid is dextrorotatory. 


Weightofmatenialy= 26 yout.) cee) sheen 00M oer. 
Volumelofthe:solution! 3)en . 2 eee, eee Pa ICice 
Observed angle (the mean of eight readings) . . +477, 
IB esorninnorss & ane Stel oe Ble fol: 100 mm. 
av 
(<4), = ia Dieoee 


On THE NUCLEOHISTONE OF THE THYMUS. 


It would seem quite easy to obtain this substance in any desired 
quantity by the very simple method which Lillienfeld describes in 
twenty lines. This method leads to a product whose solutions are 
highly opalescent, and an optical examination is not to be thought of. 
The cloudiness is so persistent that for a long time we were inclined 
to believe it a property inherent in the proteid. We finally succeeded, 
however, in obtaining solutions almost as colorless and transparent 
as distilled water. 

It is only necessary to extract Lillienfeld’s preparation with a 5 per 
cent solution of ammonium acetate and filter. The fluid filters very 
slowly, but perfectly clear and continuously. The solution was 
poured into 95 per cent alcohol, and the precipitated proteid washed 
and dried with alcohol and ether, as described in connection with 
other preparations. The substance thus obtained was submitted to 
an optical examination with the following results. The solution was 
made in very dilute ammonia. 


Weight of material 2.023 gm., 
Volume of the solution . 50 ce., . 
Observed angle. 320; 
Length of tube . 220 mm 
az a 27 =O | 
(a), he olso 


ON THE NUCLEOPROTEID OF THE SUPRARENAL GLAND. 


| 

. 
As is well known, the characteristic physiologically active con- { 

stituent of this gland forms a dark brown pigment when exposed in 

aqueous solution to the oxidizing action of the air. Aqueous extracts , 

of the gland are, therefore, always highly colored, and this coloring 

matter offers great difficulty in the preparation of substances from 

the glands which are intended for optical examination. While the 


Nucleoprotetds of the Pancreas, Thymus, etc. 455 


work on the nucleoproteid is, ‘therefore, not as satisfactory as we 
desire, it can nevertheless be stated positively that the nucleoproteid 
is dextrorotatory. 

One of us has lately described this substance, and shown that it 
yields of the xanthine bases only guanine and adenine. These two 
bases (together with thymine) are to be found among the hydrolytic 
products of the nucleoproteid in a molecular ratio which closely 
approximates 4: 1.! 

The method of isolation does not essentially differ from that 
formerly employed, except that the gland was extracted several times 
with acetic acid before removing the nucleoproteid. A substance 
was finally obtained which was too much colored for accurate polari- 
metry, but which, even in the necessarily high dilution employed, 
could easily be shown to be dextrorotatory. 


Weight of material . . . ee tees Ose eM, 
Eengthioftabe: =) « eet 3). 2 LOOtmmn.; 
Volume of solution Fee Ree ee 5 CC 
Observed angle 230. 
az 
ae — ry 2710 
(a), = Zo 48.1 


The value of this rotation is liable to revision, but its direction is 
beyond question. 


CONCLUSION. 


We have described six preparations, obtained from various glands, 
and have given methods by which several of these may be isolated 
sufficiently free from coloring matters for exact polarimetric work. 
All six of these substances yield, on hydrolysis, phosphoric acid and 
purine derivatives, and are therefore to be considered as _ nucleo- 
proteids in the wide sense of the term. 

The methods of preparation were such as to exclude all dextrorota- 
tory substances that are not of a proteid nature, and all preparations 
were shown to be free from substances which reduce Fehling’s solu- 
tion, even on prolonged boiling. Nevertheless all of these prepara- 
tions were found to be dextrorotatory, having specific rotations which 
vary from +37.5°, that of the nucleohistone of the thymus, to +97.9°, 
that of Hammarsten’s nuclein of the pancreas. 


1 JONES and WHIPPLE: Loc. cit. 


ERRATA. 


Page 337, line 23, for gm. read mg. | = 


Page 346, line 27, for 25 per cent; read 2.5 per cent, 
a . = : 


, 


—— 


IN DEAS LOs VO. VIE. 


BEL, J. J. On the behavior of ex- 
tracts of the suprarenal gland toward 
Fehling’s solution, xxx. 

ABEL, J. J. On the elementary composi- 
tion of adrenalin, xxix. 

ABEL, J. J. On the oxidation of epinephrin 
and adrenalin with nitric acid, xxxi. 

Absorption after nephrectomy, xlii. 

Adrenalin, action upon blood-vessels, xlii. 

, composition, xxix. 

——,, oxidation, xxxi. 

Alcohol, action on muscle, 61. 

AupricH, T. B. See HouGuTon and 
ALDRICH, xviii. 

AMBERG, S. The toxicity of epinephrin 
(adrenalin), xxxiii. 

Ammonia, determination in urine, 330. 

Artificial respiration in strychnine poison- 
ing, xlii. 

ATWATER, W. QO. On the sources of mus- 
cular energy, xlii. 


ACILLUS coli communis, chemical 


products, 284. 
Bacillus lactis aerogenes, chemical products, 


284. 

Blood, electrical potential, xliii. 

, laking, 441, xliii. 

Blood-corpuscles and other cells affected 
by acids, 99, 404. 

Blood-corpuscles, haemolysis, 103. 

Blood-pressure, in salmon, xlii. 

BowEN, W. P._ Exhibition of mercury- 
mercury stimulator, xx. 

Bowen, W. P. Exhibition of new form of 
platinum-mercury stimulator, xx. 

Brain protagon, 183. 


HOR, influence upon excretion of 
dextrose in diabetes, xxxil. 

CANNON, W. B. Further observations on 
the movements of the stomach and intes- 
tines, xxi. 

CANNON, W. B. Demonstration of the 
movements of the stomach and intestine, 
observed by means of the Rontgen rays, 
xli. 


CANNON, W.B.,and H. F. Day. Salivary 
digestion in the stomach, xxviii. 

CARLSON, A. J. See JENKINS and CARL- 
SON, 251. 

Cannula, tracheal, xxiii. 

Cardio-pneumatiscope, xxii. 

Coagulation of milk, xxxv. 

CusHING, H. The effect of extirpation of 
the Gasserian ganglion upon the sense of 
taste, XXvil. 


De H. F. See CANNON and Day, 
XXViil. 


Diabetes, xxxii. 

Development of eggs affected by mechan- 
ical shock, 301. 

Diaphragm, apparatus for illustrating its 
action, xlil. 

Digestion, salivary, in stomach, xxviii. 

Dopcr, R. Five types of eye movement 
in the horizontal meridian plane of the 
field of regard, 307. 


LECTRICAL convection of cells and 
nuclei, 273. 
Electrical polarity in hydroids, 294. 
Electro-magnetic signal, xxii. 
Energy, muscular, source, xlii. 
Epinephrin, xxxiii. 
, oxidation, xxxi. 
Eye movements in horizontal meridian 
plane of field of regard, 307. 


penees human muscle, 355- 
Fertilization, relation to development, 


430: 
FiscHEerR, M. H. How long does (arbacia) 


sperm live in sea-water? 430. 

FRANZ, S. I. On the functions of the cere- 
brum: I. The frontal lobes in relation to 
the production and retention of simple 
sensory-motor habits, 1. 

Frog, long survival after removal of cere- 
bral hemispheres, xlii. 

Frog-table, xxii. 

Frontal lobes, localization of habits, 1. 


457 


458 


AMGEE, A., and W. Jones. On the 
nucleoproteids of the pancreas, thy- 
mus, and suprarenal gland, with especial 
reference to their optical activity, 447, 
x]ii. 

Gasserian ganglion, extirpation, xxvil. 

Gastric secretion, stimulated by alcohol, 
etc., XVil. 

Gelatine, digestion, xxiii. 

Gigs, W. J. The influence of the H ion in 
peptic proteolysis, xxxiv. 

Gigs, W. J. A proteid reaction involving 
the use of chromate, xv. 

Gigs, W. J. Further mucoid studies, xiii. 

GiEs, W. J. See LesEM and Gixs, 183. 

GiEs, W. J. See Loren and Gis, xiv. 

GirEs, W. J. See MEL?zZER and GIEs, xlii. 

Glucophosphoric acid, xi. 

Glucothionic acid, xi. 

Glucuronic acid, origin, xiii. 

GorRHAM, F. P., and R. W. Tower. Does 
potassium cyanide prolong the life of the 
unfertilized egg of the sea-urchin ? 175. 

GREENE, C. W. Arterial blood-pressure 
in the Sacramento salmon, Oncorhynchus 
tschawtscha, xlii. 

GREENE, C. W. Osmotic changes in the 
Sacramento salmon during the run from 
the sea to the spawning beds, xlii. 

Growth on diet of skimmed milk, 197. 

Gururik£, C.C. The laking of dried red 
blood-corpuscles, 441, xliii. 


ABITS, localization, 1. 
Hemolysis, 103. 

HATCHER, R. A., and T. SOLLMANN. The 
effect of diminished excretion of sodium 
chloride on the constituents of the urine. 
139. 

Heart lever, xxxvili. 

ITeart, maximal 


contraction, “staircase ” 
contraction, refractory period, compen- 
satory pause, 213. 

Ifeart muscle, affected by sodium chloride, 75. 

, tonus, xxvi. 

Houcuron, E. M., and T. B. ALDRICH. 
A preliminary report on the pharmaco- 
logical and chemical properties of tri- 
brom-tertiary butyl-alcohol, xviii. 

Human muscle, fatigue in voluntary con- 
traction, 355. 


Hydroids, electrical polarity, 294. 


NDUCTORIUM, xxv. 
Intestine, movements, xxi, ]xi. 
lons, toxic and antitoxic effects, xiv. 


Index. 


ACKSON, H. C. On the influence of 
camphor ingestion upon the excretion 
of dextrose in phlorhizin diabetes, xxxil. 

Jackson, H. C. See MANDEL and JACK- 
SON, Xlii. 

Jackson, H.C. See WALLACE and JACK- 
SON, XVii. 

JENKINS, O. P., and A. J. CaRLson. The 
rate of nervous impulse in certain mol- 
luses, 251. 

Jennincs, H. S. Studies on reactions to 
stimuli in unicellular organisms. IX. On 
the behavior of fixed infusoria (Stentor 
and Vorticella), with special reference to 
the modifiability of protozoanreactions, 23. 

Joints, flexion and extension, xxiv. 

Jones, W. On the xanthin bases of the 
suprarenal gland, xlil. 

Jones, W. See GAMGEE and JONES, 447, xli. 

| EY, mercury, xxii. 

Kidney, glomerular fluid by filtration 
or secretion, 155. 
Kymograph, xxxvii. 


ABORATORY apparatus, xxii, Xxxv. 
LEn, Ey S..and. W. SATANT sain 
action of alcohol on -muscle, 61. 

LEE, F. S. The action of ethyl-alcohol on 
contractile protoplasm, xix. ; 

LEsEM, W. W., and W. J. Gigs. Notes 
on the “ protagon” of the brain, 183. 

LEVENE, P. A. On glucophosphoric acid, 
cele 

LEVENE, P. A. On glucothionic acid, xi. 

LEVENE, P. A. On nucleic acid, xii. 

LEVENE, P. A., and L. B. SrooKEY. On 
the biological relation of proteids and 
proteid assimilation, xxiii. 

LEVENE, P. A., and L. B. SrooKEy. On 
the digestion of gelatine, xxiii. 

Life, duration affected by potassium cya- 
nide, 175. 

LILLIE, R. S. On differences in the direc- 
tion of the electrical convection of certain 
free cells and nuclei, 273. 

LINGLE, D. J. The importance of sodium 
chloride in heart activity, 75. 

Lorn, J.,and W. J. Girs. Further studies 
of the toxic and antitoxic effects of ions, 
Xiv. 

LOEVENHAR?T, A. S. Some observations 
on the coagulation of milk, xxxv. 

LoMBARD, W. P. The action of the two- 
joint muscles of the hind-leg of the frog, 
with special reference to the spring move- 
ment, Xxiv. 


Index. 


ee , A. B. On the origin of 
1 the relation of the inorganic ele- 
ments to protoplasm, xlii. 

MANDEL, J. A., and H. C. JAcKson. On 

_ the origin of glycuronic acid, xiii. 

MATHEWS, A. P. Electrical polarity in the 
hydroids, 294. 

MATHEWS, A. P., and B. R. WHITCHER. 
The importance of mechanical shock in 
protoplasmic activity, 300. 

Mechanical shock, effect on protoplasm, 
300. 

MELTZER, CLARA. 
MELTZER, xlii. 
MELTZER, S. J., and W. J. Girs. Studies 
on the influence of artificial respiration 
upon strychnine spasms and respiratory 

movements, xlii. 

MELTZER, S. J.,.and CLARA MELTzER. The 
vasomotor influence of the third cervical 
nerve (auricularis magnus) upon the cir- 
culation in the rabbit’s ear, xlii. 

MELTZER, S. J.,and CLARA MELTZER. The 
effect of the intravenous injection of 
adrenalin upon the blood-vessels of the 
ear when deprived of their vaso-constric- 
tors, xlii. 

MEL?ZER, S. J.,.and CLARA MELTZER. The 
unmistakable vasomotor influence of sub- 
cutaneous injection of adrenalin, xlii. 

MELTZER, S. J.,and W. SALANT. - The in- 
fluence of nephrectomy upon absorption, 
xlii. : 

MENDEL, L. B., and F. P. UNDERHILL. 
New experiments on the physiological 
action of the proteoses, xvi. 

MENDEL, L. B., F. P. UNDERHILL, and 
b. Wuirr. <A_ physiological study of 
nucleic acid, 377. 

Metabolism, in salmon before spawning, 
xlii. 

Milk, coagulation, xxxv. 

, skimmed, food value, 197. 

Moist chamber, xxii. 

Molluscs, rate of nerve impulse, 251. 

Mucoid, xiii. 

Muscle, action of alcohol upon, 61. 

Muscles, classification as flexors and exten- 
SOrs, XXiv. 

—, second maximum in response to a 
series of stimuli, xxiv. 

—, heart, affected by sodium chloride, 
75: 

, human, “ gap ” in contraction, 435. 

, human, stimulation with induced cur- 


rents, 355- 


See MELTZER and 


459 


Muscles, plain, tone affected by salts, 269. 
Muscle lever, heavy, xl. 

Muscle tonus, xxvi. 

Muscles, action upon joints, xxiv. 
Muscular energy, source, xlii. 


EPHRECTOMY, influence upon ab- 
sorption, xlii. 
Nerve impulse, rate, 251. 
Nucleic acid, xii. 
, action and fate in body, 377. 
Nucleo-proteids, 447, xli. 
——,, optical properties, 447, xli. 


XYHAMOGLOBIN, inhibition of 
crystallization, xlii. 
——,, quick crystallization, xlii. 


EPSIN, 
XXXiV. 
PESKIND, S. Notes on the action of acids 
and acid salts on blood-corpuscles and 
some other cells, 99. 

PESKIND, S. The action of acids and acid 
salts on blood-corpuscles and other cells, 
404. 

Porter, W. T. New inductorium, kymo- 
graph, heart lever, heavy muscle lever, 
and square rheochord, xxxy. 

Porter, W. T. The tonus of heart muscle, 
XXVi. 

Potassium cyanide, prolongation of life, 175. 

Potential, differences between blood and 
serum, etc., xliii. 

Proceedings of the American Physiological 
Society, ix. 

Protagon of brain, 183. 

Proteid metabolism, specific characters, 
XXiil. 

Proteid, reaction with chromate, xv. 

Proteolysis influenced by acids, xxxiv. 

Proteoses, physiological action, xvi. 

Protoplasm, contractility affected by ethyl- 
alcohol, xix. 

, inorganic elements, xlii. 

Protoplasmic activity affected by mechani- 
cal shock, 300. 


action influenced by acids, 


if EICHERT, E. T. Inhibitory phe- 
nomena in the crystallization of oxy- 
hemoglobin, xlii. 

ReicHerT, E. T. Quick methods for the 
crystallization of the oxyhamoglobin of 
the blood of the dog and certain other 
species, xlil. 


RETTGER, L. F. An experimental study 


460 Lndex. 


of the chemical products of bacillus coli 
communis and bacillus lactis aerogenes, 
Bp) 284° 
Rheochord, square, xli. 


ee ae eal W. See LEE and SALANT, 
6r. 

SALANT, W. See MELTZER and SALANT, 
xlii. 

Salivary digestion in stomach, xxviii. 

Salmon, changes in the run to the spawn- | 
ing beds, xlii. 

SHAFFER, P. On the quantitative deter- 
mination of ammonia in urine, 330. 

Sodium chloride excretion in urine, 139. 

Sodium chloride retention, 155. 

SOLLMANN, T. ‘The mechanism of the re- 
tention of chlorides: a contribution to 
the theory of urine secretion, 155. 

SOLLMANN, TI. See HATCHER and SOLL- 
MANN, 139. 

Sperm, duration of life in sea-water, 430. 

STEWART, C. C. Some minor improve- 
ments in laboratory apparatus, xxii. 

STEWART, C. C. The second maximum in 
the response of muscle to stimulation, 
XXiv. 

STEWART, G. N. Differences of potential 
between blood and serum and _ between 
normal and Jaked blood, xliii. 

STEWART, G. N. The action of hemolytic 
agents at 0° C., xliii. 

STEWART, G. N. The behavior of nucle- 
ated colored blood-corpuscles to certain 
hemolytic agents, 103. 

STILEs, P.G. On the influence of calcium 
and potassium salts upon the tone of 
plain muscle, 269. 

Stimulator, mercury-mercury, xx. 

——,, platinum-mercury, xx. 

Stomach, movements, xxi, xli. 

, Salivary digestion, xxviii. 

STooKeEy, L. B. See LEVENE and STooKrEy, 


XXiii. 

SToREY, T. A. The influence of fatigue 
upon the speed of voluntary contraction 
of human muscle, 355. 

STOREY, T. A. Variations in the amplitude 
of the contractions of human voluntary 
muscle in response to graded variations 
in the strength of the induced shock, 435. 

Strychnine poisoning, affected by artificial 
respiration, xlii. ] 


Suprarenal extract, 441, xli. 
, reducing power, xxx. 
Suprarenal gland, xanthin content, xlii. 


aheete fibres, relation to trigeminus, 
XXVil. 

Tone of plain muscle, affected by salts, 269. 

Tonus of heart muscle, xxvi. 

Tower, R. W. See GoRHAM and Tower, 
175. 

Tri-brom-tertiary butyl-alcohol, pharmaco- 
logical and chemical properties, xviii. 


NDERHILL, F. P. See MENDEL and 
UNDERHILL, xvi. 

UNDERHILL, F. P. See MENDEL, UNDER- 
HILL, and WHITE, 377. 

Unicellular organisms, reaction to stimuli, 
A. 

Urine, constituents affected by diminished 
excretion of sodium chloride, 139. 

, determination of ammonia, 330. 


\/ sSenee fibres in third cervical 
nerve, xlil. 
Vasomotor phenomena of ear, influenced 
by adrenalin, xlii. 


ALLACE, G. B., and H. €. JACKSON; 
Is the action of alcohol on gastric 
secretion specific? xvii. 

WuiIrcHeER, B. R. See MatrHews and 
WHITCHER, 300. 

WuitkE, B. See MENDEL, UNDERHILL. 
and WHITE, 377. 

Wivper, B. G. A living frog from which 
the cerebrum was removed December 4, 
1899, xlii. 

WiLper, B. G. An improved apparatus 
for illustrating the action of the dia- 
phragm, xlii. 

WILSON, MARGARET B. On the growth of 
suckling pigs fed on a diet of skimmed 
cow’s milk, 197- 

WoopwortrhH, R.S. Maximal contraction, 
staircase” contraction, refractory period, 
and compensatory pause, of the heart, 
aitey 


xe bodies in-suprarenal gland, 
xlii. - ) 


665 -4 


a a = — a) a o nays _ _~ e 
oe a - ——— =; s eae ee 
er. 1% : 
Be Te ey 
‘ea = : — 
an - 


‘a a W 


‘ by Par Pein Colne Dakine gee 2 OY ee: 

+ Ser Bia ere cm fe Rae 
Pu 4 4 Ot ee Pie es o 
Cee YS re)! 


> i : “Wie . 
— = hy ri 
{ ~ 
Pa Eat 
: = ofS is 
oe 
= ne + . 
Lape 

- — 

ig 

ey, 

citar 

2 a 

WE 

a i 

e-2 Be 

3 ve 
> 
- . 
t est 
i 
»*, 


BINDING SECT. MAR4 1966 


American Journal of Physiolog 


Cope2 19 0 4 
Biological 
& Medical 
Serials 


PLEASE DO NOT REMOVE 
CARDS OR SLIPS FROM THIS POCKET 


UNIVERSITY OF TORONTO LIBRARY 


STORAGE 


Le ae 


ate woe 
tet erat Ne 
Str —. 


io 


“Sp ae ; 


aap a 
Mary 5 men ne 


Airs YS Rema se enna rere 


oe Tomee 


Ressiouen